A NEW SERIES OF PLANT SCIENCE BOOKS

edited by Frans Verdoorn

Volume IX

THE CARNIVOROUS PLANTS

Francis Ernest Lloyd u/as born in 1868 0} Welsh
parentage in Manchester, England, coming to the United
States in 1882. After graduation at Princeton in i8gi, he
taught at Pacific University in Oregon for five years and
was then appointed Associate Professor of Biology at
Teachers College, Columbia University. During this period
of ten years he studied with Goebel {Munich) and Stras-
BURGER {Bonn). In IQ06 he became Investigator, ap-
pointed by the Carnegie Institution of Washington to work
at the Desert Laboratory at Tucson, Ariz. Research here
resulted in The Physiology of Stomata {Cam. hist. Publ.
no. S3). He then entered into a contract with the Con-
tinental-Mexican Rubber Company of New York to study
the biology of guayule {Parlhenium argentatum) in the
state of Zacatecas, Mexico, and in igii the book on Gua-
yule, a Rubber Plant of the Chihuahuan Desert ap-
peared {Carnegie Institution Publication no. ijg; reissued
in IP42). After four years as Professor of Botany in the
Alabama Polytechnic Institute, where he studied boll-shedding
in cotton (Environmental Changes and their Effect on
Boll-Shedding in Cotton, Gossj^iium herbaceum, Ann.
N. Y. Acad. Sci. zg: 1-131, ig2o) he was appointed Mac-
donald Professor of Botany in McGill University, and
Emeritus in igjs. Was chairman of Sect. G., A.A.A.S., in
ig2j; President of the American Society of Plant Physiolo-
gists in ig2y; of the Royal Society of Canada in igjj atid of
the Botanical Section of the British Association in igj4.
He is a Barnes Life Member of the American Society of
Plant Physiologists, Honorary British Fellow of the Botani-
cal Society of Edinburgh, and has received the D.Sc. honoris
causa, /row the University of Wales and from Masaryk Uni-
versity. — In igoj he 7narried Mary Elizabeth Hart,
formerly Professor of Biology in The Western College for
Women, Oxford, Ohio.

The

CARNIVOROUS

PLANTS

BY

Francis Ernest Lloyd

D. Sc. L c. ( Wales); F. R. S. C, F. L. S.
Emeritus Professor of Botany, McGill University

h-^:^

1942

WALTHAM. MASS., U.S.A.

PuLlisLeJ ty tlie Clironica Botanica Company

First published MCMXLII

By the Chronica Botanica Company

of Waltham, Mass., U. S. A.

All rights reserved

New York, N. Y.: G. E. Stechert and Co.,
31 East loth Street.

San Francisco, Cal.: J. W. Stacey, Inc.,
236-238 Flood Building.

Toronto 2: Wm. Dawson Subscription Service, Ltd.,
70 King Street, East.

Mexico, D. F.: Livrarla Cervantes,
Calle de 57 No. i, Despacho 3; Ap. 2302.

Rio de Janeiro: Livraria Kosmos,
Caixa Postal 3481.

Buenos Aires: Acme Agency,
Bartolome Mitre 552.

Santiago de Chile: Livraria Zamorano y Caperan,

Casilla 362..

London, W. 1: Wm. Dawson and Sons, Ltd.,
43 Weymouth Street.

Moscow: Mezhdunarodnaja Kniga,
Kouznetski Most 18.

Calcutta: Macmillan and Co., Ltd.,
294 Bow Bazar Street.

Johannesburg: Juta and Co., Ltd.,
43 Pritchard Street.

Sydney: Angus and Robertson, Ltd.,
89 Castlereagh Street.

Made and printed in the U.S.A.

V\ 1

PREFACE

The experience which has led to the writing of this book began in ig2g
when, examining a species related to Utricularia gibba, / made an observation
of some importance in understanding the mechanism of the trap. This begot
a desire to study as many other species of the genus as I could obtain for com-
parison, primarily to determine the validity of my conclusions. My feeling
that research in this field was promising was strengthened by the discovery that
the pertinent literature was singularly barren of the information most needed,
that is to say, precise accounts of the structure of the entrance mechanisms of
the traps. And an examination of much herbarium material, because of the
meagreness of the underground parts of the terrestrial types resulting from
indifferent methods of collection, forced the conclusion that, even had other
difficulties inherent in studying dried material not intervened, it would be
necessary to obtain adequately preserved specimens. This meant a wide corre-
spondence and, if possible, extensive travel. The uncertainty of achieving the
latter made the former imperative.

The responses to my requests for help were numerous and generous from
all parts of the world, with the result that there came to me from many sources
well preserved material which fairly represented the genus, for it brought to me
some 100 of the total of 2jo or more species. The most lavish single contribu-
tion was put at my disposal by my teacher and friend, Karl von Goebel,
who gave me a collection of Utricularia collected by him in the tropics of the Old
and New Worlds, and in temperate Australia. Many others, while they may
have contributed less in amount, could have been no less generous, for the work
of collecting, preserving, packing and posting specimens is by no means an
easy job.

Travels included two journeys, one to Africa and one to Africa and Aus-
tralia, the latter made possible by a parting gift from my colleagues of McGill
University on my retirement from the Macdonald Chair of Botany in igjj.
At the university centres visited I was afforded all kinds of help: laboratory
space, guidance to promising localities and means of transportation. Several
summers were spent also at the Botanical Institute of the University of Munich
on the original invitation of Professor Goebel, seconded, after his death, by
Professor F. von Wettstein and his successor Dr. F. C. von Faber.

During my preoccupation with Utricularia / had to prepare two presi-
dential addresses, and I was thus led, as has many another in like circum-
stances, to give an account of the whole field of plant carnivory. My interests
were widened in this way, and soon I became imbued with the idea of bringing
together, and perhaps of adding to, our knowledge of this fascinating group of
plants. This extended my list of desiderata. On my requests sent to various
correspondents I received material of every group, some living, some preserved,
e.g., living material of Heliamphora nutans from the Edinburgh Botanical
Garden, where also I saw and studied Cephalotus.

On the study of the material received from many sources, therefore, the ac-
counts in this book rest, and not, in the first instance, on the published papers

Francis E. Lloyd — viii — Carnivorous Plants

of the many excellent workers who have busied themselves in this field, excepting,
however, the studies oj fungi, of digestion and, in some forms, of motility.

In view of so much help I cannot forbear from making some, if inadequate,
acknowledgement: —

First of all I should acknowledge the hospitality of the botanical stafifs of the Uni-
versities visited and made use of as centers of activity during my travels. At the Edin-
burgh Botanical Garden, Sir William Wright Smith, Mr. M. Y. Orr and Dr. J. M.
Cowan (who helped me by raising seedhngs of Utricularia); at the Royal College of
Surgeons, my friend Dr. J. Beattie; at the University of Capetown, Miss E. L. Stephens
who has been a constant help for some years. My stay there was made profitable by
the practical assistance in transportation and guidance afforded by Mrs. Frank Bolus,
Professor R. H. Compton, Mrs. M. R. Levyns and Mr. A. J. M. Middlemost; at Bris-
bane University, Professor D. A. Herbert and Mr. C. T. White; at the University of
New South Wales, Sydney, Professor John McLuckie, Dr. Pat Brough, Professor I. V.
Newman (now of the University at Wellington, N. Z.) and other members of the staff;
at the University in Melbourne, the late Professor A. J. Ewart and Miss Ethel I. Mc-
Lennan; at the Melbourne Herbarium, the late Director Mr. F. J. Rae and Mr. P. F.
Morris; at the University of Adelaide, Dr. A. E. V. Richardson (of the Australian
National Research Council) and Professor J. G. Wood; and at Perth, Professor J. C.
Armstrong and Miss Alison Baird; at the Western Australian State Herbarium, Dr.
C. A. Gardner. Without their knowledge of local conditions and immediate assistance,
always put promptly at my disposal, my work would have been much delayed and always
less fruitful.

To those who as individuals have given me various forms of help, often involving
much effort, I offer these mere thanks into which I ask^them to read my highest appre-
ciation. Dr. J. W. Adams, Morris Arboretum; Dr. A. Akerman, Svalov; Miss E. Bea-
trice Ashcroft, Auckland University College; Dr. Joji Ashida, Ky6t6; Professor L. G.
M. Baas Becking, Leiden and Buitenzorg; the late Professor Edward Barnes, Madras
Christian College, India; Mr. Charles Barrett, Editor, The Victorian Naturalist,
Melbourne; the late Mr. H. Blatter, Panchgani, India; Professor Y. Bh.\radwaja,
Benares Hindu University; Dr. K. Biswas, Botanic Gardens, Sibpur near Calcutta; the
late Dr. H. R. Briton-Jones, Trinidad; Mr. J. H. Buzacott, Maringa, N. Queensland;
Mr. E. J. H. Corner, Botanic Gardens, Singapore; Miss Lucy M. Cranwell, The
Museum, Auckland, N. Z.; Dr. J. M. Curry, Health Department, Panama (Canal Zone);
Mr. F. C. Deighton, Department of Agriculture, Sierra Leone; Professor H. H. Dixon,
Trinity College, Dublin; Mr. Wm. Dunstan, Manager, The Herald, Melbourne; Dr. J. H.
Ehlers, University of Michigan, Ann Arbor, Mich.; Miss Katherine Esau, University
of CaUfornia, Davis, Calif.; Professor M. L. Fernald, The Gray Herbarium, Harvard
University; Mr. M. Free, Brooklyn Botanic Garden; Mr. A. V. Giblin, Hobart, Tas-
mania; the late Prof. H. Gluck, Heidelberg; Professor T. H. Goodspeed, University of
California; Professor John E. Holloway, The University, Dunedin, N. Z.; Mr. R. E.
Holttum, Botanic Gardens, Singapore; Dr. F. C. Hoehne, Sao Paulo, Brazil; Dr. M.
Homes, The University, Brussels; Mr. F. W. Jane, University College, London; Dr.
W. Karstens, Leiden; Dr. S. B. Kausik, Central College, Bangalore, India; Professor
L. P. Khanna, Rangoon, Burma; Professor W. Kupper, Botanical Institute, Munich;
Mrs. M. H. Lea, Fairhope, Ala.; Frere Leon, Cuba; Dr. Gunnar Lohamm.\r, Uppsala;
Mr. Allan McIntyre, Hobart, Tasmania; Mr. C. Macnamara, Arn Prior, Ont.; Dr.
E. B. Martyn, now of Jamaica; Mr. O. Mellingen, Hanau, Germany; Dr. E. M.
Merl, Munich; the late Professor G. E. Nichols, Yale University, New Haven, Conn.;
Mr. C. E. Parkinson, Forest Research Institute, Dehra Dun, India; Dr. D. Y. Padma-
peruma. Royal College, Colombo, Ceylon; the late Mrs. Emily H. Pelloe, Perth, W. A.; Dr.
A. Quint ANiLHA, The University, Coimbra, Portugal, later of Paris; Mrs. Lester Rown-
TREE, Carmel, Calif.; Mr. E. O. G. Scott, Launceston, Tasmania; Professor em. Geo.
H. Shull, Princeton Unversity, Princeton, N. J.; Mr. N. D. Simpson, Botanical Garden,
Peradeniya, Ceylon; Dr. C. M. Smith, De Land, Fla.; Mr. J. H. Smith, Atherton, Queens-
land; Mr. H. Steedman, Perth, W. Australia; Mr. E. J. Steer, Capetown, S. Africa;
Mr. D. R. Stewart, Albany, W. Australia; Dr. G. H. H. Tate, American Museum of
Natural History, New York, N. Y.; Professor R. B. Thomson, The University, Toronto,
Ont.; Dr. J. C. Th. Uphof, Orlando, Fla.; Dr. C. A. Weatherby, Gray Herbarium,
and Professor Wm. H. Weston, both of Harvard University; Dr. Fr. v. Wettstein,

Carnivorous Plants — - ix — Preface

K. Wilhelm Institut f. Biologic, Dahlem; Professor Edgar J. Wherry, University of
Pennsylvania, Philadelphia, Pa.; Mr. J. Wyer, N. Queensland Natural History Club,
Cairns, N. Queensland.

Finally my thanks are due to the Carnegie Institution of Washington, at Stanford
University, for technical help.

From time to time during the last 50 or 60 years there have appeared in various
popular magazines and newspapers accounts giving more or less detailed descriptions of
fabulous man-eating trees. The earliest of these, apparently, is one which was written
by Dr. Carle Liche, quoted at length by Chase S. Osborn in his book Madagascar,
the land of the mafi-eating tree. This lurid title was used avowedly to "enmesh the interest
of possible readers", not to propagate the faith. A summary of this and of a number of
other yarns has been provided by Sophia Prior in a bulletin issued by the Field Museum
of Natural History in 1939. If the reader cares to inform himself concerning this lore,
these two sources will set him on his way. Miss Prior's paper is documented, and in-
cludes reproduction of some of the illustrations which constitute part of the original but
unconvincing evidence offered in the various accounts reviewed by her. Extensive use
has been made of the Field Museum bulletin by Dr. Abilio Fernandes in an article
entitled Morphologia e biologia das plantas carnivoras (see under Drosophyllum).

An amusing, perhaps also tragic, circumstance is to be found described in Liche's
account, in which a highly imaginative illustration shows that, instead of a native maiden
being sacrificed by her tribe by yielding her up to the man-eating tree (possibly a ficti-
tious kind of cycad), a beautiful magazine cover blonde was the Iamb brought to the
slaughter . . .

A certain carnivorous-plant-mindedness shown by the general public has been due
also to occasional cartoons in papers and magazines. In these it is usually the flowers
which are incorrectly if amusingly represented as the traps. Such contributions to the
more evanescent literature are happily intended less for instruction than for titillation.
The misconceptions which arise in this way, while doing little harm, awaken curiosity,
the mother of knowledge.

All the illustrations in this book are originals, prepared by the author,
unless specifically noted otherwise. The names of authorities in many cases
are not accompanied by dates. In such cases only a single publication, to be
found in the literature lists, can be referred to. Passages in languages other
than English have been translated.

The arrangement of chapters may appear illogical. The principle under-
lying it is the increasing complexity of the traps. But for this, the fungi may
be thought to appear in a strange setting.

Finally I wish to acknowledge assistance, in the reading of proofs, of Professor C. B,
VAN NiEL, of Stanford University; of Dr. Mary Mitchell Moore (Mrs. A. R. Moore).
and Mrs. F. Verdoorn, who also prepared the indices. Dr. Michael Doudoroff,
University of California, and the editors of Chronica Botanica have kindly helped me in
checking a number of references to the literature.

Caroli Goebelii

Praeceptoris Illustrissimi
Amici Fidelis
MEMORIAE

CONTENTS

INTRODUCTION

Number of species, genera i

Geographic distribution i

Kinds of traps tabulated 2

Analogs 3

Origin, evolution 7

Literature °

Chapter I: HELIAMPHORA

Heliamphora mdans 9

Discovery 9

Habitat 9

Appearance 9

The pitcher 10

Abnormal leaves 11

Other species ^ ^

Habitat "

The pitcher 12

Drainage slit 12

Histology ^3

Trichomes ^3

Glands ^4

Prey and its fate ^^

Literature ^°

Chapter II: SARRACENIA

Original description i7

Known species . . . ._ i?

Geographical distribution 17

Sarracetiia purpurea 18

History i^

Pitcher 1°

Early ideas 18

Distribution

Pitcher-leaf

Form

Structure

Outer surface

Inner surface 20

Leaves, juvenile 22

Histology

Glands

Trichomes

Sarracetiia psittacina

Form

19
19
19
19
20

22
22
22
23
23

Habitat 23

Pitcher 24

Development 24

Interior surface 25

Sarracetiia Conrtii 26

Sarracetiia ttiinor . 26

Sarracetiia Druttittiotdii . _ 27

Precarious footing for flies .... 29

Sarracetiia flava 29

Sarracetiia Jonesii 3°

Leaf 30

Morphology 3°

Digestion 32

Absorption 32

Animal inhabitants 35

Literature 3^

Chapter III: DARLINGTONIA CALIFORNICA

Distribution 4°

Habitat 4©

Habit 40

Leaves 4°

Juvenile 4°

Adult 42

Form 42

Color 43

Fenestrations 43

Entrance 43

Structure 43

Wing 43

Fishtail 43

Trichomes 44

Glands 44

Absorption 45

Locus of 45

Pitcher leaf 46

Development 4"

Digestion 4°

Absorption 48

Pitcher fluid 48

Wetting power 48

EnzjTnes 49

Bacteria 49

Literature 49

56852

Francis E. Lloyd

Xll

Carnivorous Plants

Chapter IV:

Geographical distribution 51

Habitat Si

General characters 51

Seedling S3

Morphology S3

Form of leaves Si

Adventitious shoots S4

Leaves S4

Pitcher 55

Form SS

Color S5

Mouth and lid 56

Spur S7

Peristome 57

Anatomy 57

Form, variety of 57

Morphology 59

Histology 63

Peristome 63

Glands 64

Wall 65

Anatomy 65

Vascular system 66

Interior surface 66

Waxy zone 66

Digestive zone 68

NEPENTHES

Digestion 69

Pitcher fluid 69

Hooker 69

Tate 70

Rees and Will 70

Gorup-Besanez 70

Vines 70

Dubois 71

TiSCHUTKIN 71

Goebel 71

couvreur 71

Wallace 72

Grimm 72

MoHNiKE 72

Clautriau 73

Fenner 73

Hepburn et al 74

Stern and Stern 74

De Kramer 76

Pitcher liquor 76

Antisepsis 76

Animal life of 77

Folk lore 78

Uses 78

Literature 79

Chapter V: CEPHALOTUS FOLLICULARIS

Distribution
Habit . .
Habitat .
Leaf . . .

Foliage .

Pitcher .

81

81
81
81
81
82

Development 82

Morphology 82

Anatomy 85

External surface 85

Internal surface 85

Glands 86

Glandular patches 87

Digestion 88

Literature 89

Chapter VI: GENLISEA

Discovery 90

Distribution 90

Early studies 90

Warming 90

Goebel 90

Flower 90

Leaves 90

Trap 91

Prop cells 91

Size 92

Form 93

Anatomy 93

Histology 93

Glands 94

Fohage 90 Literature

94

Chapter VII: BYBLIS

Species 95

Occurrence 95

Appearance 95

Systematic position 95

Habitat 95

Habit 95

Leaf 96

Form 96

Structure 96

Glands 96

Structure 96

Functions 97

Digestion 97

Absorption 97

Literature 98

Carnivorous Plants

xm

Contents

Chapter VIII: DROSOPHYLLUM LUSITANICUM

Occurrence . .
Habitat . . .
Appearance . .
Habit . . . ,
Leaf

Form . . .

CircinatioD .

Marcescence

99
99
99
99
99
99
99
99

Glands loo

Structure loo

Secretion loi

Digestion 102

Cytology 103

Recent work 103

QUINTANILHA I04

Literature 105

Chapter IX: PINGUICULA

Geographical distribution 106

Appearance 106

Habitat 106

Leaves 107

Glands 107

Structure 107

Darwin's studies 107

Movements 108

Secretion; Digestion 109

Uses of leaves 112

Literature 114

Chapter X: DROSERA

Species iiS

Number 115

Distribution 115

Habitat 115

Form, habit 116

Leaf-roots 117

Leaf

Unfolding movements 117

Form 117

Anatomy 118

Starch content 119

Tentacles 120

Glands of 121

Development of 123

Function of parts 124

Sessile glands 125

Function of 126

Absorption, Locus of 127

Reproduction 129

Seed 129

Tubers 130

Gemmae 131

Leaf buds 131

Condition for incidence 133

Polarity 134

Carnivory 13S

Early observation 13S

Mucilage 136

Locus of secretion 137

Movements :-

Tentacles 138

Leaf-blade 138

Direction of bending 139

Duration of response 139

Leaf -blade and stimulus 140

Stimulus

Path of 140

Intensity of (Darwin) 140

Nature of stimulant 141

Tentacles 142

Mechanism of movement 142

Specificity of reaction 143

Aggregation: Darwin, C 145

i5arwin, F., Gardiner 146

De Vries 147

Akerman 148

coelingh 149

Janson 151

Homes 152

Cytoplasm and nucleus 156

Digestion, enzymes 158

Carnivory, significance of 162

Literature 165

Chapter XI: CARNIVOROUS FUNGI

Zoophagy (Cordyceps) 169

Earliest discovery by Zopf 169

Loop snares :-

Swelling 170

Adhesive 171

Eel-bob snare :-

Zoophagus, Sommerstorffia 171

Gicklhorn's studies 172

Capture of rhizopods by:-

Adhesive alone 173

Adhesive organs 173

Literature 175

Francis E. Lloyd

XIV

Carnivorous Plants

Chapter Xlla: DIONAEA

General description i77

Discovery by DoBBS ■ ■ • 178

Original description by Ellis .... 179

Diderot 179

Later work 180

Curtis, Oudemans, Canby .... 180

C. Darwin, Goebel 180

Seed, seedling 181

Leaf 182

Trap 182

Cma 182

Lobes, posture of 183

Trichomes (steUate) 183

Glands (digestive, alluring) .... 183

Closure of trap 184

Sensitive hairs 184

Internal structure of trap 185

Physiology 186

Stimulus 186

Perception, localisation of 187

Mechanism of closure 188

ZlEGENSPECK, C. DaRWIN . . . l88

MuNK, Batalin 189

Brown, Macfarlane 190

Von Guttenberg 191

Ashida 192

Haberlandt 193

Digestion; absorption 194

Literature 210

Chapter Xllb: ALDROVANDA

General description 194

Discovery 195

Distribution 195

Leaf 19s

Seed, seedling 19S

Mature leaf 196

Bristles 196

Petiole 196

Trap, posture of 197

Development 197

Terminology (Ashida) 197

Structure 197

Trichomes, glands 199

Sensitive hairs 200

Mechanics of trap movement . . . 201

Locus of bending 201

Dionaea, comparison with .... 202

Recapitulation 204

Stimuli, responses to 205

Electrical 205

Temperature effects on 206

Chemical 206

Sugar, glycerin 206

Neutral salts, acids, alkalis . . . 207

Other organic substances .... 208

Formalin, ether, ethyl alcohol . . 208

Chloroform 208

Ethyl alcohol, chloroform .... 208

Digestion; absorption 208

Vesicles, interpretations of 209

Culture 209

Literature 210

Chapter XIII: UTRICULARIA, BIOVULARIA and
POLYPOMPHOLYX

Form of plant 213

Traps, variety of 214

Prey 214

Flowers 214

Distribution 215

Embryology 216

Seed 216

Embryo 217

Germination 217

Ulriciilaria capensis, etc 217

Uiricularia vulgaris, etc 218

Utricidaria purpurea 219

Types of Utricidaria 219

Utricidaria vulgaris type 219

Freely floating 220

Floats 221

Rhizoids 221

Foliar dimorphism 222

Dwarf shoots 222

Branching; inflorescence .... 222

Anchored forms 224

Terrestrial forms 226

Epiphj'tic forms 226

Biovularia type. 227

Utricidaria purpurea type 228

Utricidaria dichotoma type 229

Freely floating 229

Anchored 230

Other types, according to trap struc-
ture :-

Utricidaria cornuta 231

Utricidaria caeridea 231

Utricidaria rosea, Warhurgii .... 231

Utricidaria orbiculata, etc 232

Utricidaria simplex, etc 232

Utricidaria globulariaefolia .... 232

Utricidaria Kirkii 232

Utricidaria nana 232

Utricidaria Lloydii 232

Literature 267

Carnivorous Plants

XV —

Contents

Chapter XIV: The UTRICULARIA TRAP

General description 233

Terminology 233

Early ideas 234

Benjamin to Cohn, Darwin . . . 234

Later advances 237

Brocher, Ekambaram, Withy-
combe 236

Merl, Czaja 238

Hegner 240

Watertightness of trap 242

The two valves 242

Anatomjf and

Histology 242

Walls 243

Glands 244

Path of fluid through walls .... 244

Cytology 245

Entrance 245

Development 246

Threshold 246

Pavement 247

Velum, origin of 247

Door 248

Histology 249

Contact with threshold 250

Relation of velum 251

Sucking in of prey 251

Velum 253

GoEBEL 254

Roger Fry 254

Mechanical types of trap as to posture

of door and depth of entrance . . 254

Utricularia vulgaris 254

Utriciilaria capensis 255

Utricularia monanthos 256

Polypompholyx 257

Varieties of traps :-

With short tubular entrance ... 257

With long tubular entrance .... 258

Digestion 263

Prey and its fate 265

Appendix: models of the trap .... 266

Literature 267

The earliest known illustration or Nepenthes (Nepenthes mirabUis (Lour.) Merr.) prom RmiPHHJS
Herbarium Amboinense 5: 59 (published in 1747, but drawn in the second part o^ ™f J7TH
century). The plant at the right is Flagellaria indica. - The vignette on p. xv ha^ been re
PRODUCED prom Clusius' Rariorum Plantakum Historia(c/. p. ^^Y^^Z^iZ^^ll^^Z
OE A Sarracenia. The Drosera vignette on p. 271 has, by courtesy of Prof. Baas Decking, been re

PRODUCED from A PRINT, MADE DIRECTLY FROM A i6TH CENTURY BLOCK USED FOR DODONAEUS HERBALS.

INTRODUCTION

The purpose of this book is to give an historical review and sum-
mary of our present knowledge about the carnivorous or insectivorous
plants, the former being the better term. Of these there are about 450
or more species, representing 15 genera, belonging, aside from the
fungi, to six families, indicated in the present table (Table i), to-
gether with their geographic distribution.

Table i

Family and genus

No. OF

SPECIES

Geographic distribution

S
9

6S

90

Sarraceniaceae
Heliamphora
Sarracenia

Darlingtonia

{Chrysamphora)
Nepenthaceae

Nepentfies
Droseraceae

Dionaea

Aldrovanda

Drosophyllum

Drosera
ByUidaceae

Byhlis
Cephalotaceae

Cephalolus
Lentibulariaceae

Pingidcula

Utricularia

Biovularia

Polypompholyx

Genlisea
Fungi (various genera
with trapping mech-
anisms)

Roridula, formerly regarded as carnivorous, has now been shown
by me not to be so, and is excluded from the above list. The ''man-
eating tree of Madagascar" must at present also be excluded, since the
evidence of its existence is elusive.

The table shows that the carnivorous plants are divisible into two
groups, one lot {Sarraceniaceae to Cephalotaceae) belonging to the Chori-
petalae, the rest to the Sympetalae, with personate flowers. This wide
separation is a remarkable indication that the carnivorous habit has
arisen among the higher plants at two points at the fewest, (as well as
among the fungi), in the course of evolution. The methods of captur-
ing prey are in some measure common to the two lots, the greatest
height of specialization having been reached by Dionaea and Aldro-

30
275

2 (4)
10

20 or more.

British Guiana; Venezuela.

Eastern N. America: Labrador to S. E. United

States of America.
N. California and S. Oregon.

Eastern Tropics to Ceylon and Madagascar.

North Carolina and northern South Carolina,

U. S. A.
Europe, India, Japan, Africa and Queensland,

Australia.
S. Portugal, S. VV. Spain, Morocco.
Ubiquitous.

Australia, from N. W. to S. W.

Australia, extreme S. W.

N. hemisphere in Old and New Worlds.

Ubiquitous.

Cuba; eastern S. America.

S. and S. W. Australia.

W. African and E. South American tropics.

Ubiquitous.

Francis E. Lloyd — 2 — Carnivorous Plants

vanda among the Choripetalae and by Utricularia among the Sympetalae.
For this reason the arrangement (Table 2) which has been followed is
that which groups the plants according to the character of their trap-
ping mechanisms, named for their obvious analogs among human
devices. By 'active traps' is meant those which display special move-
ments necessary or contributory to the capture of prey.

Table 2

Kind (

3F TRAP

Genus

Pitfalls (passive traps), the

pitcher plants

Heliampkora
Sarracenia

Darlingtonia

Cephalotus

Nepenthes

Lobster pot (passive

trap)

Genlisea

Snares (noose, some

active;

sticky discs, etc., passive)

Certain Fungi

Bird lime or fly-paper traps
Passive

Byblis
Drosophyllum

Active

Pinguicula
Drosera

Steel-trap (active)

Dionaea
Aldrovanda

Mousetrap

Utricularia

Biovularia

Polypompholyx

The above table mentions merely the form of the trap. There are,
however, other characters which contribute in some way to the effi-
ciency of action. These include methods of attracting the prey by
means of lures: the odor of violets in Sarracenia, of honey in Droso-
phyllum, of fungus in Pinguicula; the secretion of nectar by glands
either on the traps or on parts leading to them as in Nepenthes, etc. ; the
exhibition of attractive colors and of bright fenestrations in Sarracenia,
Darlingtonia, Cephalotus; of brilliant points of light reflected from
drops of mucilage in Pinguicula, Drosera, etc. ; the secretion of mucilage
in Drosera, etc., movements of various degrees of rapidity, as in Pin-
guicula, Dactylella, Drosera, Dionaea and Utricularia. There are also,
with few exceptions, means for digesting the prey when caught: en-
zymes and acids are excreted. When these, together with the captured
prey, are accumulated in some sort of a receptacle, something much
like the animal stomach results. Involved in all this there are special
structures: hairs, glands, specialized stomata {Cephalotus, Nepenthes),
waxy excretions (Nepenthes), emergencies (tentacles of Drosera).

From the purely physiological point of view the carnivorous plants
are concerned in a somewhat special way in the acquisition of nutrient
substances containing protein, possibly vitamins and perhaps the salts
of potassium and phosphorus, and even others. In this way they re-
ceive some profit, though what they receive is no sine qua non, as it is
with many other plants. As Pfeffer pointed out, many fungi are
wholly carnivorous, as in the cases of Cordyceps, Empusa, etc. Among
the higher plants are some which get all their food materials indirectly

Intro- —3— duction

through agencies such as mycorrhizal fungi. Our so-called carnivorous
plants are therefore not peculiar in this habit.

What then distinguishes the carnivorous plants from the rest of the
plant world? Why should we still share the feelings of the naturalists
of the 1 8th century who regarded them as miracula naturae? We do so,
I think, because a carnivorous plant in the sense here meant is one
possessing a trap which, though merely a constellation of structures and
functions, many of them conmion enough elsewhere among plants, is a
special organ for the capture and digestion of animal prey, thus turning
the tables on animals, which directly or indirectly are herbivorous.

But may these traps as such be regarded as something unique? The
answer to this question must be sought in such analogs as we may find
among plants in general.

Pitfalls in the form of pitchers are of rather widespread occurrence.
In some flowers the corolla is tubular and the inner surface is supplied
with downward pointing hairs, and there is emitted a luring, if not al-
luring, smell. Flies are attracted and caught, but after effecting polli-
nation, and the hairs having withered, are released (Aristolochia). In
other flowers one or more members of the perianth are tubular and se-
crete and hold nectar {Aquilegia, Marcgravia, Delphinium). Perhaps
the closest parallel is found in the pitcher leaves of some species of
Dischidia, a tropical genus of the old world. They are invaded by
adventitious roots from nearby stems of their own plant, and are often
occupied by ants who use them as shelters. Probably in their native
habitats they often contain moisture available for their invading roots.
An inturned marginal rim surrounding the narrow mouth reminds one
of the rim of the carnivorous pitcher, but it seems to have no well
marked special function. In some species of Dischidia the pitchers are
represented merely by dished leaves facing each other. At the other
extreme Dischidia pectinoides has a double pitcher, one inside the other,
according to Goebel. Lathraea squamaria, a root parasite of Europe,
has hollow leaves, the hollow lined with glands. Goebel regards them
as reservoirs for reserve stuffs. The upper leaf lobes of Azolla are also
hollow, but these are inhabited by Anabaena azollae in symbiotic re-
lationship. Among the liverworts are species in which the leaves are
partly converted into "water sacs" (Goebel), notably Frullania
cornigera of New Zealand, though our own species offer sufficiently
good examples. Lejeunea behaves similarly, but the sacs are simpler.
Most impressive are Colura and Physiotum. In P. majus^ occur nearly
closed sacs the mouths of which are guarded by two lips closed to-
gether like the lips of a mussel shell (Goebel). Moreover one of the lips
is moveable, being provided with a hinge region, thus serving as a valve.
Precisely how this valve works is not clear. Goebel, to whose account
I am indebted, points out that such an arrangement is known only in
Utricularia, but it must be remembered that this comparison loses some of
its cogency for the reason that Goebel thought the valve of the Utric-
ularia trap to be a simple check valve. There is no evidence that these
arrangements in the liverworts indicate a carnivorous habit, though
they are inhabited, like any liverwort or moss, by protozoa, nematodes,
etc. That they are water holders is evident.

The common teasel {Dipsacus sylvestris) has been regarded in all

Francis E. Lloyd — 4 — Carnivorous Plants

probability to be a carnivorous plant by Miller Christy (1923).
This biennial herb is well known for its water catching reservoirs formed
by the connation of the opposed leaves at their bases. A large plant
attains a height of 6 feet. Eight plants, with an average height of 5 feet
8 inches, were found by Christy to retain an average of a half pint of
water. It is of interest to know that the teasel for this reason claimed
the attention of Turner (1551), who remarked the catching of "rayne"
and ''dew" (Herball, o.iiij, 1551) and Gerard (Herball, p. 1005, 1597)
wrote quaintly, as it now appears to us, "The leaves growe foorth of the
iointes by couples, not onely opposite or set one against an other, but
also compassing the stalke about, and fastened togither, and so fas-
tened that they hold deaw and raine water in manner of a little bason."

Christy rejects the ideas that the primary object of the collection
of water is the succour of the plant in times of drought, and the pro-
tection of its nectar from predatory insects. The presence of dead in-
sects, rendering the water filthy, seems to point to these as a source of
nutriment. "The cups undoubtedly form most efficient traps," Fran-
cis Darwin had said. Christy suspected the water to have some
narcotizing or intoxicating substance (F. Darwin had noticed that
beetles drown in it more rapidly than in pure water), and he further
expressed the conviction that "the plant does profit by the insects
caught in the cups". In view of the general evidence Christy draws
the conclusion that the teasel is a carnivorous plant, but without ad-
vancing any definite experimental proof.

The lobster-pot of Genlisea, though an exceedingly specialized
structure, is fundamentally nothing more than a narrow pitcher with its
interior armed with downward pointing hairs. Even the curious
method of holding the lips of the narrow slit-like mouth in rigid rela-
tion to each other by an adhesion of cells finds its parallel in other
situations such as the adhesions of algal cells and those of mycelia.
In form, the 'prop-cells' responsible find a loose analog in the cystidia
of Coprinus. But, after all, their structure and method of function is
unique.

The snares found among the carnivorous fungi — those having def-
inite traps — are more obscure in their analogies, and, it would appear,
have originated within the group. Apparently unique is the noose of
Arthrobotrys, etc. The adhesive disc is found among the orchids, in
which it is the mechanism for attaching the pollinia to visiting insects.
Obviously the orchids did not invent this originally — the fungi prob-
ably did so. The loop of the pollinia of Asclepias is a sort of noose
snare (Carry).

The snare of Zoophagus is a variant of the adhesive disc, but is re-
markable as a device resembling in its manner of working a common
fish-line and hook, or perhaps better an 'eel-bob.'

The plants which catch their prey by means of a viscid secretion
are only a few of a multitude of others that excrete sticky substances
by which small insects are caught. These substances are in general of
three kinds: oily (often aromatic), resinous and mucilaginous. Among
the carnivorous plants, only the last is found, as a watery medium is
the only one that can carry an enzyme, as in Drosera. Adhesive (mu-
cilaginous or resinous) glands are very common, and often small insects

Intro- —5— tiuction

are captured, as, e.g. in the case of the catch-fly {Silene). Suspecting
that many such plants might turn out to be carnivorous Darwin
investigated the behavior of some of them: Saxifraga umbrosa, S.
rotundijolia (?), Primula sinensis, Pelargonium zonale, Erica tetralix,
Mirabilis longifolia and Nicotiana tabacum. But while he thought to
have proved that the hairs of these plants can in some instances absorb
organic nutrients, he regretted that he did not try if they could "digest
or render soluble animal substances." Fermi and Buscaglione in 1899
tried some of these and still others {Martynia, Hydrolea, Sparmannia)
for digestion with negative results, whereas those of the recognized
carnivorous plants which they tried were positive. This brings into
relief the fact that there are many plants which resemble our carniv-
orous plants so closely that we can decide about them only through
experiment.

Though the glands involved are in structure similar in some cases
{Byhlis, Pinguicula) to those found among other plants, those of Dro-
sophyllum and Drosera, fundamentally the same in structure in both,
are unique as entireties. Those of Drosera are raised on emergencies
which display motility in no respect different, except perhaps in speed,
from that of ordinary growth. The histological elements of the glands
are common enough; again it is the constellation of characters which

stands out.

The most complete analog to a carnivorous plant of this type is one
which was until recently regarded as one itself. This is the Roridula,
of which there are two species, in South Africa. I myself included it
among the carnivores in an account pubhshed in 1933. Since that
time, on receiving material preserved in formalin from Munich, it was
at once apparent that the secretion which appears as glistening drop-
lets in the living plant, was intact and still adherent to the glands, and
could therefore not be a mucilage. Had the preservative been alcohol
this might have escaped attention. The leaves bear many tentacles
superficially similar to those of Drosera. Examination showed them to
be anatomically quite different, and that they exude a resinous secre-
tion. There are no other glands, so that on this evidence the carniv-
orous habit seems to be quite excluded (Lloyd, 1934)- These two
species, relatives of Drosera, are, like them and Byblis, used by certain
insects (certain bugs and crab spiders) as habitual feeding grounds.
When insects are freshly caught, they are attacked for their body juices.
How these commensal forms avoid capture is another matter, but an
interesting one.

The trap of Dionaea and Aldrovanda, with its close resemblance to
a steel-trap, has been, and still is regarded by some as "perhaps the
most marvellous in the world," to quote Morren (1875) who, in say-
ing this, was only repeating what Darwin had already said. It appears
to be quite unique when regarded as a total mechanism. But an analog,
in some measure at any rate, was suggested by Delpino, quoted by
Hooker in his Presidential Address at the meeting of the British
Association for the Advancement of Science in 1874. Hooker had
already described a plant from Tierra del Fuego under the name Caliha
dioneaefolia "which," Delpino in effect remarked, "is so analogous in
the structure of its leaves to Dionaea, that it is difficult to resist the

Francis E. Lloyd — 6 — Carnivorous Plants

conviction that its structure also is adapted for the capture of small
insects".

Contributory structural features of these traps are the glands, phys-
iologically of two kinds, but of identical {Dionaea) or different (Aldro-
vanda) structure, and the sensitive hairs, which are local points of
greater sensitivity. The latter only are unique in structure. There is
also motile tissue which, startling in rapidity though its movements
are, appears to work in much the same way as that of tissues exhibit-
ing geotropic responses, according to Brown. But should Caltha
dioneaefolia, as Delpino suggested, turn out to be quite parallel to
Dionaea, it would only add another member to a very small unique
group.

A curious case of an insect-catching grass, Molinia coerulea, briefly
described by F. Ludwig in 1881, may here be indicated as an analog,
albeit a very loose one, to the trap of Dionaea. It appears that this
grass can catch small insects between its paleae, which act as the jaws
of a spring trap, after the fashion of a 5-cent spring mousetrap. It is
well known that, during flowering, the paleae are forcibly separated
by the swelling lodicules, and are held there for the space of anthesis.
The lodicules then shrink and allow the paleae to close. If now, during
the period, an insect, attracted by the shining, sappy and turgid masses,
attacks them by biting or puncturing, the resulting reduction of turgor
is sufficient to allow the outer palea to close, which it does "with sur-
prising swiftness" (Hildebrand, fide Ludwig) thereby trapping the
offending insect. This action, Ludwig points out, disadvantages the
plant in curtailing the time during which the flower would normally
remain open. No compensating benefit seems to accrue.

The trap of Utricularia, minute though it is, is compared in the
present account to a mousetrap. There are mousetraps and mouse-
traps however, from simple to complex in structure, from a 5 cent
dead-fall to an elaborate, automatic self-setting one, which catches
them as fast as they come. If to this should be added a disposal plant
(Prof. Tracy I. Storer informs me that such a mousetrap has been
invented) so that nothing is left at last but hair and bones, the com-
parison would be fairly complete, especially if the trap should work in
any position and at the same time under water. These are constructed
of rigid parts, while that of Utricularia is composed of soft, yielding
parts. Previous to 191 1, the Utricularia trap was thought to be rela-
tively simple: a soft "pitcher" or vesicle guarded by a simple check-
valve; now it is known to have two valves, a tripping mechanism, a
spring which opens the door (one of the valves) which then automati-
cally closes, barely to indicate the complexity of the mechanism, the
complexity and perfection of which are extraordinary. In 1891 Goebel
said that the Utricularias are "among the most interesting of plant
forms, whether we view them from the point of view of their morphol-
ogy, anatomy or biology." If this was true at that date, it is, because
of the added knowledge about the complexity of the trap, much more
true now, as indeed Goebel personally admitted to me in conversation.

For my own edification I have attempted to indicate the structure
of a mousetrap which, as closely as may be, duplicates the trap of
Utricularia without, however, hoping for the reward for him who invents

Intro- — 7 — duction

a mousetrap, promised, I believe, by Emerson. In this I am yielding
to the importunities of many of my friends, whose urgings I must as-
sume to be disinterested. I have, somewhat apologetically, relegated
the drawings and description (very necessary I fear) to an appendix
to the chapter on Utricularia.

For such a mechanism we cannot find an analog among other
plants. Though it has moving parts, the property of irritability is not
used. Particularly, after the door opens, which it does only passively,
it instantly recovers its original position all in 1/33 second. Though its
movements are made possible by its turgidity, there is no change of
turgor — hence the instant reversibility of movement. But this again
depends on a structure which finds some analogy in the walls of anthers,
but only a partial one. Without further amplification we may regard
Utricularia as unique.

It is not without interest to note that among the Lentihulariaceae
we find examples of the simplest traps (Pinguicula), the most complex
of the pitfall type, (in the lobster pot of Genlisea), and the incomparable
trap of Utricularia, whose only rival is that of Dionaea. Which of the
two is the more "wonderful" (I refer now to Darwin's statement that
he thought Dionaea the " most wonderful plant in the world ") will
perhaps be a matter of opinion, but the evidence seems to favor Utricu-
laria.

How all these traps work and how we came to know about them,
it is the purpose of this book to tell. But I have not confined myself
to the traps, for it seemed necessary to present an adequate picture of
the plants as a whole. This was especially true of Utricularia, as, in
spite of many studies, a survey of the entire genus (and those of Poly-
pompholyx and Biovularia) has not been made since the 1891 publica-
tion of GoEBEL. No genus more fully substantiates the saying of
Caruel "La pianta cresce ciascuna alia sua idiosincrasia", for which
allusion I am indebted to Professor Goebel. The survey presented
seems to indicate with some fairness the extraordinary variety of form
and behavior of these plants, but necessarily as briefly as possible in
the interest of space saving.

About the origin and evolution of the carnivorous plants, however
much these questions may intrigue the mind, little can be said, nor
have I attempted to discuss them. The evidence from fossils is meagre,
for these plants, even the most prolific of them, have seldom been pre-
served. A Utricularia {U. Berendii Keilhack) is recorded from the old-
diluvial of Oberohe (Engler and Prantl). No others, so far as I
know, have been recorded. The water lilies are recorded for the Ter-
tiary, and it is probable that Utricularia was contemporary. The fact
that they have originated at two or more distinct points in the phylo-
genetic tree is of major importance. How the highly specialized organs
of capture could have evolved seems to defy our present knowledge.

J. G. Peirce (1926) remarks that the wide distribution of the car-
nivorous plants and the permanence of their peculiar morphological
and physiological characters mark them as descendants of ancient
forms, but we have to add that only some of them are widely dis-
tributed {Drosera, Pinguicula, Aldrovanda (Old World only) and Utric-
ularia) while others, though related in one taxonomic group or the

Francis E. Lloyd — 8 — Carnivorous Plants

other, are of restricted, sometimes very restricted, distribution. The
two categories exist side by side in that ancient continent, Australia.
Are the latter young scions derived from the more ancient stocks? And
may we regard the Australian types of Utricularia as ancient t3^es and
in some measure as analogs of ancient animal forms of that continent?
Since we cannot answer these questions, it is perhaps as well to say no
more.

Literature Cited:

Carry, T. H., On the structure and development of the g3mostegium and the mode of

fertilization in ^iic/e^/a^ CorM7</j Decaisne. Trans. Linn. Soc. London, II, 2:173-208,

1881-1887.
Christy, M., The common teasel as a carnivorous plant. Journ. Bot. 61:33-45, 1923.
Cramer, C, tJber die insectenfressenden Pflanzen. A lecture. Ziirich, 14 Dec. 1876.

Seen. No more expHcit data given.
Darwin, F., On the protoplasmic filaments from the glandular hairs on the leaves of the

common teasel {Dipsacus sylveslris). Quart. Journ. Micros. Sci. n. s. 25:245-272, 1877.
Drude, O., Die insectenfressenden Pflanzen. Schenk's Handbuch der Botanik, Breslau,

1881, 1:113-146.
Fermi and Buscaglione (see under Utricularia).
Franca, C, La question des plantes carnivores dans le passe et dans le present. Bol. Soc.

Brot. 1:38-57, 1922.
Goebel, K., Organographie der Pflanzen, 3. Auflage, Jena, 1928-33.
Hooker, J. D., Address to the Department of Zoology and Botany, 1874. B. A. A. S.,

Report of the Forty-fourth Meeting, 1874:102-116, 1875.
Jones, F. M., The most wonderful plant in the world, with some unpublished correspon-
dence of Charles Darwin. Nat. Hist. 23:589-596, 1923.
Lloyd F. E., Is Roridula a carnivorous plant? Can. Journ. Res. 10:780-786, 1934.
LuDWiG, F., Molinia coerulea als Fliegenfangerin. Bot. Centralbl. 8:87, 1881.
Morren, Ed., La theorie des plantes carnivores et irritables. Bull, de I'Acad. Roy. Belg.

II, 40:1040 seq. (seconde edition revue et amelioree dans Bull. Fed. Soc. Hort. 1875).
OsBORN, C. S., The land of the man-eating tree. Pp. 442, New York, 1924.
Peirce, J. G., The physiology of plants. New York, 1926.
Pfeffer, W., tJber die fleischifressenden Pflanzen und iiber die Ernahrung durch Aufnahme

der organischen Stoffe iiberhaupt. Landw. Jahrb. 6:969-998, 1877.
Planchon, J.-E., Las plantes carnivores. Rev. des deux mondes, 13:231-259, 1876.

Chapter I
HELIAMPHORA

Discovery. — Appearance of H. nutans. — Discovery of other species. — Habitat. —
Leaf structure. — Leaf forms. — Comparison of species.

The genus Heliampliora is based on a plant, H. nutans, collected by
ScHOMBURGK who found it growing "in a marshy savannah, at an ele-
vation of about 6000 ft. above the level of the sea on the mountain of
Roraima," "the fruitful mother of streams," on the borders of British
Guiana. (Bentham 1840; Schomburgk 1841). Im Thurn (1887) de-
scribed its habitat and appearance. " the most remarkable

plant of the swamp is the South American pitcher plant, Hel-

ianiphora nutans Benth., which grows in wide-spreading, very dense
tufts in the wettest places, but where the grass happens not to be long.
Its red-veined pitcher-leaves, its delicate white flowers raised high on
red tinted stems, its sturdy habit of growth, make it a pretty little pic-
ture wherever it grows it attains its full size and best de-
velopment up on the ledges of the cliff of Roraima and even

on the top." (Im Thurn 1887).

For many years only the one species, H. nutans Bentham {i — i), was
known. The meagre hterature deals almost solely with this plant.
Four other species are now known. Three of these were discovered on
Mt. Duida, Venezuela, by Dr. G. H. H. Tate just previous to 1931,
and were described by Dr. H. A. Gleason in 1931. An examination
of the herbarium material (all t3^e specimens in the herbarium of the
New York Botanical Garden) was made possible for me through the
kindness of Dr. Gleason, and my notes on these species have been
published (1933). A fourth species, also discovered by Dr. Tate in
1937-8 in the same general region, is H. minor (Text fig. i). The
three Mt. Duida species, H. Macdonaldae (i — -2), H. Tyleri and H.
Tatei, are closely related and furnish a striking example of closely
related species arising within a restricted region (Lloyd 1905). Fur-
ther exploration may show the genus Heliaynphora to be as prolific
of species as the North American Sarracenia, or even more so.

Heliamphora grows in a region of vast rainfall, under extremely wet
conditions of soil and air. It is cultivated only with difficulty, but
successfully at the Edinburgh Botanical Garden, and I am indebted to
Sir William Wright Smith for both preserved and living material of
the species H. nutans, on which the present account is based. The
name Heliamphora means swamp-pitcher.

The accounts of H. nutans given us by Zipperer (1885), Mac-
farlane (1889, 1893), GoEBEL (1891), and by one of his students,
Krafft (1896), leave Httle to be added. The plant consists of a rosette
of basal leaves arising from a strong rootstock. It produces a simple
racemose inflorescence of white or pale rose colored apetalous flowers.
The 4-6 sepals are ovate-acuminate in form, with numerous stamens
and a trilocular ovary with a single style. The normal mature leaves

Francis E. Lloyd ^10 — Carnivorous Plants

may attain a length of 30 cm.; in cultivation, they are rarely longer
than 15 cm. On side shoots of the rhizome arise depauperate branches
bearing leaves in various stages of arrested development (2 — 2-4),
which have been duly described by the above authors.

Structure of the leaf. — The normal pitcher leaf is an insect trap of
the pitfall type. Its form is that of a gracefully curved funnel, widen-
ing above the base to contract somewhat just below a leafy expansion,
the bell (2 — i; Text fig. i). The apex of the bell normally ends in a
spoon-shaped, thick- walled structure resembling superficially the lid of
Nepenthes, which stands upright as represented by Bentham, and not
bent forward as Goebel suggested as a protection of the nectar se-
creted against rain. The spoon is lacking in some of the leaves on
plants which I have seen growing in greenhouses, where many of the
leaves are small and lacking in vigor, but seems to be normally present
and upright in position in wild plants, judging from a photograph
made by Tate (Lloyd 1933).

Mrs. Arber (1941) regards the spoon ("hood" she calls it) as the
continuation of the rudimentary curve-over of the pitcher-lip. The
indications of this curve-over are stated to be "very slight". My ma-
terial, derived from the same source as hers, shows no such indication.
To be sure, the sides of the spoon are the incurled margins of the leaf
apex, but these have no relation, it seems to me, to anything corre-
sponding to a putative "outward roll-over".

The flaring bell is oval in form, the margins sweeping forward to
meet at once or to run parallel for a short distance before joining. At
this point they continue into two narrow wings running down the mid-
ventral surface of the pitcher, toward the base of which they enlarge
and spread to form a wide membranous spreading and clasping base.
Above the wings are closely approximated, and in this respect Heliam-
phora, according to Macfarlane, occupies an intermediate position be-
tween Nepenthes, with widely removed wings and Sarracenia and
Darlingtonia with a single keel (representing fused wings). The outer
surface of the pitcher bears numerous twin unicellular trichomes which
have inverted V-shaped spreading arms quite unique and peculiar to
this plant (2 — 6 and Text fig. i), stomata, and many minute glands
which probably secrete nectar (2 — 10). These are numerous on the
wings, and are probably part of a general lure for insects. The inner
surface can be divided for the purpose of description into four zones.
Zone I (2 — i) is the spoon, which is quite smooth on its concave sur-
face, but here bears a number of glands which are larger than else-
where, and some of them very large, usually three or four on each
flank. These are nectar glands. Zone 2 begins just below the spoon
and is indicated by a dense clothing of downward pointing, delicate
hairs in great numbers covering the whole of the surface of the bell
(which occasionally is smooth), and the upper constricted portion of
the tubular part of the pitcher. They are found in reduced numbers
for a short space just below the constriction, and here they are very
long and straight. Below a certain point, however, the hairiness sud-
denly ceases, and gives way to the next zone (3) in which the epidermis
is glistening smooth. While in zone 2 there are very numerous nectar
glands, interspersed between the bases of the hairs, here in zone 3 there

Chapter I — 11 — Heliamphora

are none at all. Below the lower limit of the smooth area there be-
gins zone 4, which again is clothed, but more sparsely than the bell,
with downwardly directed hairs. These are very stout and claw-like
{2 — 5), while those of the bell are longer and flexible. Both kinds
are longitudinally ridged with delicate folds of cuticle, the hairs of the
bell especially so. The difference of stature and rigidity is related to
their functions. The hairs of the bell afford an unstable footing for
insects which are trying to get at the nectar on the surface between
them, and their flexibility adds to the instabihty.. while those in the
depths of the pitcher have the role of retention, and for this purpose
the stronger they are the better. There are no glands in zone 4.
There are no digestive glands present in this trap.

The abnormal leaves on the reduced side shoots are of various forms
{2 — 2-4). The base of the shoot is enveloped in scale leaves. Fol-
lowing these there may be underdeveloped leaves, described by Krafft,
in which the tube is very slender and would have a very limited func-
tion as a trap, and acts simply as a petiole. The bell is relatively large
so that the whole behaves as nothing more than a photosynthetic organ.
A spoon is not developed. The inner surface of the blade (bell) is free
of hairs, but carries nectar glands. At the lower limits of the bell
there is a trapping zone of hairs, but the rest of the tube is quite
smooth. It is to be noticed that the "tube" is not truly such, since its
edges are not fused. This I judge from Krafft's drawing, though he
does not state so specifically. Goebel described a sort of juvenile leaf
at the base of normal shoots consisting of a closed tube, winged with
two wings as in the normal, but with a much reduced bell with out-
turned edges and no spoon. Krafft added some details, pointing out
that the whole of the inner surface, save a small zone at the base, is
lined with downward pointing hairs. The bell, with the exception of
the outturned edges, is also hairy, but the spoon — this is not evi-
dently such, nor did Goebel recognize it — is without hairs or glands.
Goebel described a still simpler and more elementary condition in
leaves seen on much reduced side shoots. These were small, with an
undeveloped bell, and the tube was open, though appearing closed by
the juxtaposition of the edges of the leaf margins, or perhaps some-
times closed by the concrescence or adhesion of these edges, a mode of
development which he argues is different from that of the normal leaf
which arises as a peltate structure. This idea has been elaborated by
Troll who generally supports Goebel's thesis that the tube of the
pitcher of Sarracenia and Heliamphora is fundamentally a peltate leaf.
The condition of concrescence or adhesion of the leaf margins is ac-
tually realized in the case of H. Tyleri in its fully developed normal
leaves, as we shall see.

The latest described species, H. minor Gleason was found by Dr.
G. H. H. Tate on Mt. Auyan-Tepui, Venezuela in December, 1937 at
an altitude of 2,200 meters. Generally similar to H. nutans, it differs
in the more sturdy and less graceful leaves, 10 to 12 cm. long. The
spoon is larger and deeper, and orbicular in form. The bell is densely
hairy only along the marginal zone, with a few scattered small down-
ward directed hairs on the general surface, with many nectar glands.
The slight constriction at the base of the bell is hairy, with slen-

Francis E. Lloyd — 12— Carnivorous Plants

der hairs, and very glandular. The twin hairs on the outer surface are
finely tuberculate (Text fig. i).

The other recently described species H. Tatei, H. Tyleri and H.
Macdonaldae (Gleason 1931) are tall shrubby plants (4 feet), but
otherwise present only a few differences which concern us here._ The
leaf is in all three much elongated, the major elongation being in the
bell which becomes tubular, expanding only at the top where it is sur-
mounted by a rather large and massive overhanging lip-Hke appendage,
which, hke the spoon in //. nutans, carries large nectar glands. The
most divergent of the three is H. Macdonaldae, in which the inner sur-
face of the bell is quite smooth except along the free margin and for a
narrow zone at the lower hmit of the neck about 2 cm. wide. The
distribution of small glands on the outer surface is much the same as
in H. nutans except that they do not occur on the outer faces of the
extensive stipular wings.

Glands are absent from the interior surface of the bell where the
surface is smooth, while even when in H. nutans the surface of the bell
is smooth, as it sometimes is, glands occur nevertheless.

An adaptive feature of very great interest is one reported by Tate :—

"The question arose as to how the pitchers, closely packed

and unable to bend over as they were, maintained a constant water
level and succeeded in getting rid of the excess water poured into them
during the frequent heavy rains. Upon examination it was found that
each leaf had a small pore in the seam (opposite the midrib) placed
just at the juncture of the basal water containing part of the pitcher
and the terminal portion, through which the excess fluid_ might run
out. This observation was made on H. Macdonaldae, but in all prob-
abihty holds for other species as well. " (Tate, quoted by Gleason,

1931).

I have examined the material collected by Dr. Tate with much
care and interest, so far as it was permitted, and I have found that
the condition described above is to be found in its most pronounced
form in H. Tatei. To understand the morphology involved _ let us
compare the structure of the leaf with that of H. nutans, in which the
closed tubular portion ends abruptly in a bell, sht for a short dis-
tance along the ventral border, the margins running downward as the
ventral wings as above described, to be continued as the stipular
wings. In H. Tatei the leaves are from 40 to 50 cm. long, and the
stipules, in accordance with the shrubby habit of the plant, are very
long, clasping the bearing stem and adjacent leaf bases. The ventral
wings are short and fohose, while the bell is very long, being slit
down the ventral side for some distance. The edges of the trumpet
are here also confluent with the ventral wings, but above the morpho-
logical limit of the slit they are concrescent except for a space of about
one centimeter at its lowest limit. Here the edges of the bell remain
free, forming a short elongated-oval slit, which we may call the drain-
slit or pore (Text fig. i). The head of water released by flowing
through the slit amounts to 5 cm. in some leaves, or possibly more,
often less. This would have the effect of lowering the center of gravity
of the water loaded leaf very considerably. In //. Macdonaldae, in a
leaf of nearly the same length (37 cm.), the region of fused edges was

Chapter I

13

Heliamphora

much shorter in the only leaf I could examine, while in H. Tyler i (of
which I examined two leaves) there is but a shght commissure above
the drain slit, or it may be quite open, as it is in H. nutans. The
length of the slit is about i cm. more or less, and has cihated margins,
since it lies within the hairy zone of the bell near its lower limit.
Owing to the fact that the specimens were dried and pressed "types",
I was limited in my examination. Where the slit of the bell in H.
Macdonaldae seemed to be open, this may have been due to the sepa-
ration of the fused margins in drying (Lloyd 1933).

Fig. I. — I, Heliamphora Tatei; 2, H. Macdonaldae; 3, H. minor, the zonation of the
interior surface is indicated by the numbers 1-4; 4, H. minor; 5, H. nutans; 6, retrorse
hair of H. minor; 7, twin hairs of H. minor.

It is quite evident that the presence of the drain slit discovered
by Tate would render tall plants, which grow to the height of four
feet and inhabit a very rainy habitat, far less top-heavy. The up-
right position of the leaves is further assured by the ample, tightly
clasping, stipular wings. At the same time it is to be observed that
the adaptation is not equally expressed in all three species, being least
so in H. Tyleri, in which the slit may not be present at all, the bell
being slit all the way down to the limit of the hairy zone. So far as
the effect is concerned, this amounts to the same thing, since the
water is drained off to the lowest open point of the bell in any case.

Trichomes. — There are two kinds of hairs to be found on the inner
surface of the pitcher, those of the bell and adjacent tube, that is of
the conducting surface of Hooker, and those of the basal portion of
the tube. They are of identical morphology, but differ in important

Francis E. Lloyd — 14 — Carnivorous Plants

details of structure. They are unicellular hairs with the bases embed-
ded in a raised mass of epidermis and underlying parenchyma. They
taper from the base gradually to the sharp tip, and the cuticle is
raised to form folds or ridges, beginning at or near the base and con-
verging on the apex, where they gradually fade away. These ridges
are more pronounced on the hairs of the conductive surface, and are
much weaker to scarcely distinct on the hairs of the detentive surface.
The former are relatively longer, more slender and distinctly flexible,
the more so in the lower parts of the conductive surface. Krafft
suggested that these ridges have use as strengthening elements, but
the effect is not to make the hairs rigid. They may make it more
difficult for flies to use their foot organs. Their chief eflficiency lies,
I imagine, in their number and flexibility, so that an insect cannot
place its feet on the bell surface, but can only hook onto the hairs
which by their flexibility give way and permit the insect to slip into
the pitcher. The hairs of the detentive surface on the other hand
are short, very thick walled and rigid, thereby making escape difficult.
Bentham (1840) said of these that ''they have all the appearance of
ordinary secreting hairs," but this was a mistake. They are cer-
tainly not secretory.

The structure of the nectar gland of the pitcher in Heliampkora
has been described, though not quite correctly, by Krafft (1896).
There are, for purposes of description, two kinds: (a) those which
are relatively small and have few cells, about 12 in number (2 — 10).
These are found scattered over the whole of the bell, and on the outer
surfaces, including the wings, and are regarded as nectar glands
by GoEBEL, who detected a sweetness in the excreted fluid; and
(b) large glands found only on the inner surface of the spoon (2 — 7).
Krafft spoke of three kinds of glands, distinguishing those on the
inner from those on the outer surfaces. But they are all quite the
same in structure, differing somewhat in size, those on the outer sur-
faces being shallower, the glandular cells of a gland having little more
depth than that of the surrounding epidermis.

The structure of the smaller glands is as follows: They appear in
surface view to consist of six cells, four in a single course in the level
of the epidermis, covered partially by two cells, the "cover cells" of
GoEBEL. The area of the exposed surface of the gland is about equal
to that of a stoma. It is difficult to resist the theory that the glands
are derived phylogenetically from stomata, but there is no support
for this beyond the suggestive appearance of the two cover cells.
Beneath the course of four cells, there is a second inner course of cells
which appeared to Krafft to be four in number. One usually can
count at least six cells, two cells lying beneath the two cover cells.
But the glands are of irregular structure, and one cannot say definitely
that there are only so many cells. These constitute the gland proper,
and are all derived from the epidermis. They are surrounded by a
cuticularized membrane, except at the base, which, though being
partially covered with this membrane, is not entirely so, there being
left a "window" (Krafft used this word), so that the active gland
cells lie in direct contact with usually two, sometimes one, or three
to four, parenchyma cells below.

Chapter I — 15 — HeUamphora

The walls of these cells which lie in contact with those of the
gland immediately above (if we assume an orientation of up and
down, the cover cells being up) are always strengthened by curved
thickenings like those of xylem vessels. Krafft seems to have re-
garded these cells also to be cuticularized, and the wall thickenings
to be on the lateral walls in contact with surrounding parenchyma
cells. On both these points he was mistaken. The wall thickenings
are found on the walls only where the cells impinge on the gland cells,
and are not cuticularized in any part. This arrangement is found also
in the glands of Sarracenia, as Goebel showed, to be described be-
yond. The function of these parenchyma cells is not known, but it
serves to call them transmitting cells. But whether they do more
than permit movement of substances from the leaf tissues to the
gland, is not known. There is no reason to particularize as to the
distinction between the glands of the inner and outer surfaces, be-
yond the fact that where the outer surface glands impinge on the
underlying parenchyma, the parenchyma cells in immediate contact
are sometimes so large that it is very easy to recognize the fact that
the wall thickenings occur only where the gland cells are in contact
with the parenchyma cells. In cases where the section has run through
the gland at right angles to the common axis of the contingent pa-
renchyma walls beneath the gland, it becomes apparent why cell
walls carrying the thickenings appear to be other than those exactly
in contact with the glands, namely, because of the oblique position.
In making a drawing one is usually forced to show them as if they
were anticHnal, instead of periclinal walls. For this reason one can
understand how Krafft may have been misled. Viewing the gland
from beneath, possible in tangential sections, leaves no doubt about

the facts.

The glandular cells and the contingent parenchyma cells were
regarded together as constituting the gland by Krafft, and the
whole was attributed by him to an epidermal origin. But the pa-
renchyma cells are certainly not of glandular nature, judging by their
meagre protoplasmic contents, and are of equal certainty not of epider-
mal origin. One hardly needs to see developmental stages to draw
this conclusion. Tenner's account of the origin of the Sarracenia
nectar gland, which is of the same structure as that of HeUamphora,
also includes the parenchyma cells.

The large glands are found only on the inner surface of the spoon.
Krafft correctly traces this type of gland to an epidermal origin,
but does not show that in this gland also there are to be found the
xylem-Hke parenchyma cells. He attributes to the gland an identity
with the large glands found in Cephalotus which, as we shall see when
we discuss the latter, is not justified. These glands occur to the
number of about 20, larger and smaller in size, the larger ones being
in a more lateral position. And these are very large. In surface
view they consist of a number of cover cells and first course cells,
underlain by about four courses of thick walled cells beneath which
lies the mass of glandular cells proper. The periphery of this mass
is irregular, as if there was a tendency of the glandular mass to branch
or lobe. The contingent parenchyma by the same token intrudes

Francis E, Lloyd — 16 — Carnivorous Plants

cells between the lobes. So large are these glands that they must
contain about looo cells. There are only a half-dozen such large
glands, the rest within the spoon being of various sizes, but all smaller
and showing a structure more obviously like that of the rest of the
glands of the inner and outer surfaces. One of the smallest I show
in Plate 2 — 7, in which it is seen that some of the parenchyma cells
in contact with the gland have spirally and reticulately thickened walls.
One finds them in the large glands, but only occasionally; they are
very difficult to find, however. In any event the wall thickenings
seem to be less pronounced and distinct. The large gland is often
in close contact with the vascular tissue.

The cuticularization, as in the small glands of the general surfaces,
extends around the gland, with, however, areas furnishing contact
with the surrounding parenchyma.

GoEBEL leaned towards the opinion that the glands above de-
scribed are different in structure from those of Sarracenia, but the
evidence to be later deduced will, I think, show otherwise. Mac-
FARLANE described the glands as being depressed and surrounded
by thick walled neighbor cells, "the upper part of which may over-
hang" the gland. He did not study the gland structure further,
as Krafft did later, and suggested that the glands of Heliamphora
stand in an intermediate position between those of Sarracenia and
those of Nepenthes. This idea seems lacking in justification.

Prey and its fate. — That insects are caught by the pitchers of
Heliamphora was not known to Bentham, who handled the first ma-
terial to reach London. I found insect remains in the pitchers of
H. Tyleri and Krafft did so in the pitchers of H. nutans sent to him
from England by Veitch and Sons. The odor accompanying decay
of insect bodies was noticed by Krafft. As there are no digestive
glands to be found, we must conclude that the proteins of animal
bodies are made available to the plant only by means of bacterial
digestion. That no work has been done on this plant can be ex-
plained by the rarity of the material, due to the difficulties of cul-
tivation.

Literature Cited:

Arber {see under Cephalotiis).

Bentham, G., Heliamphora nutans, a new pitcher plant from British Guiana. Transact.
Linn. Soc. 18, 1840.

Fenner, C. a., Beitrage zur Kenntnis der Anatomie, Entwicklungsgeschichte und Biologic
der Laubblatter und Driisen einiger Insektivoren. Flora 93:335-434, 1904.

Gleason, H. a. and E. F. Killep, Botanical results of the Tyler-Duida Expedition. Bull.
Torr. Bot. Club 58:277-586, 1931.

Gleason, H. A., Brittonia 3:164, 1939 (Description of Heliamphora minor).

GoEBEL, K., Pflanzenbiologische Schilderungen. Marburg, 1891.

Im Thurn, E. F. and D. Oliver, The botany of the Roraima Expedition of 1884. Trans.
Linn. Soc. London II, 2:249-300, 1887.

Kraeft, S., Beitrage zur Kenntnis der Sarraceniaceen-Gattung Heliamphora. Diss. Mu-
nich, 1896.

Lloyd, F. E., Isolation and the origin of species. Science, N. S. 22:710-712, 1905.

Lloyd, F. E., The carnivorous plants. Trans. R. S. C. 27:3-67, 1933.

Macfarlane, J. M., Observations on some pitchered insectivorous plants, I. Ann. Bot.
3:253-266, 1889; 11. Ann. Bot. 7:403-458, 1893.

ScHOMBURGK, R. H., Reisen im Guiana . . . Leipzig 1841.

Troll, W., Morphologic der schildformigen Blatter. Planta 17:153-314, 1932.

ZrppERER, Paul, Beitrag zur Kenntnis der Sarraceniaceen. Diss. Erlangen, 1885.

Chapter II
SARRACENIA

Discovery. — Known species. — Descriptions of: 5. purpurea. — S. psiUacina. — S.
Courtii. — S. minor. — S. Drumtnondii. — S. flava. — S. Jonesii. — Morphology of the leaf.
— Digestion and absorption. — Animal life of the pitchers.

The genus Sarracenia is based on Tournefort's original description
(1700) of a plant sent to him by Dr. M. S. Sarrazin from Quebec,
Canada. The name was adopted by Linnaeus (1737). The earliest
known illustration of Sarracenia is to be found in de l'Obel's Nova
Stirpium Adversaria, evidently of a leaf of 5. minor, probably re-
ceived from some Spanish explorer in Florida (p. 430, 1576 ed.). Ac-
cording to Uphof (Engler and Prantl, 2 ed.) there are nine species.
Wherry (1933) distinguishes between the northern form of S. pur-
purea and a southern form, namely, the subspecies S. purpurea gih-
bosa and 5. purpurea venosa, respectively. All species are distinguished
by the possession of pitcher leaves either upright or decumbent, of con-
siderable variety of form, to be detailed later. The following is a
list of the species and their geographic distribution, according to
Wherry (1935).

Species with upright tubular pitcher leaves: —

S. oreophila (Kearney) Wherry. Green pitcher plant. Of very
hmited distribution: Taylor Co., Georgia, and in the Appalachian
Mountains of Alabama (Cherokee, DeKald, and Marshall Cos.).

S. Sledgei Macfarlane. Pale pitcher plant. S. Alabama, Missis-
sippi, Louisiana and E. Texas. One or two colonies are reported
to survive in the Cumberland Plateau of Tennessee, presumably in
its ancestral home before the rise of the peneplain of the Cretaceous.

S. flava L. Yellow pitcher plant (1 — 7). N. and S. Carolina,
Georgia, extreme N. Florida and S. Alabama, in the coastal plain.

S. Jonesii Wherry. Red pitcher plant. There is a singular and
striking survival of this plant in an isolated spot in Buncombe and
Henderson Cos., N. Carolina. Otherwise it is found chiefly in S.
Alabama and in restricted regions nearby in Florida and Mississippi.

S. Drummondii Croom. (7 — 6). White-top pitcher plant. Chiefly
S. Alabama, with slight extensions into Mississippi, Georgia, and N.
Florida. It forms two isolated colonies in Georgia (Sumter Co.) and
in Florida (Madison Co.).

S. rubra Walter. Sweet pitcher plant. North Carolina (from
Moore Co. southward) through S. Carohna into Georgia, away from
the coast except in N. Carolina.

S.minorW&W. {1 — 9). Hooded pitcher plant. " Fly-traps " (Mel-
LiCHAMp). Eastern half of the peninsula of Florida, in the north of
that state, southern Georgia, western S. CaroUna and slightly into
N. Carolina.

Species with decumbent leaves: — •

5. psittacina Michaux (/ — 8). Parrot pitcher plant. From a

Francis E. Lloyd —18— Carnivorous Plants

short distance E. of New Orleans, Mississippi, through S. Alabama,
S. Georgia to the coast and in N. Florida.

S. purpurea venosa (Rafinesque) Wherry. Southern pitcher plant.
This species has an interrupted distribution from southern New Jersey
to S. Mississippi. Only small isolated colonies are to be found be-
tween N. Carolina, where it is widespread, and a similarly wide-
spread area in S. E. Georgia, extreme N. E. Florida, S. Alabama,
from which a narrow tongue extends to near the Mississippi River
N. of New Orleans.

S. purpurea gibbosa (Rafinesque) Wherry. (i—SS)- Northern
pitcher plant. Found throughout a vast area, beginning with a nar-
row strip embracing the coastal regions of Maryland, Delaware and
New Jersey, it spread westerly through N. Pennsylvania, N. Ohio,
N. IlHnois, Wisconsin, through the whole region north and east to
the Atlantic coast, and N. W. through the region of Winnipeg into
uncharted regions. The northern limits are not known.

Sarracenia purpurea Linn, has had a long history, and we are in-
debted to Hooker (1875) for digging out the facts. From_ an early
sketch by an unknown author, which found its way to Lisbon and
thence to Paris, Clusius (Rariorum pi. historia, 1601, p. boodj)
published a figure, which thirty years later was copied by Johnson
in his edition of Gerard's Herbal, in the hope that someone would
find the plant. The hope was reahzed when John Tradescant,
whose name is perpetuated in the genus Tradescantia, ^ found it in
Virginia and succeeded in bringing it ahve to England in 1640. In
1700 Tournefort described the plant, naming it Sarracenia (or Sar-
racena) in honor of Dr. M. S. Sarrazin, who had sent it to him from
Quebec. The name was adopted by Linnaeus in his Hortus Clif-
fortianus, 1737. The plant in question is then called Sarracenia
purpurea L., and is the best known of all the species chiefly by reason
of its above mentioned wide distribution.

Quite naturally the structural features of these peculiar plants
were the first to attract attention. The terminal lobe or flap not
only looked like a Hd, but was believed by Morison (Plantarum
Historiae, 1699, 3:533) to be hinged and capable of movement, as
many non-botanists believe today. Linnaeus and others adopted
this idea, thinking this behaviour to conserve the water within. Bur-
nett (1829) seemed very sure of this. "In many instances the ap-
paratus is fitted with a lip or Hd, by which the mouth may be shut or
opened; the machinery of which limb is so contrived that, when the
cavity within is well supphed, it closes to prevent evaporation; and
when the stock is diminished or consumed, the lip is raised, so that the
mouth is again raised to receive the falhng rain or rising dew."
Catesby (Nat. Hist, of CaroHna 2:69, 1743) had the idea that the
hollow leaves were a refuge for insects from the animals (frogs, etc.)
which might devour them. William, son of Charles Bartram, in
his Travels in N. and S. Carolina, Georgia and Florida (1791) re-
corded the objection suggested earlier by Collinson (see Smith,
182 1) that many insects, on the contrary, are caught and destroyed
in the pitchers.

Later more meticulous observations on S. adunca led Macbride

Chapter II — 19 — Sarracenia

(1817) to find that the tube of the pitcher leaf is lined with down-
wardly pointed hairs, which he could "plainly see at the bottom of
the tube," and he saw also that there is about the mouth of the tube
a viscid substance which attracts flies.

Then Mellichamp, a physician like Macbride, resident of the
same region and a contemporary of Hooker, did the first real ex-
perimentation on this plant and compared the rate at which fresh veni-
son showed disintegration in the pitcher fluid and in distilled water,
concluding that bacterial action was at work. He found also that
the pitcher fluid did not allow the escape of flies when they fell into
it as water does, indicating that there is an "anaesthetic action."
There is also, he saw, a nectar baited pathway up the outside of the
pitcher to its mouth. Thus Mellichamp's work opened up the field
of physiological research to which reference will be made later.

Since then there have been numerous descriptions of the structure
of the various plants of the genus, not always of unimpeachable ac-
curacy. We now consider this aspect of the matter in what follows.

Sarracenia purpurea is the most widely distributed, and longest
and best known, species, ranging from Labrador to Florida along the
Atlantic seacoast of N. America, and westward to Wisconsin and
Minnesota, successfully withstanding the rigors of the northern win-
ters. It has been successfully introduced in Switzerland. With S.
psiUacina Michx. it is associated in the section Decumbentes Uphof,
both being characterized by having their leaves more or less spread-
ing as a rosette. It is to be found in bogs, usually in company with
much Sphagnum, anchored therein by its strong ascending rhizomes
clothed with the remnants of dead leaf bases and sending out thick
fibrous roots. It may often be found in company with other plants,
making large floating or semi-floating masses of vegetation about the
edges of ponds, as described by Macfarlane and Steckbeck for
Davenport Lake, near Toms River, N. J., U. S. A. (1933). In com-
mon with carnivorous plants in general the roots are devoid of mycor-
rhiza (MacDougal 1899).

It is a beautiful plant (/ — 3). Its leaves, which have a very grace-
ful form, are clustered into a rosette and are deep green with rich
crimson markings along the venation of the "flap" and more or less
uniform similar coloration in the upper portion of the body of the
leaf, depending on the exposure to fight.

Form and structure of the leaf. — The pitcher leaf of Sarracenia
has many times been the subject of description from the anatomical
point of view by Vogl (1864), Macbride (181 7), Mellichamp (1875),
Hooker (1875), Zipperer (1885), Goebel (1891), Macfarlane
(1889, 1893).

Aside from the cotyledons, which present no especially peculiar
features, there are two forms of the pitcher leaf, a juvenile and a ma-
ture, both mentioned by Troll (1932). The mature form may be
likened to an elegant cornucopia curving however only in one plane.
Arising from a winged base, the wings embracing the bearing stem, it
becomes cylindrical for a shorter or longer distance according to cir-
cumstances. The lower part of this is soHd; in its upper part may
be found the deepest portion of the hollow interior of the pitcher.

Francis E. Lloyd — 20 — Carnivorous Plants

Along the upper (ventral) surface of the cylindrical petiolar portion arises
very gradually a single ridge, the ala ventralis, which attains consider-
able depth further up the leaf (/ — 4; 2 — 12). At the same time
the leaf becomes expanded into a curved conical hollow vessel, ex-
tending to the mouth. As this point is approached the ventral wing
begins to show evidence of a double character in that its edge is longi-
tudinally fissured to form two parallel ridges which are continuous
with the edges (nectar roll) of the bell, which in this species has a very
peculiar form. The abaxial two-thirds are expanded into a cordate
"flap," the sides of which where they meet the adaxial part of the
bell become helicoidal. The edge of the helix can be seen to continue
as the edge of a convolute margin of the pitcher — the adaxial part
just mentioned which I have called the nectar roll, and which is noth-
ing more or less than the adaxial part of the bell. Troll's (1932)
description says that the abaxial part of the pitcher is lengthened as
the lid. This is true enough as far as it goes, but this and all other
descriptions, as far as I am aware, neglect to point out the nature of
the curious rolled margin around the adaxial limb of the opening.
At the midpoint of this rolled margin there is on the surface no sign
of its nature. Examination by means of sections, however, shows that
the two ridges of the ventral wing spread to right and left to continue
as its involuted margins. One is reminded of the volutes of the
capital of an Ionic column. To sura up in a word, the flap and nectar
roll are to be taken together as the edge of the pitcher, wide and
leaf-life abaxially and tightly rolled outwardly adaxially.

But this comes out quite clearly in the juvenile form of pitcher,
found on seedlings and delicate shoots. These are slender, the tubular
portion being narrow and only very gradually widening toward the
mouth (2 — 13). This is surmounted by an overhanging hood-like
expansion, the margins of which do not become rolled along the ab-
axial reach of the mouth, but run obliquely, meeting to form the
ventral wing. According to this description the ventral wing is a
single structure below and double above, so far as external evidence

goes.

Mrs. Arber (1941) has recently argued that the lid in Sarracenia
"is merely a localized development of the collar". This may be
questioned on the evidence above stated, that the collar (nectar roll) is
not present in juvenile leaves.

The whole of the outer surface of the mature pitcher is supplied
with scattered nectar glands (2 — 15, 16). It is also somewhat rough
and hairy with scattered trichomes (j — 2) with peculiar thicken-
ings in the form of waves of surface expansion rather than continuous
ridges, such as occur on the trichomes of the interior. The external
glands may be regarded, as Macfarlane suggested, as alluring in
function, leading creeping prey to the mouth.

The internal surface shows, as Hooker pointed out, distinct zo-
nation. He recognized four zones and described them as follows.
Zone I (2 — 12) embraces the cordate emarginate flap. The epider-
mis carries stomata, glands and strong downwardly directed hairs.
The lower limit of this zonation is clearly marked by an irregular line
where the character of the epidermis cells abruptly changes. In zone

Chapter II — 21 — Sarracenia

2, the epidermis cells have very thick outer walls each ending in an
umbo, and more or less imbricated with its neighbors below (2 — 14).
There are numerous glands here. The appearance to the eye is vel-
vety, the surface being broken up by the imbrication and by the very
numerous, fine, downwardly directed ridges concentering on each outer
cell wall on the umbo. This zone forms a collar about i cm. in width.
While zone i is highly colored with red along the venations, green be-
tween, zone 2 is less colored, though the red still follows the main
veins. There are no stomata here.

Zone 3 is smooth and glassy and reflects the light strongly. The
epidermal cells have wavy thick walls, and except for a narrow strip
just below zone 2, there are throughout numerous glands (I count
15-20 per mm 2), but no stomata. This zone occupies about one-
half the whole interior. Below it is zone 4, which is devoid of cuticle
(Batalin 1880) except for a small space surrounding the base of each
hair. These are numerous, downwardly pointed, long, slender and
glassy, and are effective in the detention of prey. The lack of cuticle
can be very easily demonstrated by exposing the interior surface of a
leaf to a weak solution of methylene blue (or other suitable dye) or
potassium permanganate. It shows some discoloration, being brown-
ish as compared with the rest of the surface. At the lower limit of
zone 3 the sinuous walled epidermis abruptly changes to an epidermis
with plain walls, and the cells appear strictly isodiametric. There are
neither glands nor stomata.

For the above zones we may adopt Hooker's descriptive terms,
which, for zone i, is attractive, zone 2, conducting, zone 3, glandular,
and zone 4, detentive, even though these terms are incomplete in sig-
nificance. Zone I is not only attractive but is also a place of very
insecure foothold, because of the form and direction of the hairs.
Zone 2 is also both attractive because of the nectar secreted, as in
zone I, and affords a precarious foothold. Zone 3 has a hard, glassy
surface, extending the glacis of zone 2, all three zones forming a. facilis
descensus Averno. Zone 4 is probably not only detentive, but also
absorptive.

And to these should be added a fifth zone. This is a relatively
narrow zone below zone 4, in which the cuticle is permanent, and
which is hairy only in its upper part, the lower being completely
smooth. There are no glands, and the epidermal cells are quite like
those of zone 4. Fenner (1904) calls this zone (but he was describing
S. flava) an absorption zone, but I think without good reason. It is
true that these cells do not completely resist the entrance of methyl-
ene blue, but this enters them much less easily than into the cells of
zone 4, though evidently more easily than into those of zone 3, which
are completely resistant.

MacDougal recognized the zones as I have described them.
When subjected to total darkness, the petiolar region of the leaf elon-
gates greatly, while the upper zones become shorter, zones i and 2
showing the greatest reduction in size. Fully etiolated leaves are
twice the length of normal ones, but the petiolar region, including
the basal part of the pitcher, is five times the normal length. The
ventral wing does not develop, the leaf being wedge-shaped in trans-

Francis E. Lloyd — 22 — Carnivorous Plants

verse section. The ascidium is present. Corresponding dimensional
changes take place in the component cells, but there is also, according
to MacDougal, an actual increase in the number of cells in the
elongated portion of the detentive region (which partakes in elonga-
tion with the petiole) and an actual decrease in the number of cells
of the conductive surfaces (zones i and 2). Thus the pitcher of Sar-
racenia behaves, in relation to light, as if it were a leaf blade (zones
I, 2 and the upper part of 3) and the rest as if it were petiolar (1903,
pp. 173-6). Iris, when grown in the dark, grew only slightly in excess
of the normal. We recall that Goebel compares the Sarracenia leaf
with that of Iris.

The juvenile leaves, which are also pitcher leaves, differ in some
details from the mature. While they display the same zonation, the
characters as they have been described for the mature leaves overlap.
They can be described as follows:

Zone I is the same as in the mature leaves. The margin of the flap
is ciliated with more or less curved blunt hairs. The epidermal cells
are sinuous walled, the glands present but few. Stomata are present
and the hairs stout and curved.

Zone 2. The epidermis of zone i changes abruptly in that all the
cells become trichomatous, but very short and produce the effect of
imbrication, as described above. Glands are large and numerous,
more so towards the lower limit of the zone, where the epidermis
again changes.

Zone 3. The epidermal cells are again sinuous walled but, unhke
the mature leaf, there are numerous trichomes. Glands are here also
present, but no stomata.

Zone 4. The epidermal cells become straight, the cells isodiametric,
with numerous very slender hairs, and no glands, and no cuticle.

Zone 5 has the same sort of epidermis as zone 4, but is devoid of
hairs.

As we shall see, the, juvenile leaf of S. purpurea resembles in struc-
ture that of 5. psittacina.

We come to the details of the glands and trichomes. The glands
(j — i) are all of one type (the ''Sarracenia type," Goebel, 1891).
Viewed as part of the general surface, each gland exposes normally
six cells to view, four in a rough circle and two, the cover cells (Goe-
bel) in the middle (2 — 15). The cover cells overlap the four bor-
dering cells to a greater or less extent. Those on the outer surface
have relatively large cover cells, which jut out further beyond the
general level of the surface. Those of the glands of zone 2 are rela-
tively much smaller.

The outer surface glands (2 — 15, 16) are the smallest and simplest
in structure, derived from a single epidermal cell, according to Zip-
perer and to Fenner. Undoubtedly the gland is of epidermal ori-
gin, and the idea is not precluded that it may represent originally a
stomatal apparatus, the cover cells arising originally as guard cells.
But this is admittedly speculation. The four peripheral cells lie in
the general level of the surrounding epidermal cells, while the two
cover cells are conical and are wedged in between the peripheral cells.
Against the interior faces of the peripheral cells there is usually one,

Chapter II

23 — Sarracenia

sometimes two cells, the adjunct cells, which appear to be derived
from the parenchyma. The walls of the adjunct cell, or cells in con-
tact with the gland cells, are variously thickened, reticulately and
circularly, as represented by Goebel (1891). Fenner thought that
the adjunct cells are also derived from the epidermis (Fenner called
them "Durchlasszellen"), but their denser protoplasmic content in-
dicates that they have more than a passive role. The outer walls
of the cover and peripheral cells are all cuticularized, staining with
fat stains (Congo Red, etc.) except a part of those making the contact
with the adjunct cell or cells. These walls resist sulfuric acid along
with the outer walls of the surrounding epidermal cells. But meth-
ylene blue easily permeates these glands in the living leaf.

The interior glands (j— i) are of similar structure, but are of
two courses, evidently, as Fenner indicates, derived from originally
four peripheral cells. The gland, aside from the adjunct cells, there-
fore consists of normally 10-12 cells, the outer course having four periph-
eral and two cover cells, the inner course usually four to six cells. The
cover cells are slenderly conical, are wedged in the middle of the outer
peripheral cells, extending inwardly till more or less in contact with
the inner course. Their walls are considerably thickened and cuti-
cularized throughout. The base of the gland is in contact with two
to four or five adjunct cells with reticulate, circularly or spirally
thickened walls of contact which give a positive reaction with phloro-
glucin and HCl. Here, however, there is no suberization of the gland
cells so that there is left a "window" (Fenner's term) allowing
communication by ready diffusion between the adjunct and gland
cells proper. The glands of the nectar roll, while identical in struc-
ture with those elsewhere, show a certain distortion consequent on the
growth movement resulting from torsion during the development.

The structure of the glands in whatever zone they occur is the
same, though they function differently. On the outer surface of the
pitcher and on the inner in zones i and 2, they secrete nectar. Though
Hooker was in doubt on this point, I am sure of it from my own
observation. I have also, on a warm, sunshiny day watched flies in
numbers busily sucking the nectar and some of them getting trapped
by slipping down the surface of zones i and 2. The glands of zone
3, however, probably secrete digestive ferments, judging from the
results of Hepburn et al. (1927), to be discussed later.

Sarracenia psittacina Michx. — This species (i — 8) is associated
with 5. purpurea in the Decumhentes by Uphof, because of the posi-
tion of the leaves which lie more or less parallel with the ground. The
leaf considered as a trap, is, however, quite different in this species,
and is much more efficient mechanically — or at least it appears so.
No account I know of quite brings this point out. The leaf which
Goebel described as the "first pitcher leaf" is a juvenile form (3 — 5).

Its habitat is low, wet, sandy meadows subject to inundations
by the acid waters of nearby swamps (Wherry).

It is regarded taxonomically as associated with S. purpurea, but
it is as much or more like Darlingtonia. The pitcher consists of a
narrow, tapering curved tube (j — 4), somewhat flattened dorsiven-
trally, with a wide ventral wing, and with the top of the pitcher

Francis E. Lloyd — 24 — Carnivorous Plants

curved over to form a hood, with an entrance of narrow caliber facing
horizontally, instead of downwards as in Darlingtonia. It is deep
red in color, mottled by angular white fenestrations which allow
diffused light to enter the tube on all sides, more especially on the
ventral aspect of the tube which, because of the decumbent position
of the leaf, hes uppermost. In some other species, the fenestrations
occur on the dorsal aspect of the tube. The pitchers are rigid, and
are of striking shape, suggesting the specific and common name. The
above mentioned fenestrations have a more opaque look than those
of Darlingtonia, and on examination they were found to have extensive
intercellular spaces, the effect of which is to diffuse the Hght, thus
producing a snowy whiteness.

The form and structure of a pitcher can be best seen in one cut
sagittally (j — 4). The upper part of the tube is strongly curved, so
as to direct the opening toward the leaf base. The end of the mid-
vein is indicated by a low umbo, the organic apex of the pitcher. Be-
yond this point the hood is closed by the forwardly curved lobes of
the "flap," the margins of which are closely apposed, sometimes even
to mutual adherence, though sections show their histological inde-
pendence (j — 6). The more ventral reaches of these lobes are en-
larged and bend inwards to form the entrance tube. Inspection of
this shows that it is formed partly by the upper short stretch of the
ventral pitcher wall and the proximate parts of the lobes, a condition
dupHcated in the aberrant juvenile leaf of Darlingtonia (5 — 3). Thus
is formed a short cylindrical entrance tube (j — 6), making the trap
of the lobster-pot type. The inner free edge of the entrance tube is
stiffened not only by the strong epidermis, but by a weal running
parallel to the free edge {3 — 4). This weal is continuous with the
exact edge of the lobes above the entrance tube, and must be re-
garded as the morphological margin. The edging beyond this forms
a shelf bearing numerous nectar glands, and is clothed on both sides
with tessellated epidermis of umbonate, striated cells, characteristic
of the inner general surface of the entrance tube, where also glands
are found.

This shelf corresponds exactly with the inwardly curved nectar
roll of Darlingtonia (5 — 10), from the pitcher of which that of S.
psittacina differs in the fact that the organic apex of the tube lies
within the periphery of the hood, while in Darlingtonia it lies beyond.
The two lobes of the fishtail appendage of Darlingtonia correspond
to the two lobes of S. psittacina. Macfarlane regarded this species
as the most aberrant of all the species of Sarracenia, and its similar-
ity to Darlingtonia supports this view. It is on the whole more
similar to Darlingtonia than to 5. purpurea.

The earlier stages in the development of the leaf are practically
indistinguishable from that of other species examined. The feature
peculiar to this species, however, the infolded edge of the flap, is a
character which appears quite late in the course of development. In
a leaf which, though embryonic in form, was large enough to be ex-
posed to the Hght, the hood measured about 0.75 mm. and in this
the fold has just commenced to develop (5— 11). In one with the hood
2 mm. long the ingrowth was marked, but far from fully developed.

Chapter II —25— Sarracenia

In this early stage the structure resembles that of Darlingtonia, save
for the forward developing fishtail appendage, which here does not

appear at all.

The interior surface of the pitcher presents a zonation which cor-
responds roughly with that in S. purpurea, but is by no means as
distinct as in that species. This lack of distinction arises from the
overlapping of the zones 3 and 4, and the restriction of zone 2 to the
inner surface of the entrance tube, the ventral portion of which is part
of the pitcher tube proper. This point must be appreciated as other-
wise it would be difficult to recognize zone 2 at all.

Zone I (j — 4) is the whole of the inner surface of the hood ex-
cepting the inturned edge of the re-entrant tube. The epidermis is
of wavy-walled cells with stomata and many nectar glands, and scat-
tered, relatively few downwardly directed, curved weak hairs, as com-
pared with the flap of 5. purpurea, or Heliamphora. Since the chief
mechanism for the capture of insects is the re-entrant tube, these hairs
are of little importance. This is compensated for by the presence of
very many stiff hairs in the dorsal aspect of the dome, to which zone
3 reaches.

Zone 2 corresponds to this zone in S. purpurea in function, though
it is not a complete zone geometrically speaking. It embraces a short
reach of the ventral wall of the pitcher with the contiguous sides of
the re-entrant tube, including the shelf, which is within the morpholog-
ical edge of the lobe. The shelf and the inner surface of the tube
sides are clothed with an epidermis of tessellated straight walled cells
each with a low striated umbo, pointing inward and downward, as in
zone 2 of S. purpurea. Glands are present, in greater number proxi-
mally than distally. They are absent under the shelf, but the shelf
itself bears a great many. This is the principal lure evidently, but the
insect which advances into the re-entrant tube to sip the nectar is
invited to enter further by the shining white fenestrations, mullioned
in red, of the pitcher wall. The outer surface bears many small glands
also, which, with those of the general outer pitcher surface, constitute
a general lure.

Zone 3. Like zone 3 of 5. purpurea this carries some stomata and
many glands, but, unlike that species, the whole surface is clothed
with a dense felt of downwardly pointed, slender stiff hairs, continuous
with those of zone 4. It may be described as an advance of zone 4
to overlap zone 3. The parallelism between this species and S. pur-
purea is seen in the many glands of this zone. These are possibly
peptic glands, though the evidence is at present not conclusive {see
p. 34). The epidermis is of a mixture of wavy- walled cells, and smaller
straight walled cells, these becoming more numerous as the next zone
4 is approached. It is underlain by a course of wavy-walled cells, the
walls of which are thick and afford stiffening to the pitcher wall._

Zone 4 is devoid of glands, but has a dense clothing of trapping
hairs down to the very end of the pitcher tube, where they are shorter,
fitting better the reduced bore of the tube, and leaving a lumen.
Macfarlane described the whole of zones 3 and 4, as above delimited,
as the detentive zone, but having glands in the upper one-third. In
S. purpurea zone 3 is secretive, and has a glissade surface, while in
S. psittacina the surface is hairy.

Francis E. Lloyd — 26 — Carnivorous Plants

The structure of the glands is like that in S. purpurea (3 — 3).

GoEBEL described the seedling leaves which appear directly fol-
lowing the cotyledons. In form they resemble closely the juvenile
leaves of S. purpurea. The leaves of the two species in the juvenile
state are regarded by W. P. Wilson (1888) as indistinguishable, and
he regarded them as closely related. The presence of the umbo, how-
ever (5 — 5), clearly separates S. psittacina from S. purpurea. The
forward margin of the mouth is simple, and the inturning valvular
nectar roll with its marginal thickening is absent. Absent also are
the two lateral lobes. The interior surface is divided into four zones:
(i) the under surface of the hood, with scattered retrorse hairs with
interspersed glands; (2) the ghding zone with a Hning of imbricated
cells with downward directed points; (j) a wide zone with many long
downward pointed hairs with glands between; and {4) the bottom
zone with smooth epidermis. Goebel's zone (2) corresponds to zones
2 and 3 in my description of the adult form of leaf above.

Sarracenia Courtii. — This is a hybrid between S. purpurea and
S. psittacina, and in its structure reflects the characters of both in the
zonation of the pitcher leaf. As in S. purpurea, the conductive zone,
zone 2, is broader than in 6*. psittacina, and occupies a transverse band
around the interior of the leaf, narrowing dorsally, thus separating
zones I and 3 almost completely. Zone 3 is much less hairy than this
zone in S. psittacina, but is glandular, as in that species. The general
aspect of 5. Courtii resembles that of 5. psittacina, but the plant is
larger.

Sarracenia minor. — In this species the leaf stands in a vertical
position, and the opening is overhung by a wide, domed lid (i — 9;
3 — 7-9)- The wall of the pitcher opposite the opening, and for some
distance up and down, is fenestrated with white patches as in Darling-
tonia. These are slightly thinner areas of the wall, devoid of chloro-
phyll, and there is no palisade tissue anywhere. These white spots
may be regarded as a visual lure for insects. The lower edge of the
mouth is thickened by an outwardly reflexed edge of the wall, as in
5. purpurea, to form the nectar roll. The body of the pitcher is a
tapering tube slightly curved, and carrying a wing in front — the
ala ventralis — which, has a double edge above, the edges flowing
right and left into the edge of the hd, but a single one below, and is
not confluent with the stipular wings of the leaf base. The ventral
wing starts at the top of the tube and attains its greatest width about
half way down.

The outer surface is sparsely hairy with short, curved hairs with
finely tuberculated walls. There are numerous glands scattered all
over the outer surface, and these are especially active in secretion along
the upper part of the edge of the wing, where drops of nectar which
have been excreted by them may be seen (5 — 9). I have not seen
nectar collecting visibly elsewhere on the outer surface. The interior
surface presents a zonation visible to the eye but somewhat dif-
ferent from that of 5. purpurea and more like that of 6". psittacina.
Mellichamp (1875) recognized three "belts" or zones: (/) embracing
the internal honey secreting portion; (2) a belt hned with soft and
velvety pubescence affording no foothold for most insects; and (j)

Chapter II — 27 — Sarracenia

that of coarse straw-colored hairs extending to the bottom of the tube
where a watery fluid is secreted. Essentially correct, this description
does not specify closely enough the distinction which actually exists.
Though the zonation does not stand out so clearly as in S. purpurea,
we can nevertheless recognize four zones.

Zone I, that of the under surface of the lid which ends in an obhque
line extending obliquely upwards from the lower margin of the mouth.
This is covered rather densely with curved hairs downwardly directed.
Interspersed are numerous nectar glands which are evidently active,
for one can see minute drops of nectar studding the surface. At the
line of demarkation the epidermis abruptly changes to a smooth, con-
tinuous surface of tessellated cells, each of which is downwardly sharply
umbonate (j — lo). Interspersed are very numerous glands which
are very active. In the upper part stand large drops which run to-
gether to form a flood of nectar. This is continuous along the lower
lip of the mouth and for a centimeter down from there and elsewhere
around the tube. This I call zone 2. At its lower limit the umbonate
cells give way gradually to cells of identical structure, but having
the umbo lengthened into a longer slender spike, still with many
glands scattered between. This zone measures about 3 to 4 cm. in
depth. It is evidently glaucous, with a white sheen. This is zone 3.
Below begins zone 4, recognizable to the eye by the pale green, non-
glaucous appearance. It soon becomes brownish in color and bears
numerous scattered long slender hairs with a detentive function (j —
11), with more or less straight walled cells between. In the upper
part of this zone there are a few glands, but none much below a
depth of I cm. Below the epidermis in zones 2, 3 and 4, the pitcher
wall is conspicuously strengthened by a hypodermis of wavy walled
cells with walls rather thin above, but in the general region of zone 4
very thick and underlain by a second course of wavy walled cells with
thinner walls. It is obvious that these cells add materially to the
rigidity of the tubular wall. The epidermal walls themselves are
straight or only very slightly wavy. The edges of the lid are con-
tinuous with the true edge of the lower reach of the mouth border.
The nectar roll, as in 5. purpurea, is covered with tightly imbricated
umbonate cells with numerous nectar glands (j — 10, 12). On the
whole, the pitcher hning is similar to that of the juvenile pitchers
of that species, in which, however, zone 3 is much shallower than in
S. minor. In general form this species resembles S. purpurea with its
bell turned forward so as to shade the opening, but as far as the
epidermal lining is concerned is more like either the juvenile leaves
of S. purpurea, or the mature leaves of 5. psittacina. In the course
of evolution it is possible that S. purpurea has been derived from a
plant resembling S. minor simply by a change in the posture of the
leaves from vertical to spreading, and by an extension of one zone
at the expense of another. A change from 5". minor to S. psittacina
could have been accomplished by an additional elaboration of the
region surrounding the mouth by extending the dimensions of the
nectar roll, and reversing it, curling it to the inside instead of to
the outside. This is speculating, of course.

Sarracenia Drummondii. — This is a tall species, the trumpet-

Francis E. Lloyd — 28 — Carnivorous Plants

shaped leaves attaining a length of 2-3 ft. and standing in a strictly-
erect posture (/ — 6). The tapering tube gradually widens to the
top, but contracts somewhat just below the opening, the bulge being
scarcely wider than the opening, which is oblique. About two-thirds
of the margin is occupied by a nectar roll, the free edges of which
are continuous with the two edges of the ventral wing, as in other
species already described. The abaxial third of the edge of the open-
ing is extended into a spreading, over-hanging lid supported by its
broad stalk. The posture changes somewhat with age, passing from
a more horizontal to a more oblique position, and, according to Goebel
and others, serves to divert rain water from the interior of the tube.
The lid and upper portion of the tube are highly colored in a motley
of white and red, with green except in the white fenestrations, which
occur here as in S. minor and Darlingtonia.

The external surface is supplied with small nectar glands and is
roughly hairy on the upper part of the tube. The hairs, which are
like those in S. purpurea, point in various directions, and not uni-
formly in one direction.

The internal surface is clearly divisible into zones. Zone i is the
inner surface of the lid as far as a distinct line crossing the isthmus
supporting it. There are many glands and stomata, and the surface
is studded with curved, downwardly pointed hairs of slender form
and bending under the pull of a fly's foot. Zone 2 starts at the line
across the stalk of the hd, and, including the nectar roll, extends down-
ward inside the tube a distance of 18 cm. in a leaf 60 cm. tall. The
surface is clothed with oval cells which form short sharp hairs, retrorse
as elsewhere. There are large nectar glands on the nectar roll and
in the upper one-half of the zone, but none below. There are no
stomata (Zipperer). This is the conducting zone (Macfarlane).
Zones 3 and 4 are not separable to the eye except for the fact that,
in the upper region of the combined zones, which are detentive, there
are a few glands in a narrow belt just below the lower limit of zone 2.
Though to the eye the line of demarkation between zones 2 and 3 is
distinct, the transition under the microscope is not a very sudden one,
since the change from very short hairs (zone 2) to the very long ones
of zone 3 is gradual. In zone 2, every epidermal cell is a trichome
(except of course where glands occur); in the zone below this is true
only of relatively few epidermal cells. In the upper part of zone 3, the
epidermal cells have more or less sinuous walls, giving way soon to
cells with oval outline, and this region we may recognize as zone 4.
The bottom of the tube is quite smooth for nearly five cm. From the
whole of zones 3 and 4 the cuticle is absent. The tube throughout its
length is greatly strengthened mechanically by the presence of a wavy
thick walled second layer.

Macfarlane denied the presence of glands in the upper part of
the detentive zone, saying that there are stomata surrounded by
groups of cells. I think he was mistaken in this. The glands, which
are few in number, are somewhat distorted, owing to the character
of the epidermis of many imbricated hairs, but are nevertheless clearly
glands. There are no stomata (Zipperer).

In addition to the normal pitcher leaves there are others in which

Chapter II — 29— Sarracenia

the tube is not developed. The leaf then consists of a cylindrical
stalk with a wide ventral wing. At the apex there is more or less
of a depression. Such leaves are photosynthetic only, and have been
compared by Goebel to the unifacial leaf of Iris. These leaves de-
velop later in the growing season (Goebel).

Having favorable material at Munich, I arranged the upper part
of a leaf under a bell jar and put a blue bottle fly inside. The leaf
stood vertically in shallow water in a vial. The fly was soon attracted
by the nectar secreted by the glands of the external surface and gradu-
ally worked his way by an erratic path to the rim. Mounting this
he began to sip the nectar, either on the under surface of the lid as
far as he could reach without letting go his hold with his hind legs
on the edge of the lid, or of the rim. Swinging about he explored
the surface inside the rim, always hanging on by his hind legs; and
it was evident that he was aware of the precarious foothold, for he was
loth to free his hind foot or feet. But on getting what seemed to be
a foothold and reaching as far as possible for the nectar, he would
let go and then invariably fall plump into the abyss. A bit of the
tube was cut away above the water level so that he could escape,
and in consequence he performed for me again and again. His actions
were repeated a dozen times without failure of being trapped. If
he ventured on the under side of the lid, as he sometimes did, he
could remain there as long as he grasped the edge with one foot, seiz-
ing the hairs with the other; but the moment he let go of the edge,
down he fell into the tube. There is therefore no question but that
the surface of zones i and 2 is one which gives no foothold to such
flies and, to infer from the variety of prey found in the pitchers, to
most other insects as well.

Sarracenia flava. — This species resembles S. Drummondii in many
respects, but is stouter and coarser, and its prevailing color is greenish-
yellow, with the latter color quite dominant. The tube tapers gradu-
ally from the mouth down, being widest at this point. The lid is
more erect and has a narrower and stouter neck, and is backwardly
recurved at the edges. The apex, instead of being emarginate as in
some species, is acute (7 — 7). A leaf 24 cm. long, examined at Munich,
shows zonation as follows.

Zone I is the under surface of the lid and carries many short,
stout, downwardly pointed hairs and many nectar glands. The lower
limit of this zone is very irregular, the hairiness following the promi-
nent veins, the spaces between being smooth and glaucous and con-
tinuous with zone 2, which is lined with an epidermis of imbricated
pointed cells with many glands scattered over the surface. This zone
extends about 2 cm. downwards, and includes most of the neck, the
nectar roll and a narrow zone below it, as in 5. minor. This is deep
yellow, glaucous and the imbricated cells of the epidermis are short
pointed. There are glands present in great numbers, and, quite as in
other species described, this is a dominant place of lure. The lower
limit of this zone is not defined but fades into zone 3 (8 cm. deep),
in which the imbricated cells have longer retrorse points. The num-
ber of glands is reduced so that in the lower regions there are none
to be found. The whole area is glaucous. The lower limit of zone 3

Francis E. Lloyd — 30 — Carnivorous Plants

is very irregular, but readily distinguished by the eye by the change
in color, due to absence of cuticle in zone 4. Under the microscope
there is a sudden transition from imbricated apiculate hairs to scat-
tered, very long, curved ones, characteristic of zone 4. Underlying
both zones there is a hypodermis of wavy, thick walled cells. The
lowest portion of zone 4 is quite smooth, is lined with a small, straight
walled epidermis, and is 6 cm. in depth.

Sarracenia Jonesii. — Living material from Flat Rock, North
Carolina, collected by Dr. L. E. Anderson through the courtesy of
Professor F. A. Wolf, was examined and showed quite the same char-
acters of the epidermis and of zonation as have been described for S.
Drummondii and S. flava. Such differences as occur are those of the
shape of the hd, which is smaller and ovate, similar to that of S.
minor save that it is more erect and apiculate. The color is green,
veined with red, with no fenestrations. There are no glands in the
lower part of zone 3, or in zone 4.

Morphology of the leaf. — Troll (1932) has summarized our knowl-
edge on this subject, adding his own views.

Baillon (1870) compared the leaf of Sarracenia with that of
Nelumbo, expressing the opinion that "the wide but shallow cone
of the (peltate) leaf of Nelumbo becomes in Sarracenia deeper and
narrower in such a manner as to produce definitively the form of a
long, obconical trumpet," thus recognizing the relation between the
epiascidiate pitcher of Sarracenia with peltate leaves.

As we have already seen, the pitcher consists of a spreading bi-
facial leaf base surmounted by a tubular, gradually widening pitcher
bearing a strong ventral wing and the foliaceous flap continued in
front as the nectar roll. When the ascidium fails of development, as
it sometimes does (as in S. flava, etc.), the leaf presents a certain
likeness to that of Iris and the phyllodia of Acacia sp. Asa Gray
(1895-7) designated the ventral wing or keel as a "phyllodial wing."
Various earlier authors (Lindley 1832, Saint-Hilaire 1840, Morren
1838, Duchartre 1867) regarded the pitcher as a leaf blade with the
margins fused, some of them thinking the tube to be the winged
petiolar region and the flap the leaf blade. Gray accepted this view,
saying, "they are evidently phyllodia." Of them, however, Morren
believed the flap to be only the apical portion of the leaf blade, the
most of which is involved in the tube. Macfarlane (1889-90) was
firmly of this opinion. He regarded the keel as compound of the
fused leaf edges, comparing the condition with that in the leaf of
Iris as he interpreted the anatomy of this. The single keel of Sar-
racenia is equivalent to the pair of apposed wings of Heliamphora
and the more widely separated ones in Nepenthes, according to the
nature of the fusion. The flap of Sarracenia is to be regarded as
compound of two pinnae, as is the lid of Nepenthes (1893). Troll
regards this view as false, based on a misconception of the mor-
phology of the Iris leaf, which he insists is congenitally a strictly uni-
facial leaf. He accepts Goebel's intrepretation that the Sarracenia
pitcher is a unifacial leaf in the form of a tube and turns to the de-
velopment of it for support. He recalls that in plants with sword-
shaped leaves the leaf blade (Oberblatt) arises as an outgrowth of the

Chapter II — 31 — Sarracenia

back mass (abaxial side) of the leaf base (Goebel i88i), which
then enlarges its own apex behind and above the "primary leaf apex,"
that of the leaf base proper.

The development of the embryo leaf follows the same course except
that the upper side of the primordium of the lamina is not completely
suppressed but is limited to a minute depression between the leaf
base and the apex of the leaf blade (Oberblatt). This is limited
below by an as yet extremely narrow transverse zone, which cor-
responds to the transverse zone of a peltate leaf, and at the same
time may be taken as an indication of a unifacial petiole primordium,
one, however, which experiences no further development. It is im-
portant to note that the leaf blade apex does not rise much above
that of the leaf base and that therefore the leaf blade stands nearly
normal to the base. This, however, soon changes. The apex now
grows rapidly and the primordium widens, flattening dorsiventrally,
and changes in curvatures ensue to produce the helmet-shaped form
of the lid. Meanwhile the adaxial side elongates and the keel appears,
below which the transverse weal of the leaf-base runs; otherwise said,
the edges of the leaf base are not concurrent with the keel. The
similarity to the Iris leaf is unmistakable, only allowing that the leaf
blade is hollow. Thus the unifacial character of the leaf blade comes
into expression, not as a shallow saucer but as a narrow furrow. The
upper side of the lamina primordium is at first confined to the adaxial
side of the leaf blade. Growth progresses less by spreading than by
upward growth of the margins to form the tubular leaf. The petiole
is not developed, but instead the blade greatly lengthens, in S. flava
to the extent of i meter. The terminal portion, whose edges are
free, becomes the lid.

I have examined the development of the leaf of 5. minor and can
generally agree with Troll. In this species, however, the primordia
stand out perhaps more clearly because of the greater tendency to
grow in length, as compared with S. purpurea. In this, however, all
leaves do not act ahke, for there is sometimes a greater lengthening
of the petiolar region, so that even in this species the petiole is not
absent, though normally much reduced (2 — 12).

In S. minor the earliest stage of development available was a leaf
only 0.1 mm. in height in the form of a low cone, its base reniform,
about 0.4 mm. in greatest diameter, its convex edge adaxial (3 — 13).
The extremities of this edge had a talus-like slope, the cone melting
into the overhanging front of the body of the leaf. This rose to a
very low apex, evidently a growing point, scarcely evident as a
distinct boss. There could be seen a very shallow groove from the
top down the adaxial face of the overhanging mass. In this very
early and undifferentiated condition there is recognizable a leaf base
with very thick edges, a groove, a sign of the coming invagination,
and a very low boss. It is not, however, possible to locahze a definite
growing point for the leaf base, while even at this young stage, the
apex of growth of the leaf blade is just visible.

In a leaf 0.3 mm. (j — 14, 15) in height the apex stands out well,
and below it adaxially one now sees a short groove, the leaf base
margins distinctly passing transversely across, some distance below

Francis E. Lloyd — 32 — Carnivorous Plants

the groove. But there is no conclusive evidence that the only apex
ever seen, the leaf-apex, is secondary, as in the case of Iris, though I
hesitate to deny the parallel accepted by Troll.

In a leaf 0.8 mm. long (j — 16) the initial groove is surrounded
by a Hp which is evidently the rim of the pitcher, surrounding the
mouth. With further growth this is raised upward on a laterally
compressed stalk on the adaxial side of which the wing now appears
as a solid longitudinal outgrowth. In these later stages it is quite
evident that the margins of the leaf base meet transversely, and that
the wing arises quite independently of it. The zone between the lower
end of the wing and the upper end of the leaf base must, I think, be
regarded as petiolar. It is solid, and has unifacial structure. In the
mature leaf one can see that the wing is, near the mouth, slightly
doubled (j — 17, 18).

Digestion and absorption. — It was once thought that the tubular
leaves were a device for holding water. Collinson wrote to Linnaeus,

"the leaves are open tubes, contrived to collect rains and dews,

to nourish the plant in dry weather." This prompted Linnaeus
to regard Sarracenia leaves as derived morphologically from those
of Nymphaea, but adapted to holding water for its needs, thus enabHng
it to occupy drier situations, incidentally providing water for thirsty
birds. But as Goebel, from whom we have drawn these notes, re-
marks, Sarracenia lives in swamps, a fact with which William Bar-
tram was familiar, but who yet thought that the water caught by the
hollow leaves was for the " refreshment " of the plant.

Goebel, however, showed that it is easy enough to demonstrate
that the pitchers can and do absorb a not inconsiderable amount of
water: 6.8 cc. out of 20 cc, and 2 cc. out of 10 cc. in two cases. Fibrin
which he introduced remained unaffected, and meat extract, neutral-
ized with sodium carbonate, was attacked by bacteria. These and
other similar experiments led Goebel to the conviction that while
absorption can take place, there is no digestion beyond that attribut-
able to bacteria, and there is no antiseptic action. Previous to
Goebel several authors had expressed suggestions, opinions, even
convictions about the matter. Sometimes the remarks made did
little more than show that a question had arisen in the mind, as in
the case of Macbride (1817). Hooker (1875) merely recognized a
possibility that digestion occurs. Mellichamp (1875) was the first
to do some experiments which, though crude, led him to conclude that
the fluid of the pitchers hastens the decomposition of insects, without
at all evaluating the role of bacteria. Batalin (1880) interpreted the
exfoliation of the cuticle in the deeper zone of the pitcher as evidence
that in the absence of glands, which he incorrectly stated to be absent
from Sarracenia, digestive stuffs (Losungsmittel) for the solution of
proteins were released. No experiments to prove the occurrence of
digestion were done. Schimper (1882) showed that changes, due to
the absorption of food materials, occur in the epidermal cells similar
to the changes called aggregation by Darwin, but he could not find,
from experiments, that there is evidence of digestion aside from bac-
terial action. That nitrogenous compounds are absorbed was shown
by HiGLEY (1885). Zipperer (1885), concerned chiefly with anatomy,

Chapter II — 33 — Sarracenia

did only two superficial experiments, one proving to his mind that
diastase, and a second that a pepsin is present.

Following GoEBEL, Lambert (1902) showed absorption to take
place in certain regions, by following the entrance of methylene blue
or fuchsin. That digestion occurs was no more than a conviction
without proof. Fenner (1904) like Schimper observed that absorp-
tion takes place and is followed by cytological changes in the epider-
mal cells. He observed quantitatively considerable amounts of fluids
absorbed. As to digestion he expressed the opinion that Sarracenia
flava is an insectivorous plant with a digestive enzyme. Robinson
(1908) found evidence that sucrose and starch can be digested to,
presumably, simple sugars by the pitcher fluid, confirming Zipperer
as to starch, but none that there is either fat or protein digestion.

Thus stood the evidence when in 1918 Hepburn, St. John and
Jones started their exhaustive studies on these physiological questions
presented by Sarracenia (and Darlinglonia, mentioned elsewhere).
These authors did an immense lot of work and travelled extensively
for the purpose of doing field experiments. To their results we now
turn.

For the purpose of learning about the presence or absence of
digestion, they examined 5. flava, Drummondii, Sledgei, rubra, minor,
psittacina and purpurea. Tests were made on the fluid from both
unopened and opened pitchers, with and without the addition of weak
acid (HCl) and alkali (sodium carbonate), and always in the presence
of trikresol as a bactericide. Carmine fibrin was used as a substrate
in the field, and this, edestin, casein and coagulated egg albumin in
the laboratory. The evidence was strengthened by duplication, trip-
lication, or even quadruplication of the tests. Generally composite
samples of fluid drawn from a number of pitchers were used. All
experiments were done quantitatively, even to measuring the amount
of substrate, a matter of importance often disregarded.

In S. flava a protease was shown to act on fibrin in both closed
(that is, still unopened) and open pitchers. It is more active in weak
acid (0.2%), than in weak alkali, but there was no action in their
absence. Edestin was digested in 1.5 to 2 hours. The fluid from closed
pitchers was vigorous, acting completely in 30 min. at 37.5° C. and
almost so at room temperature. Casein was partially digested in 2
hours. With coagulated egg-white negative results only were obtained.

S. Drummondii and S. Sledgei. — When the fluid of closed pitchers
was acidified and then tested the results were purely negative. With
that of open pitchers, six experiments of eight were negative, but
partial or complete digestion occurred in two on sustained exposure
(49-57 days). When neither acid or alkali were added the fluid of
closed pitchers failed to act even after 50-55 days in four experiments;
in two others digestion occurred in 7 and 21 days respectively for S.
Drummondii. For 5. Sledgei one experiment was negative, while
others acted slowly but completely in 37 to 49 days. The results
stand in marked contrast with those in which the fluid was modified
by alkali, showing that digestion occurred in 1.5 to 18 days in 20 ex-
periments, while in two others it required 32-36 days, the slowness
being due to a reduced concentration of sodium carbonate, which

Francis E. Lloyd — 34 — Carnivorous Plants

was probably neutralized by the acid proper to the fluid which had
been shown to be present. There was no marked difference between
the fluid of open and of closed pitchers. It was concluded that a
protease was present which, however, acted best in an alkaline medium.

Edestin was slightly attacked by the fluid of open pitchers. Casein
was digested completely in 2-8 hours at 37.5° C, and coagulated
egg-white, in the absence of acid or alkali, was not attacked, nor was
it in the presence of acid, but, on the addition of alkah, "incipient
digestion was noted in 24 hours, marked digestion in 48 and 144
hours and advanced digestion in 168 and 216 hours respectively" for
two experiments.

S. rubra. — Only a single sample of fluid was available, but this
on division showed that in the presence of alkali digestion proceeded
rapidly, o.oi gram of fibrin being completely digested in 1.3 cc. of
fluid in 2 hours. In the part with acid added "partial solution was
noted at the end of 9 days, complete solution at the end of 50 days."

S. minor. — Experiments were done in the field. No digestion oc-
curred in 30 days (trikresol present) in the absence of acid or alkali,
but it did occur when either acid or alkali was added to the fluid,
but less vigorously in the alkaline, than in the acid medium. This
species therefore stands in contrast to the others aforementioned.

S. psittacina. — Field observation showed that fluid was being se-
creted in the pitchers, but in such small amounts that it had to be
collected by dilution with water (0.5 cc. in each of 50 pitchers free
of insects). The results were inconclusive, but indicated that a pro-
tease is present which is active in the presence of acid.

Sarracenia purpurea. — The secretion in this species is very small
in amount, being found as beads of moisture on the walls. Experi-
ments were done by flushing the pitchers, emptying, and adding 10
to 15 cc. of water to each pitcher. After some days this was removed
and tested. The fluid thus obtained showed an ability when alka-
line to digest fibrin to a marked extent in 8, 24, 72, and 87 hours
(4 samples), and completely in 42, 48, 120, and 135 hours, respectively.
Three other experiments gave positive results in 42 and 87 hours.
In acid, the results were equivocal.

The general conclusions reached by Hepburn and his colleagues
are that a protease is present in all the Sarraceniae, but that in most
cases it acts best in an alkaline medium. In some cases, however, it
acts in an acid medium.

It was shown experimentally that pitcher fluid retains its power
of digestion after being kept at room temperature for as long as 370
days, either with or without a bactericide; and that this is also re-
tained on dilution, which is of importance in view of the fact that
dilution by rain may always be expected in the habitat. In contrast
with Nepenthes, neither mechanical nor food stimulation was found
to have any effect.

In the fluid of closed pitchers there is evidence that there are both
invertase and lipase, while maltase, emulsin, diastase, urease, and
esterase were present. It will be recalled that Zipperer claimed the
digestion of starch and Robinson that of sucrose and starch.

It is generafly understood that the water of swamps and sandy

Chapter II — 35 — Sarracenia

soils is acid, though not always. Wherry found bog water to be
sometimes alkahne, though the sphagnum hummocks would contain
acid water. Generally he found that when growing in acid water,
the pitcher fluid was acid, but not always in the same degree. But
occasionally the fluid would be alkaline or neutral. On the same
plant the fluid of a young pitcher might be alkaline, and of old pitchers
acid — this for Sarracenia purpurea. Hepburn et al. found for some
southern species acidity and alkalinity about equally distributed for S.
Drummondii and S. Sledgei. Five pitchers of S. flava all held acid fluid.

Hepburn and St. John examined the bacterial content of closed
and open pitchers. Closed pitchers were always sterile. The fluid
of open pitchers which had captured prey always contained bacteria,
as is to be expected, and these always digested several different sub-
stances. The interesting fact is pointed out, however, that these
"bacteria digested the proteins so slowly that their part in the di-
gestion of prey must be a minor one in the genus Sarracenia, the
protease of the pitcher liquor playing the leading role." "The bac-
teria apparently live in symbiosis with the Sarracenias, drawing their
nutriment from the digested insects, and aiding, to a certain extent,
in the digestion of the prey." An exception must be made for Dar-
lingtonia where there is no digestion by the pitcher fluid proper (see
beyond). Reference is made to the odors of putridity and of am-
monia and amines noticed when such bacteria are active.

The earHer observations of others that water is absorbed was
verified, and it was further found that, when nitrogenous substances
in solution were used, the absorption of the solutes proceeded more
rapidly than that of the water. In the presence of a phosphate buffer,
the nitrogenous compound would be absorbed while the total amount
of water increased. When a neutral phosphate solution was used,
the absorption of the phosphate was less rapid than that of the water.
The percentage of compound absorbed usually increased with the
length of the period of absorption. The actual entrance of substances
into the tissues was demonstrated by the presence of the lithium ion,
after introducing lithium citrate into the pitcher fluid. It is thus
indicated that the products of proteolysis and phosphates are ab-
sorbed by the walls of the pitcher and utilized by the plant. Data
on the chemical composition of the tissues are also furnished, but
are of only secondary interest here.

This brief summary of the work of Hepburn and his colleagues,
it is gratifying to record, has furnished a comprehensive view of the
physiology of the pitchers of Sarracenia in its relation to digestion,
which might have remained unwritten but for their evident enthusiasm
and dihgence.

Animals which live in the pitchers of Sarracenia and Darlingtonia. —
It is a matter of common observation that the pitcher plants attract
a horde of insects of all kinds: "ants, wasps, bees, butterflies and
moths" by their nectar, and other forms (beetles, spiders) for other
reasons. Spiders frequent the opening of the pitcher in the not vain
hope of visits of small insects which may be caught by them. Ed-
wards observed this in Darlingtonia. They occur on Drosera, Byblis,
Nepenthes and probably others for the same reasons. Minute wasp-

Francis E. Lloyd — 36 — Carnivorous Plants

like creatures parasitize pitcher inhabitants. The caterpillars of a
moth live in burrows formed by feeding on the tissues of the rhizome,
and form characteristic, more or less upright, above-ground tubes
of the debris. This is Papaipema appassionata, the moth being maroon
and yellow in color. Of especial interest here are those which habitu-
ally use the pitchers as their homes, and live in them and nowhere
else. They fijid their food either in the tissues of the pitcher walls,
or in the mass of dead insects caught by the trap, or merely live in
the water from which, however, they must derive their food. Animals
other than insects include a small tree-toad and a small chameleon
lizard, whose bones are sometimes found in the inclosed debris. The
toad rests just inside the mouth of the pitcher, doubtless awaiting
the chance of capturing prey. But to come to the obligate inhabit-
ants :

A small mosquito {Wyeomyia Smithii) lays its eggs in the pitchers
of Sarracenia purpurea. In the pitcher fluid (always diluted by
rains) the larvae grow to maturity, hibernating frozen in ice during
the winter. It is harmless to man. It is said not to breed elsewhere,
and is found well beyond the Canadian border, though it is tropical
in its affinities.

Similarly the larvae of a minute gnat, Metriocnemus Knabi, breeds
in the same manner. A closely related species, M. Edwardsii, de-
scribed by Jones (191 6), was discovered to occur in the pitchers of
Darlingtonia californica by Mrs. Austin about the year 1875. The
larvae are minute thread-Uke "worms" circulating in the decaying
insect debris, which appears as writhing masses, so numerous are the
larvae. According to Hepburn and Jones (1919), such forms pre-
serve themselves against the digestive action of the surrounding fluid
by means of antienzymes. In this connection it may be mentioned
that, in discussing the anthelmintic action of papain (in the crystal-
line form), Berger and Asenjo (1940) tested the effect of this enzyme
on Ascaris lumhricoides (from pig intestine) and found that these
organisms were attacked, that is, ulcerations were formed on them
inside of 16 hours in a 0.02% solution in a phosphate buffer of pH 5,
and the animals were completely digested in 16 hours in a 0.11% solu-
tion. The same authors showed that the bromelin of fresh pineapple
juice acts similarly. Evidently Ascaris has no sufficient protection
against this enzyme in the concentrations used, though presumably it
has an antienzyme which protects it within its normal environment.

The most intriguing of all the animal associates are three closely
related species of a small moth of the genus Exyra, because of the
striking adaptations which they display to a special environment
(Jones 192 i). These moths lay their eggs singly or in groups in the
mouth of the pitcher. When laid singly, the hatched larva enters
the tube and feeds on the superficial tissues of its wall. This is true of
the species Ridingsii and semicrocea. If more than one larva happens
to occupy a pitcher, one of them ruthlessly drives out or kills its
neighbors. The third species, Rolandiana, lays a group of eggs in a
pitcher of a single plant (of S. purpurea), and when hatched the
larvae spread to various closely placed pitchers, readily possible in
this species because of the dense massing of the pitchers (z — 3).

Chapter II — 37 — Sarracenia

Eventually only a single larva occupies a pitcher. The wide separa-
tion of pitchers of 5. jlava, rubra, Drummondii, Sledgei and minor are
a practical hindrance to such movements of very young larvae from
one pitcher to the other, and it is in these that the other species of
£^>'m lay their eggs singly. "Thus," says F. M. Jones, "the habit
of growth of the food plant determines the egg-laying habit of the
associated insect" (1921).

The newly hatched larva is very small (2.6 mm.), and being trans-
lucent and half buried in the tissues on which it feeds and partly
covered by debris enclosed within the tube, seems pretty well pro-
tected without further ado. E. Ridingsii on hatching retires to the
grooves in the hd-stalk of S. flava, and there forms for itself a small
tent of silk and frass, on the floor of which it continues feeding. The
older larvae of all three species make use of a method of isolating
themselves from the outside world as follows. They spin a diaphragm
of silk webbing across the mouth of the tube, either transversely or
more or less obliquely according to the position of the hd, and in
5*. psittacina, across the mouth of the entrance tube. Any accidental
openings are closed by webbing, and thus they immure themselves
in a food chamber from which rain is prevented entrance. Larvae
of a spring brood, when they find themselves in young tender pitchers,
use another quite extraordinary method of insuring for themselves a
safe retreat. The young larva then eats away a ringing groove near
the top of the pitcher. Above this the pitcher wall dies, dries, and
becomes indurated, sagging over and barring the entrance. In the
chamber thus formed the larva feeds and hibernates. In the pitchers
of S. flava, which die down during the winter, the larva retires to the
lower regions of the pitcher, and there ensconces itself in a chamber
plugged by webbing and frass, where it awaits the spring. A curious
variant of this habit is displayed by the caterpillar of Exyra Ridingsii,
which before pupation prepares for the future by cutting an emergence
hole above its point of pupation, so that the moth may easily escape,
and below a small hole for the drainage of water, so that its pupation
chamber may not be flooded. It then forms its chamber by webbing
spun loosely so as to allow water to pass, and then spins its cocoon
of webbing and frass. Exyra semicrocea, when it pupates in the pitchers
of S. psittacina, handles its situation somewhat similarly, but with
special attention to the peculiarities of the host plant. Usually when
the larva intends to pupate it passes into an uninjured pitcher. Since
that of 5". psittacina has a lobster-trap entrance, out of which escape
would be difficult — not because of the size of the opening, but be-
cause of its re-entrant character — the larva first cuts an escape hole
in the roof region of the hood.

After hibernation the larvae (the third instar), are voracious, and,
emerging in the spring, attack not only the pitcher, but the flowers
and young fruit which they devour. When ready to pupate the larva
cuts a hole in a young growing leaf still unopen, ascends the tube,
and feeds on the inner tissues. This causes the tops of the pitchers
to wither and the dead portion to topple over. E. Rolandiana does
the same in S. purpurea. The larvae of these moths have lateral
tubercles or "lappets" which, according to Jones, prevent them from

Francis E. Lloyd — 38 — Carnivorous Plants

entering and so getting pinched in too narrow spaces. The species
{E. Rolandiana) pecuhar to S. purpurea, with its wide, amply spacious
pitcher, does not possess the lappets. These in the very young larvae
are scarcely more than bristles, but with successive instars the tubercle
becomes larger, and armed with a prominent bristle.

A solitary wasp, Chlorion Harrisi, habitually makes use of Sar-
racenia pitchers for its nest of several stories which are supplied with
food and each an egg. Dr. Jones informs me, however, that this
insect is not confined to Sarracenia. He found it in 1939 nesting in
abandoned beetle-burrows on Martha's Vineyard, Vineyard Island,
where Sarracenia does not occur.

A fly, Sarcophaga, produces large white maggots which feed upon
the remains of insects in the pitchers. The protective enzyme studied
by Hepburn and Jones was extracted from Sarcophaga larvae. Sev-
eral species of this genus, peculiar to Sarracenia, were described by
Riley.

There is a minute fly (2-3 mm.), Dorniphora venusta, which is found
in the pitchers of 6*. flava late in the season when they are relatively
dry and have lost their trapping abilities, as they apparently do
(Jones, 191 8). The larva feeds upon the captured insects.

Another small fly, 3-3.8 mm. in length, described under the name
Neosciara Macfarlanei by F. M. Jones (1920) has similar habits, and
is found in the vertical tubed Sarracenias. Its presence is betrayed
by a frothy looking product of the larvae about to pupate, which
fills the pitcher tube just above the mass of dead insects on which
they have fed. Both these flies, as well as the Sarcophagids above
mentioned, appear to be confined to Sarracenia, but some doubt
remains as to this.

The purpose of the above brief account is merely to point out
the more general facts about the constant insect associates of Sar-
racenia and Darlingtonia. To exhaust the present knowledge of all
the insects which attack and feed upon and pollinate these plants
would go beyond our purpose. A general summary of this knowledge
is supplied by Jones in Walcott's book of illustrations of the Sar-
racenias, in which a bibliography is to be found. It is upon this
author that I have depended for these notes. It may be added as a
matter of speculation that further investigation would certainly dis-
cover many other associates, including Crustacea, protozoa and pro-
tophyta, some of which might turn out to be obligate inhabitants,
as in the case of Nepenthes.

Literature Cited:

Arber {see under Cephalotus).

Baillon, H., Sur le developpement des feuilles des Sarracenia. C. r. Acad. Sci. Paris
71:630, 1870. Also in Adansonia 9:331, 1868-1870 {through Troll).

Bartram, Wm., Travels in N. and S. Carolina, Georgia and Florida. Philadelphia 1791.

Batalin, a., tJber die Function der Epidermis in den Schlauchen von Sarracenia und
Darlingtonia. Acta Hort. Petropol. 7:345-359, 1880.

Berger, J. & C. F. Asenjo, Anthelmintic activity of crystalline papain. Science II, 91:
387-388, 1940.

Burnett, G. T., On the functions and structure of plants, with reference to the adumbra-
tions of a stomach in vegetals. Quart. Journ. Sci. Lit. and Art, Vol. for Jy.-Dec,
1829:279-292.

Chapter II — 39 — Sarracenia

Canby, W. M., Darlingtonia calif ornica, an insectivorous plant. Proc. A. A. A. S. 1875,

6:64-72.
Catesby, Nat. Hist, of Car., Vol. 2, p. 69, 1743 {through Hooker).
CoLLiNsoN, (see Smith 1765).
Darwin, (see under Drosera).

DucHARTRE, P. E., Elemcns de botanique. Paris 1867 (through Troll).
Fenner (see under Nepenthes). .

GoEBEL, K., Blattentwickelung von Iris (under the heading of "Litteratur"). Bot. Zeitung

39:95-101, 1881.
GoEBEL, K. (see also under Nepenthes).

Gray, Asa, Sarraceniaceae. Synoptical Flora of North America 1:79, 1895-7, New York.
Hegner, R. W., The protozoa of the pitcher plant (Sarracenia purpurea). Biol. Bull. 50:

271-276, 1926. . J II- J

Hepburn, J. S., F. M. Jones, & Eliz. Q. St. John, The absorption of nutrients and allied

phenomena in the pitchers of the 5(Z?-raceH/aceae. Journ. Franklin Inst. 189:147-184, 1920.
Hepburn, Jones & St. John, The biochemistry of the American pitcher plants (Secondary

title: Biochemical Studies of the North American Sarraceniaceae). Trans. Wagner

Free Inst, of Science 11:1-95, 1927- Full bibhography.
Higley, Bull. Chic. Acad. Sci. 1:41-55, 1885 (through Hepburn).
Hooker, J. D., Address to the Department of Botany and Zoology. Rep. 44th Meeting

Brit. As. Adv. Sci. Belfast 1874:102-116, 1875.
Jones, F. M., Pitcher-plant insects. Ent. News 15:14-17, 1893.
Jones, F. M., Pitcher-plant insects, II. Ibid. 18:413-420, 1907.
Jones, F. M., Pitcher-plant insects. III. Ibid. 19:150-156, 1908.
Jones, F. M., Two insect associates of the California pitcher-plant, Darlingtonia califor-

nica. Ibid. 27:385-392, 191 6.
Jones, F. M., Domiphora venusta Coq. in Sarracenia flava. Ibid. 29:299-302, igis.
Jones, F. M., Another pitcher-plant insect. Ibid. 31:91-94, 1920.
Jones, F. M., Pitcher plants and their moths. Nat. Hist. 21:296-316, 1921.
Jones, F. M., Pitcher plants and their insect associates. In Walcott, 1935, pp. 25-34-
Krafft (see under Heliamphora).

Lambert, Ann. de Hygiene et de Med. col. Paris 1902, 5:652-662 (through Hepburn).
Lindley, J., Introduction to Botany. London, 1832.
Macbride, James, Trans. Linn. Soc. London 12:48-52, 1817.
MacDougal, D. T., Symbiotic saprophytism. Ann. Bot. 13:1-47, 1899.
MacDougal, D. T., The influence of light and darkness upon growth and development.

Mem. N. Y. Bot. Gard. 2, 319 pp.. New York 1903.
Macfarlane, J. M., & D. W. Steckbeck, Sarracenia purpurea var. stolonifera, A note-
worthy morphological and ecological type. Bull. Misc. Inform. Kew, No. 4, 1933:161-169.
Macfarlane (see also under Nepenthes).
Mellichamp, J. H., Notes on Sarracenia variolaris. Proc. .\m. Ass. Adv. Sci. 23 meeting.

MoRREN,' Ch., Morphologic des ascidies. Bull. Acad. R. Belg. Bruxelles 5:430, 1838

(through Troll). ■ , / j

Riley, C. V., On the insects more particularly associated with Sarracema varwlus (spotted

trumpet-leaf). Proc. A. A. A. S. 1875, 6:18-25.
Robinson, Winifred J., A study of the digestive power of Sarracema purpurea, lorreya

8:181-194, 1908.
Saint-Hilaire, a. DE, Legons de botanique etc. Paris 1840 (n.v.).
ScHiMPER, A. F. W., Notizen uber insectenfressenden Pflanzen. Bot. Zeitung 40:225-234;

241-248, 1882.
Smith, Correspondence of Linneaus. Vol. i, p. 69, 1821. Reference to Collinson

(through Hooker, 1875).
Troll, W., Morphologic der schildformigen Blatter. Planta 17:153-314, i932-
Uphof, J. C. Th., Sarraceniaceae. Die naturlichen Pflanzenfamilien, 2. Aufl., Vol. 17b:

1-24, 1936. . . .

Vogl, a., Phytohistologische Beitrage, II. Die Blatter der Sarracenia purpurea l>mn.

Sitzungsber. Wien. Akad. Wiss. Math.-Wiss. Kl. 50:281-301, 1864.
Walcott, Mary V., Illustrations of North American pitcher plants. Smith. Inst. Wash.

1935 (containing contributions by Wherry and by Jones).
Wherry, E. T., Acidity relations of the Sarracenias. Journ. Wash. Acad. Sci. 19:379-39°, 1929-
Wherry, E. T., The geographic relations of Sarracenia purpurea. Bartoma No. 15,

Wherry, E. T., Exploring for plants in the southeastern states. Sci. Mo. 38:80-85, 1934-
Wherry, E. T., Distribution of the North American pitcher plants. Walcott, 1935, Pp. 1-23.
Wilson, W. P., On the relation of Sarracenia purpurea to S. variolaris. Proc. Acad. Nat.

Sci. Phila. 1888:10-11. ivt • u

Zipperer, Paul, Beitrag zur Kenntnis der Sarraceniaceen. Diss. Univ. Erlangen. Munich

1885, Pp. 34-

Chapter III
DARLINGTONIA CALIFORNICA

Discovery. — Distribution. — Habit. — Leaves : kinds. — Structure. — Place of absorp-
tion. — Development of leaf. — Digestion and absorption.

The genus name Darlingtonia is used here because of its wide
familiarity and use in horticultural literature. Under the International
Rules of Botanical Nomenclature this name is invalid because of being
a later homonym, and is to be replaced by Chrysamphora Greene.

This highly localized pitcher plant of Oregon and California called
locally the ''cobra plant'' was discovered in October 1841 by Mr. J. D.
Brackenridge, Assistant Botanist of the U. S. Exploring Expedition,
under Captain Wilkes, on a journey from Oregon to San Francisco.
It was found in a marsh bordering a small tributary of the Upper
Sacramento River a few miles south of Mt. Shasta. In the opinion of
John Torrey, who described it in 1853, it was sufficiently different from
Sarracenia to warrant the new generic name which he gave it, dedi-
cating it to his "esteemed friend" Dr. Willl^.m Darlington, of West
Chester, Pa., "whose valuable works have contributed so largely to the
scientific reputation of our country." The range of this species is now
known to extend into the Siskiyou Mountains of S. Oregon, down to
sea-level along the coast (I found it 6 miles north of the town of
Florence) and in the contiguous region of California. As an example
of a restricted geographical distribution, this is comparable to that of
Cephalotus follicularis in S. W. Australia.

Darlingtonia has the same general habit of growth as that of the
other Sarraceniaceae, a strong perennial rootstock, bearing a sort of
rosette of leaves and clothed with the dead remains of older leaves.
The larger leaves attain a length of 2-3 feet ("3 ft. 6 in.", Edwards)
and present a unique appearance, owing to the torsion of their tubes
and the large motley domes with their fishtail-shaped appendages.
"The leaves are most beautiful and singular, having a fanciful re-
semblance to a number of hooded yellow snakes with heads erect
in the act of making a final spring, suggesting the name 'caput ser-
pentis'," wrote Edwards in 1876. He states that the leaves all twist
in the same direction, which is not the case (Kurtz) {4 — 1-5).

There are two kinds of leaves, juvenile, produced by seedlings
and by small shoots, and the leaves of maturity. The juvenile leaves
(5 — 1-4, 6; 6 — 17), which have been described by Goebel, follow
directly on the very simple lanceolate cotyledons, and on small lateral
shoots of restricted growth on rhizomes. They are tubular, tapering
downwardly, with a clasping base. The opening is oblique, the leaf
being drawn out on the abaxial side into a tapering acute or bifid
apex. The edges of the opening are simple, that is, are not curved
in- or outwardly. On the adaxial aspect the opening is bayed or
sometimes slit downwards. The whole outer surface of the leaf is
studded with somewhat raised stomata and many nectar glands which

Chapter III — 41 — Darlingtonia

scarcely exceed the stomata in size. These glands are found even
on the outer surface of the overhanging apex. The inner surface can
be roughly divided into three zones. The uppermost embraces the
whole of the apex and some distance into the interior from the open-
ing. The inside surface of this portion is furrowed longitudinally from
the extreme apex to a point well within the tube. The floor of this
furrow is lined with smooth epidermis which for some distance forms
a low swelling on each side of the furrow. Among these cells are a
very few groups which have distinctly the structure of a nectar gland,
but I have not been able to determine positively that nectar is se-
creted. They seem not to be quite so highly speciaHzed in form at
least as the glands of a mature or adult type of pitcher. The epider-
mis in general of this zone is of the "fishscale" type, that is the cells
are imbricated and are downwardly sharply pointed, the more sharply
the more deeply placed in the pitcher. As zone 2 is reached, fewer
of the cells are trichomes, which are now much longer, the remaining
cells being quadratic and elongated to some extent. In the depths
of the pitcher, hairiness ceases, and the epidermis, in zone 3, is quite
smooth. Like the senile leaves, the juvenile tube leaf is twisted
through 180 degrees from base to apex, so that the opening comes
to face downwards, more or less. In size they may be as small as i
cm. and up to 10 cm.

Occasional juvenile leaves display aberrancies from the normal
course of development. Rather frequently one finds a leaf with the
apex forked, and having no median vein, clearly corresponding to
the fishtail appendage of the adult leaf. Accompanying this con-
dition there may or may not be developed an ala ventralis. In most
cases the rim of the mouth remains simple, but in one leaf I found
that a distinct nectar roll had been developed along both sides, but
not meeting anteriorly, as in the adult leaf, to fuse {5 — 3)- This
indicates pretty clearly that the place where the two sides of the
nectar roll meet is a site of concrescence in the fully elaborated leaf.
As in all these forms of juvenile leaves there are nectar glands, which
are confined to a broad band along the inside of the hood and apex
but not elsewhere on the inner surface ((5— 18). The epidermis is
all of tesselated umbonate cells above, becoming longer pointed further
down. The glands are not quite so elaborate here as in the adult
leaf. Those on the outer surface are typical in appearance.

A single juvenile leaf was found in which there was a closure of
the mouth for only a short distance above the base. What might
have been the tube was laid quite open, and formed no trap at all.
Glands were present on the apical appendage and along a midband,
as usual. The epidermis was tesselated. That is, the leaf was a nor-
mal juvenile leaf in all respects except that the edges remained free

(5-6).

Cases of this kind might be used as evidence that the pitcher
arises by fusion of the leaf margins (Macfarlane) but can as well
be explained as resulting from disharmony of growth. On various
grounds another explanation is to be preferred. (See beyond).

It may be noted in the juvenile leaves that the margins of the
"total stipule" (Troll) run far up the petiolar region. In a juvenile

Francis E. Lloyd — 42 — Carnivorous Plants

leaf about 30 mm. long, the ends of the stipular margins were encoun-
tered about half the way up (ca. 15 mm.). This very gradual running
out of the stipular margins conveys the impression that the edge of
the wing is doubled throughout its length, and inasmuch as in the
adult leaf the edge of the wing in its upper reaches is, as a matter of
fact, also double, this doubhng seems continuous with that of the
stipular wings. We shall, however, see elsewhere that the one had
nothing to do with the other. In the juvenile leaf, however, the wing
is single above so that the end of the stipulation is clear. This in-
volves the study of transverse sections (5 — 4).

At the same time it can be seen that the outgrowth to form the
wing had already started its growth beneath the stipular margins.
A study of the development of the leaf shows why this takes place.

The adult leaves have been described a number of times, by Tor-
REY, Hooker, Macfarlane, Kurtz, Goebel and others, but despite
this, the precise morphological relations of the parts about the mouth
of the pitchers remain only vaguely comprehended.

Adult leaves are produced both on short shoots, when they may
be quite small (1.5 cm. to 10 cm.) and on large vigorous rhizomes when
they attain a stature of a meter more or less. When seen in its native
habitat, growing thickly in large clumps, with its tall leaves standing
straight up, it affords a spectacular sight. The picture which is seen
reproduced {4 — i) was taken in an open glade on a steepish wet
hillside in the mountains east of Crescent City, Calif, in August 1938,
when many of the leaves were just approaching maturity. The seeds
were already fully ripe, since the flowers (4 — 2, 3) are produced in
early spring, before the leaves start to grow.

The pitcher arises from a clasping base, the wings of which appear
concurrent with the ventral wing, the edge of which is doubled as in
Sarracenia ((5 — 15). The tube is tapering, widening upward. At
the top the tube spreads suddenly and at the same time is bent sharply
forward to form a dome, bringing the mouth into a horizontal posi-
tion underneath. From the front of the mouth a prominent forked
appendage, of ''swallow-tail" (Lemmon) or fishtail form, hangs
down with a forward curve. In the largest leaves the dome rnay be
10 cm. long, 6 broad and 5 deep, while in a very small, but still per-
fectly formed pitcher 1.5 cm. long, the dome measures only 2.5 mm. in
length. A feature peculiar to Darlingtonia is the twisting of the tube
either to the right or left so that the helmet-shaped dome is turned
about 180 degrees from the axis of the plant. All the leaves then
are turned outwardly, a position conceivably of advantage in attract-
ing prey. The small leaves often lie more or less prostrate and the
fishtail appendage lies on the surface of the ground forming a ramp
leading small creeping things to the opening {6 — 14).

When the leaf is yet immature, but of full extent, the tissues of
the dome are still soft, and the two sides lie against one another. In
attaining their final shape the sides expand, the dome is inflated, and
then becomes indurated, so that, supported by the sclerotic cell walls
and other mechanical tissues, the dome attains a marked firmness,
like a hard hat. The wings of the appendage spread to form a plat-
form leading to the opening, its ventral surface secreting much nectar

Chapter III — 43 — Darlingtonia

as a lure. Light green at first, the color gradually deepens and at
last becomes splashed with red. The roof of the dome and the back
of the upper part of the tube are mottled with numerous white flecks,
devoid of chlorophyll, glands and hairs, and, to an insect at the
mouth, form a visual lure {4 — 4, 5). Such fenestrations or areolae
are found in Sarracenia minor, S. psittacina, and S. Drummondii and
perhaps some others. The transparency of an areole is traceable to
the entire absence of chlorophyll-bearing tissues and of intercellular
spaces. Each areole lies in a mesh of the vascular tissue surrounded
by an irregular edging of chlorophyllous tissue with, inside, stomata,
glands and blunt curved downward pointing hairs, the latter encroach-
ing a little more on the clear tissue, which is composed of wavy walled
epidermis on both surfaces, with three or four courses of thick-walled,
perfectly clear cells. There is no pigment of any kind, except in old
leaves, when a yellow tinge may be detected. So complete is the ab-
sence of air-spaces and pigment, that the areole is quite glassy. A
few coarse starch grains occur especially toward the margins.

We will now consider the structures about the mouth. We note
in the first place that the ventral wing just below the mouth as in
all the Sarraceniae has a doubled edge, less conspicuous in some species
(S. purpurea) than in others (S. minor). This is most conspicuous in
Darlingtonia, in which the double margin may be traced down a long
way and appears, as it did to Torre y and to Kurtz, to be continuous
with the edges of the basal stipular wings. The embryology shows,
however, that this is not the case {See beyond). If it were not so,
then we should have to explain a condition in Darlingtonia which is
not common to the Sarraceniae. As the evidence indicates, the ven-
tral wing or keel originates in the same way in both these genera.
The condition in Heliamphora, which has a pair of independent wings,
cannot in the absence of embryological evidence be brought into
comparison, as Troll also remarked.

Now these two admittedly shallow free edges of the keel mark
the margins of the mouth. In Darlingtonia they may be traced along
the nectar roll and marking its outer limb. The wing edges are ac-
companied by the major wing veins, and these run forward to the
base of the fishtail, and enter it, one on each side, where they branch.
The appendage receives the end of the midvein also, but this
immediately branches. The fishtail is evidently due to deep emargi-
nation, as Goebel maintained, and is not a pair of pinnae, as Mac-
FARLANE believed. The condition in Darlingtonia is not parallel to
that in S. psittacina, in which the inrolled edges of the flap lobes
form valves with a weal along the edge of each representing the nectar
roll, but not of the same form. This receives only a minor branch of
the keel veins, which continues along the margins of the flap lobes.
In Darlingtonia the nectar roll results from hypertrophy of the leaf
edge in a lateral direction. The strong venation is correlated with
the supply necessary to the fishtail with its large number of active
glands and its large size. As already remarked, the veins running
along the outer Hmb of the nectar roll (5 — 7-10) pass forward to
enter the fishtail near its outer margins, there branch and furnish
the main supply lines. One readily infers that the outer marginal

Francis E. Lloyd — 44 — Carnivorous Plants

zone of each lobe of the fishtail is a continuation of the nectar roll
on its own side. Its position and topographical relations in the defini-
tive pitcher leaf show that it gets these as a result of torsion and con-
traction of the tissues at its base. I have been prompted to make
a guess as to what a primitive condition of the Darlingtonia leaf might
have been. Plate 6 — 13 represents such a hypothetical condition.
In order to get B, which with a little more forward curvature would
represent the modern pitcher, all that need occur is the transverse
contraction of the base of the flap accompanied by bending forward.
It should be noted that there is no fusion of the two sides of the nec-
tar roll in front, so that the inner superficies of the hood are con-
tinuous through the gap between the forward ends of the two sides
of the nectar roll with the ventral (upper) surface of the fishtail. That
a change of this nature has occurred in the process of evolution is
indicated by the case above mentioned of a juvenile leaf with a nectar
roll and an emarginate apex, but not contracted transversely at its
base (5 — 3). In this case the edge of the nectar roll is clearly con-
tinuous with the edge of the apical appendage. This is an objective
example of the hypothetical primitive condition presented in 6 — 11.

The fishtail appendage on its outer (dorsal) surface has stomata
and simple glands in great numbers. The inner or ventral surface has
no stomata, but there are numerous glands, and a good many stiff,
thick, blunt hairs turned morphologically downwards, but, because of
the upsidedownness of the hanging appendage, point upward and
furnish a rough surface which assists, rather than impedes, a climbing
insect, lured by the abundant nectar. To the presence of this there
is abundant evidence in the living plant. The converging folds of
the appendage, as an insect crawls upward, {4 — 5; 5 — 9), guide it
toward the entrance into the hood, where it meets the inturned nectar
roll. Once inside, the insect has to face the dangers of the inner
surface. It is not to be supposed that insects will insist on using the
appendage. Nectar glands occur everywhere on the outer surface.
The ventral wing, as well as the appendage, may act as a wing-fence
to guide them to the opening. Meeting the heavy exudation of nectar
on the nectar roll is an added spur to entrance, however they may
have been attracted thus far.

To turn to the conditions found in the interior of the pitcher. The
forward (upper) portion, the dome, is lined with many stiff, coarse
hairs so directed as to urge insects toward the depths of the tube. These
are largely absent from the areolae, though small ones may occur.
They are most plentiful on the floor, where there are no areolae. In-
termingled with the hairs are many nectar glands, so that the whole
forward portion of the floor of the dome serves as a feeding ground,
from which also insects can feed with great convenience on the nectar
roll, as from a table. The rear of the dome, however, the surface of
which extends down into the tube, has no glands, but the imbricated
epidermal cells are elongated, each into a sharp downwardly pointed
hair, which offers no foothold. This continues far into the tube, as
far as a point where there are no more fenestrations in the wall. Here
the character of the hairs gradually changes, and they become fewer
and longer. In the extreme depth of the tube the hairs are absent.

Chapter III — 45 — Darlingtonia

and there are no glands. If the absence of glands indicates anything
it is that in Darlingtonia the only digestion which may occur is that
induced by bacteria, and that this at least takes place has been testi-
fied by J. G. Lemmon in a letter to Canby who mentions the obser-
vation in a paper in 1875. Lemmon remarked that he detected a
strong smell of decay at some distance, as did Jones and others later.

The structure of the nectar glands is quite unique, though they
evidently may be regarded as conforming to the Sanacenia type.
On the surface a gland appears as one of the epidermal cells, or if
compound from two to five or six such cells {6 — 18, 19, 22). It ap-
pears filled with cytoplasm and a nucleus is always distinctly visible,
sometimes two or three (in the thin superimposed cells). Focussing
more deeply the gland cells become larger and rounded in outhne.
The reason for this is understood when a section is examined {6 — 20,
21, 23). It is then seen that the diameter of the glands increases
with depth and is composed of a row of flat cells, evidently derived
by periclinal division of an original epidermal cell. Underlying each
gland (if simple) is usually a single parenchyma cell, which in the
glands of the outer surface is quite deep, suggesting to Macfarlane
the adjective "globoid" {6 — 20, 21).

When the gland is compound there will be seen in section two
(rarely more because of the unfavorable chances of such a section)
tiers of flat cells. These glands are not only compound but are much
larger than those on the outer surface, where they are invariably small
(about the size of the stomata) and simple. Compound glands occur
in great numbers on the nectar roll, and, to a less extent, on the for-
ward interior face of the dome.

When a pitcher is allowed to lie in a weak solution of methylene
blue, the glands of the outer surface become stained throughout,
though the surrounding epidermal cells remain colorless. There is
evidently ease of diffusion through the external cells. Macfarlane
explained this by the absence of cuticle from the outer gland cell, say-
ing that he could observe the torn edges of the cuticle in a surface
view, but I have been unable to verify this. By the evidence of ex-
posure to methylene blue it also appears that the wafls of the gland
are cutinized (Goebel) except at the base, as is the case with the
glands of other genera of the Sarraceniaceae.

In the absence of digestive glands, but on the presumptive nutri-
tion of the plant from the decaying insects which are caught in great
numbers (Edwards counted 33 spp.), the question as to what part, if
any, of the interior surface of the tube can absorb the products of such
decay, is pertinent. We have seen that zone 4 in S. purpurea is devoid
of cuticle. In Darlingtonia it is surprising to note that the whole of
the surface from the lower limit of zone i, that is, below about two-
thirds of the dome, is capable of absorption. When a leaf is plunged
into a weak methylene blue solution for 20 hours the tissues, as far
as and including the outer part of the third layer of parenchyma, become
dyed, while no dye enters through the outer surface epidermis, except
through the nectar glands. There can hardly be any question, there-
fore, that the inner surface of the pitcher is capable of absorbing so-
lutes which result from the decay of insects within it. This is due,

Francis E. Lloyd — 46 — Carnivorous Plants

probably in a large part, to the absence of cuticle from the whole area
occupied by the long detentive hairs, according to Batalin (1880)
who observed the loosening of the cuticle from the free surface of the
cells by the formation of blisters (in Sarracenia flava). Batalin even
suggests that this non-cuticularized epidermis takes over in the ab-
sence of glands, their function, not only of absorption but also of di-
gestion, since throwing off the cuticle seems to be indicative of the
excretion of some substance, possibly digestive. The condition in
Darlingtonia does not seem to be wholly parallel to that described
by Batalin for Sarracenia. I placed a pitcher in methylene blue over-
night and found the whole inner surface stained deeply in the morn-
ing. On sectioning, the whole inner epidermis was found deeply
colored. On staining with Sudan III there was distinct evidence of
cuticularization, especially in the radial walls. The outer walls were
thinly stained, sometimes not at all, while the cuticle of the outer
epidermis was obviously thick and richly stained. I could not, how-
ever, find clear evidence that the matter stands as Batalin describes

it.

Development of the leaf. — Material for the study of the development
of the leaf in Darlingtonia was obtained on May 22, 1938, growing in a
sphagnum swamp 6 miles N. of Florence, on the coast of Oregon at a
few feet above sea level. At that time the plant was in full flower, and
in some plants very young leaves were beginning to make their ap-
pearance. New leafage would be achieved in the course of a month,
the present leaves having persisted since the previous season. In the
depths of the pitchers were to be found merely the chitinous remains
of insects long since caught, and no odor, such as has been detected
by others during the active season, was noticed.

The morphology of the leaf is easily the most complicated of all
the pitcher plants of the Sarracenia type. This is because of the
torsion of tissues which occurs at the outer (distal) extremities of the
two sides of the nectar roll, and the edges of the fishtail flap. The
nectar roll appears to be extended as an infold of the outer edges of
the fishtail flap, which hangs down from the distal sector of the open-
ing, its ventral face being that one which faces the tube of the pitcher.
We may follow the development of the leaf in examining the follow-
ing series of stages, chosen conveniently.

Case I. A very early stage of development (<5 — i) in which the
whole leaf consists of a flat cone 0.3 mm. high. This may be regarded
as identical with the corresponding early stage of Sarracenia purpurea
as represented by Troll (1932) and earlier by Goebel (1891), though
in Goebel's figure the mouth of the beginning pitcher is too wide,
and the leaf -base is not shown. The mouth is not set so nearly hori-
zontal as in Sarracenia. The margins of the leaf-base wings are con-
tinuous transversely from one side to the other. A small stretch of
tissue separates this from the edges of the mouth, already well marked.
The rim of the mouth is continuous all around, making peltation com-
plete.

Case 2. Leaf 0.7 mm. tall {6 — 2). The mouth and its continuous
rim form a definite papilla, the upper margin taking the lead in up-
ward growth. The tissues between the lower transverse rim of the

Chapter III — 47 — Darlingtonia

mouth are somewhat raised to form a low ridge. The twisting, charac-
teristic of the Darlingtonia leaf, has already begun.

Case 3. A leaf 1.5 mm. long {6 — 3). The leaf base has elongated,
carrying the margins of its wings up some distance. Above, the rim
of the mouth has been extended down as a low double ridge and the
lateral reaches of the rim now begin to form the two sides of the ter-
minal fishtail of the mature leaf {6 — 6). The ascidium reaches well
down into the leaf base.

Case 4. Leaf 2.6 mm. long {6 — 7). The wings of the leaf base
have now developed so that the distinction between this and the leaf-
blade is sharp. The double ridge, continuous with the two sides of
the mouth is longer and is raised up on the edge of the ala ventralis.
The close apposition of this with the apex of the leaf-base wings shown
by Troll for Sarracenia does not occur here. It has now become
clear that the double character of the edge of the ala ventralis is de-
rived from the rim of the mouth. If not so extended in Sarracenia,
yet the origin of the double edge is the same. In this case the twist
of the leaf is to the right.

Case 5. A trifle older than case 2, not so old as case 3, in sagittal
section {6 — 4). Here can be clearly seen the identity of the side lip
of the mouth and the edge of the keel. The pore of the mouth is still
small. The section being truly sagittal, the other keel edge is not seen.
No indication of the nectar roll is yet visible. Advance beyond this
stage consists of the enlargement of the lateral reaches of the lips of
the mouth concomitant with the laying down of the nectar roll and
its continuation along the outer margins of the fishtail.

Cases d, 7 and 8 {6 — 5, 6, 10). Successive stages following on
case 5, showing the development of the fishtail from the sides of the
mouth, the apex being now arrested and of slower growth. In cases
7 and 8, the outer marginal roll of the one side of the fishtail is seen,
and that it is continuous with the nectar roll which has also appeared.
The fold between the distal ends of the nectar roll has begun develop-
ment.

Case g {6 — 16). Surface view of a somewhat later stage, about
like that shown by Goebel. The difficulty of interpretation is obvious.

Case 10 {6 — 8). The dome has begun development and the tube
is twisted through 90 degrees. The distinction between the edges of
the wings of the leaf base has become obscure, except in transverse
sections (5 — 5). Seen in sagittal section the dome is represented in
6 — 9. The fold {6 — 10) has now come into a vertical position as the
dome has enlarged fore and aft, and the outer marginal roll of the one
side of the fishtail is seen continuous with the nectar roll, which has
pushed forward. The ventral surface of the fishtail is continuous with
the inside surface of the dome.

In a word, all parts are now clearly defined, and the glands have
appeared. The final condition may be seen in various figures illustrat-
ing the mature leaf. At the time growth is complete, the leaf has
twisted through an angle of 180 degrees, though it may be as small as
90 degrees or as large as 270 degrees. The torsion does not involve the
dome. It is either to the right or left in any given plant (antidromy
of McClosky).

Francis E. Lloyd — 48 — Carnivorous Plants

Digestion and Absorption. — Edwards (1876) and Goebel were of
the opinion that true digestion, that is, by means of a secreted enz3mie,
does not take place in Darlingtonia. More recently Hepburn and
his collaborators St. John and Jones (1920, 1927) examined the fluid
of unopened, cotton-plugged and open pitchers with regard to its
effect chiefly on carmine fibrin and fibrin in the presence of a bacteri-
cide (0.2% trikresol). Of a total of 57 experiments in the laboratory
and field, none gave a definitely positive result, occasional, very slight
aberrancies being due probably to the presence of bacterial ferments.
On anatomical grounds this is to be expected, though as above noted,
Batalin made a suggestion that the non-cuticularized cells of the
depths of the pitcher might take over the function of the glands. But
that the function of the secretion of a protease could be one seems,
in view of the above cited results, to be out of the question. That
insects are disintegrated by bacteria is obvious, and that their products
are available as nutriment to the plant is indicated by the fact that
absorption of various substances can and does take place as shown
also by Hepburn and his colleagues, and as would appear to be the
case in view of the non-cutinized tissues of the pitcher through which
methylene blue readily passes. Hepburn, St. John and Jones showed
that water is absorbed, and dissolved lithium was found to have been
taken up by the tissues. When various nitrogenous substances were
introduced, both these and the solvent were absorbed, but in the
presence of a phosphate buffer the water might increase though the
compounds were absorbed. Mrs. Austin had found (1876) that when
stimulated by the introduction of bits of meat, the amount of fluid
increased in the pitchers. Her results were quoted by Asa Gray
(1876). Though the experiments were done in the field, there is as-
surance of the exclusion of rain which, if any fell, which is quite un-
likely, could gain no entrance into the hooded pitchers. Hepburn
et al. investigated this point, also in the field (Plumas Co., Calif.) and
found that when milk was introduced into the pitchers, there was in-
variably an increase in the amount of fluid ranging from 20 to 1242
per cent in periods of 1-7 days. They studied 77 pitchers, and the
amount of increase of volume varied independently of the time, so
that some pitchers were much more active than others. When beef
broth was used, there was an increase of from 302 to 387 per cent in
fluid content in five days. When bits of meat were used the results
depended on whether the meat was cooked or raw. If cooked there was
little if any increase, because only small patches of the surface were
affected. If raw, an increase of volume of from 48 to 157 per cent
was observed. No results were obtained with raw or coagulated egg-
white, nor with cheese, casein or fibrin "possibly for the same reason
as with meat." When acids and alkalis in very dilute solutions were
introduced, there was no very "marked tendency" for the volume of
fluid to "increase or decrease", but it was noted that, as in the human
stomach, the fluid returned to neutrality whatever the nature of the
introduced reagent.

Has the fluid of pitchers the power of wetting insects, when im-
mersed, more than pure water? While positive evidence was ob-
tained for other species of Sarraceniaceae, that from Darlingtonia,
from experiments done in the field, was purely negative.

Chapter III — 49 — Darlingtonia

Experiments done by the same authors to determine if other
enzymes than protease might be detected in Darlingtonia gave nega-
tive results except for diastase, of which, however, only a trace could
be detected. Maltase, invertase, emulsin and urease were absent.
It seems, therefore, indisputable that this plant depends solely upon
the activity of bacteria to provide the absorbable protein and other
nutrients, if any, through the pitcher walls. Edwards' opinion, ex-
pressed in 1876, turned out to be correct.

The presence of bacteria and their activities were observed by
Hepburn ct al. A chemical study of the pitcher fluid was made by
these authors who found that in closed, plugged and open pitchers, a
small amount of nitrogen could be recovered, viz. 0.027% from closed
pitchers, 0.015 % to 0.009% from plugged pitchers and 0.034 f; to 0.049 %
from open pitchers. The fluid studied has a specific gravity of 1.003
at 15 degrees C. and contained 0.213% solids, 0.104% ash, and 0.046%
calcium oxide (lime) forming 44.23 % of the ash. Chlorides were present.
No reducing sugars could be found, though it is quite probable that
such may sometimes be present by contamination with the nectar
found elsewhere on the walls of the pitcher.

Literature Cited:

Ames, Mary E. P., Calif. Horticulturalist and Floral Magazine 10:225-229, 1880. Quotes

a letter from Mrs. Austin re increase of fluid in pitchers of Darlingtonia.
Arbee {see under Cephalotus).
Austin, R. M. L., Brief an Dr. K. Keck, iiber Darlingtonia. Oester. Bot. Zeitschr. 1876:

1 70-171.
Barnhart, J. H., Brackenridge and his book on Ferns. Journ. N. Y. Bot. Card. 23:117-

124, 1919. .

Batalin, a., tjber die Function der Epidermis in den Schlauchen von Sarracema und Dar-
lingtonia. Acta Hort. Petropolitani 7:346-359. 1880.

Braun, a., Uber Darlingtonia californica Torrey. Sitzungsber. d. Gesellsch. naturf.

Freunde, Berlin 1873:73-75- , ^ * a a c •

Canbv, Wm. M., Darlingtonia californica, an msectivorous plant. Proc. A. A. A. bci.
1874:64-72, Salem, Mass. 1878. Reprinted in Oester. Bot. Zeitschr. 1875:287-293.

D.ARWiN, C, Insectivorous Plants. London 1875.

Edv.'ards, Henry, Darlingtonia californica Torrey. Proc. Calif. Acad. Sci. 6:161-166, 1875
(published in 1876).

GoEBEL, K., Pflanzenbiologische Schilderungen. Part 2, V. Insectivoren. Marburg, 1891.

Gray, Asa, (Description of the seed of Darlingtonia). Amer. Journ. of Science and Arts,
2 ser. 35:136-7, 1863.

Gray, Asa, Darwiniana. Appleton, New York 1876, 330 pp. (Cites Austin's Observa-
tions on fluid in pitchers of Darlingtonia).

Hepburn, J. S., F. M. Jones & Eliz. Q. St. John, Biochemical studies of North Ameri-
can Sarraceniaceae. Trans. Wagner Free Inst. Phila. 11:1-95. 1927- A very full bib-
liography.

Hooker, J. D., On the carnivorous habits of some of our brother organisms — plants.
Rep. Brit. Assoc. Adv. Sci., Belfast 1874. .

Kurtz, F., Zur Kenntnis der Darlingtonia californica Torrey. Verhandl. Bot. Vereins
Brandenburg, meeting June 2, 1878, 24 pp.

Lemmon, J. G., Brief an Dr. K. Keck iiber Darlingtonia. Oester. Bot. Zeitschr. 1876: 35.

Macbride, J., On the power of Sarracenia adnnca to entrap insects. Trans. Linn. Soc.
London 12:48-52, 1817 (read in 1815).

Macfarlane, J. M., Observations on the pitchered insectivorous plants, I. Ann. Bot.
3:253-266, 1889, 1890.

Macfarlane, J. M., Observations on the pitchered insectivorous plants, II. Ann. Bot.
7^.03-458, 1893.

Mellichamp, J. H., Letter to Dr. Hooker on the CaUfornia pitcher plant. Gard. Chron.

1871:46.
Mellichamp, J. H., Notes on Sarracenia variolaris. Proc. A. A. A. S. 23 meeting, 1874.
1875:113-133. An earlier communication appeared in Gard. Chron. 1874:818-819,
earlier published in the N. Y. Tribune by Asa Gray.

Francis E. Lloyd — 50 — Carnivorous Plants

ToRREY, John, On Darlingtonia californica, a new pitcher plant from Northern California.

Smithsonian Contrib. to Knowledge 5:1, 1853. (Year of discovery given as 1842.

According to Barnhart, 1919, the year must have been 1841).
Troll, W., Morphologie der schildformigen Blatter. Planta 17:153-314, 1932.
VoGL, A., Die Blatter der Sarracenia purpurea. Sitzungsber. Wien. Akad. Wiss. 50, Oct.

1864.

Chapter IV
NEPENTHES

Geographical distribution. — Habitat. — General character. — Morphology of the leaf
and the seedUng. — Development of the leaf and adventive shoots. — The pitcher (Mor-
phology; Variety of form, color etc.; The mouth; The lid; Spur; Special anatomy). —
The rim or peristome. — Histology of the peristome. — The glands: their histology. _ —
Anatomy of the pitcher-wall (Vascular system; The interior surface; Wax zone; Digestive
zone; Rim). — Digestion. — The animal life of the pitchers. — Folklore, uses. — Antisepsis
of pitcher fluid.

The species of Nepenthes are found scattered throughout the tropics
of the Old World with the center of distribution in the region of
Borneo, being found as far East as N. Austraha and New Guinea,
and to the West in Ceylon and in Madagascar, its extreme outpost
(Danser). Madagascar, indeed, was the scene of its first discovery by
the Governor, Flacourt, in the middle of the 17th century, and it was
reported from Ceylon a little later by Paul Hermann, a physician,
who sent the specimens to Commelin in Amsterdam. (Wunschmann

1872).

They grow with rare exceptions only in moist or very moist situa-
tions, and they are successfully cultivated in greenhouses only if the
relative humidity is kept very high; in particular, a slightly reduced
humidity inhibits the development of pitchers. In their vertical dis-
tribution they occur from near sea-level to 9000 ft. altitude {Nepenthes
Rajah and villosa, on Kina Balu, Borneo). They are chiefly jungle
plants, though one species at least {N. destillatoria in Ceylon) grows
in wet savannahs where it climbs on scattered shrubs. A^. gracilis was
found by Korthals (1839) in "dry sandy, stony ground" though it
was found to prosper better in other, moister situations. The de-
mands of the plant are for wet soil and hot to cool temperatures ac-
companied by a high humidity of the air.

It is most rarely that they can be successfully cultivated outdoors
in temperate regions but it was reported some years ago at a meeting
of the Naturalists Club of Sydney, N. S. W. that two unidentified
species were grown out of doors on a trellis, at Parramatta, not far
from Sydney. This is a region where staghom ferns are grown out of
doors by everybody, and the Nepenthes species above mentioned may
be especially hardy.

In general appearance the species of this genus are pretty uniform,
the more striking differences being found in the size and shape of the
pitchers. The plant consists of a creeping rhizome from which spring
coarse, clambering vines with thick, glossy leaves of frequently con-
siderable length (i meter) arranged in a "^5 phyllotaxy, though one
species (A. Veitchii) is wholly distichous (Troll 1939). The leaf con-
sists of a spreading winged base narrowing into a short isthmus
beyond which it spreads into a hgulate to orbicular blade beyond which
extends a short or long tendril which can twine about a support and
ending in a pitcher with a lid overhanging the mouth, behind which

Francis E. Lloyd — 52— Carnivorous Plants

is a small or larger spur. The pitcher is always held in an upright
position. When young the various parts are clothed with a tight rusty
pubescence of curiously branched hairs. In cKmbing, often to the
crowns of tall trees (i6 to 20 meters: N. bicalcarata, Rafflesiana, etc.
according to Macfarlane), the plant supports itself by means of the
stout tendrils. It sometimes grows epiphytically, as in the case of
N. Veitchii (Burbidge, 1880). Such species may have cHmbing stems
3 cm. in diameter. Troll (1932) has given us an excellent^ word
picture of the appearance of N. ampullaria {4 — 9) in its habitat.

"I came across N. ampullaria among the massive vegetations of a
swamp-forest on the island of Siburut, off the west coast of Sumatra.
It was a fabulous, unforgettable sight. Everywhere, through the
network of lianas the peculiarly formed pitchers of this species gleamed
forth, often in tight clusters; and, most remarkably, the muddy, moss
overgrown soil was spotted with the pitchers of this plant, so that one
got the impression of a carpet. How is this pecuUar behavior to be
explained?

"iV. ampullaria develops a rhizome which creeps in the earth or
between clumps of moss. This sends out one or more hana-like shoots
which cHmb high into the trees, and at their ends, where they can en-
joy bright illumination, they become leafy. The leaves of these long
shoots are of the usual type — they possess a well developed lamina
and a functional tendril. Elsewhere the Manas are bare or have re-
mains of dead leaves clinging to them.

"Of quite a different appearance are the pitcher leaves which are
found on the ground. True, the pitchers are well developed, but the
tendrils are always short and serve only to hold them in an upright
position.

"If one searches for the attachments of these simplified leaves,
they will be found to occur on short branches, just as Goebel de-
scribed them. It has been overlooked, however, that they are not con-
fined to the main rhizome but spring also from numerous prostrate
stems which attain a considerable thickness. Such branches may be
followed for a distance of several meters along the soil surface quite
easily because of the numerous dense clusters of pitchers which are
strung along them." {Translated).

Earlier observers in some cases thought that the lid of the pitcher
is capable of motion, and so to close and open its mouth. Loureiro
is mentioned by Sims (1826) to have held this view. But this of
course is not the case — the Hd attains a quite fixed posture, usually
overhanging the mouth of the pitcher, but sometimes turned quite back.

The morphology of the very highly specialized leaf of Nepenthes
can best be considered by a comparison of the mature condition with
that met with in the leaves of seedUngs and of adventitious shoots
on cuttings. The former have been studied by Dickson, J. D. Hooker,
Goebel, Macfarlane and Stern. In spite of a general uniform-
ity of evidence, with exceptions to be noted, there is a wide divergence
of opinion as to the homology of the parts, Macfarlane regarding
the leaf as a p'nnate structure and Goebel as a simple leaf with a
highly specialized region forming the ascidium or pitcher. These and
other interpretations will be considered.

Chapter IV — 53 — Nepenthes

Seedlings. — The primary leaves of the seedling (first described by
BiscHOFF in 1834), the cotyledons, are elongate oval and present no
noteworthy features. The following leaves, which will for convenience
be called primary, consist of a short spreading and clasping base,
narrowing briefly to expand at once into a pitcher (Korthals) with the
edges of the leaf base extending up its ventral (adaxial) face as two
wings which either meet transversely somewhat beneath the rim of the
pitcher mouth (Hooker, 1859, Dickson, Macfarlane), or end
abruptly without meeting (Goebel). Stern, restudying Goebel's
material, verified this but pointed out that he found a row of gland-
like tentacles (7 — 5) and these might indicate a transverse connection.
Troll strongly favored the view that there occurs actually or funda-
mentally a union of the wings below the rim to express '"total stipula-
tion." The edge of the mouth of the pitcher is armed with a transverse
rim usually well developed, and occupies about one-half to two-thirds
of the peripher>^ the rest being taken up by the base of a lid, that is,
in the primary leaves the lid base is very broad (7 — 7, 9) while in
the adult leaf type it is narrow, with the consequence that the veins
are spread apart in the former and crowded together in the latter.
The venation of the lid appears quite evidently to be an extension of
the plan of that of the pitcher, and not secondary as is that of the rim,
if we may lean on juvenile leaf forms arising on small forced shoots.
The lid bears a number of tentacle-hke emergencies at its edges and
upper surface, and behind it extends an appendage which is properly
regarded as the organic apex of the leaf, the "spur." With the advance
of age, the region betw^een the leaf base and the pitcher elongates, so
that a blade now intervenes, with its margins continuous with the wings
of the pitcher. The intercalation of a tendril at this region is indicated
in the narrowing of the blade (7 — 7, 11), and in the more mature
condition a tendril is realized. The leaf then consists of an expanded
base, a blade, generally of some length, a tendril which becomes
functional as such, supporting at its end the pitcher which is always
winged, though less obviously, it may be, than in the seedling {4 — 7,
8). In some species the pitchers on the higher parts of the plant have
the wings reduced to mere ridges.

The early development of the pitcher leaf has been described by
J. D. Hooker (1859), Bower, Stern, who, as to the facts, agree.
In the very early condition, there is to be observed a depression just
below the apex of the yet merely low conical structure (7 — i). The
lid develops as a transverse ridge at the distal limb of the depression
and is independent of the true apex (7 — 2). The lid is therefore not
the tip of the leaf, but an outgrowth on the ventral face of the leaf
near its apex (Hooker). It grows downward over the opening, which
in the meantime becomes deeper to form the acidium. It has the
appearance of a two lobed affair (7 — 4), and that it is really such
has been thought by Bower and by Macfarlane who cite in support
of their view the fact that the lid in the mature leaf is often emargi-
nate. The conical apex continues its development into an expanded
leaf tip which may at length bear one to several expanded lobes
{N. ampullaria), "pinnae" as they have been called, and Macfarlane
regards them as supporting evidence of his theory that the whole leaf

Francis E. Lloyd — 54 — Carnivorous Plants

is a pinnate structure obscured by secondary changes. They are more
or less conspicuous on mature leaves in some species {N. ampullaria)
while on others the spur, as it is called, is a tapering simple conical
projection often much displaced by the secondary growth of the tissues
beneath it so that the lid is moved forward to occupy an apparently
terminal position (4—10; 7 — 23). Meanwhile the leaf blade de-
velops more or less in front, i.e. on the ventral surface, of the enlarging
ascidium in two usually deep ridges, the margins of which are con-
tinuous to the base. From their position it appears clear that the
ascidium is formed by the expansion chiefly of the lower moiety of
the midrib, so that at full growth the leaf margins mark the limits of
the upper surface of the midrib.

In adventitious shoots produced by forcing cuttings, good material
of which I obtained at Munich, various embryonic conditions of the
leaf are preserved in the mature condition, which are always small
and embryonic ("juvenile") in appearance as in fact. This is to be
referred to the failure locally of the incidences of growth which would
mold the leaf into the mature form, such as the failure of the leaf to
elongate in the region giving rise to the tendril; or the continuation of
growth where it is normally suppressed, such as in the narrowing of
the blade at the base of the ascidium. The former is shown in Fig.
7 — II which is nearly mature, the leaf blade being here narrowed in
the region which in a completely developed leaf would have become
the tendril. The second condition is shown in Fig. 7 — 13 in which
it is seen that the leaf blade has expanded, beginning to do so at the
middle point of the ascidium instead of below the base. In both these,
as in other early stages of development, the apparent "two-lobed"
condition of the lid, seen by Bower and others, stands out. That this
is more than appearance may be doubted. It may be contended that
the lobing may be an appearance due merely to the infolding of the
middle longitudinal zone, the marginal zones resting on the rim of the
pitcher, which during the earlier stages of development is laterally
compressed so that the sides of the mouth, that is of the rim, are
close together and parallel (7 — 24; 8— 19). The presence of emargi-
nation is not by any means general, and at best, as Goebel points out,
its presence is not an indication of lobation. In any event emargina-
tion may easily occur when it does, from the manner of longitudinal
folding by mutual pressure of the rim and Hd apex.

The spur (we continue to treat of juvenile leaves of short shoots)
is usually broad and lobed, and, being the organic leaf apex (Hooker)
receives the terminal part of the mid vein, which does not pass into
the lid, so that this is devoid of a midvein (7 — 7-10). Below the base
of the spur, however, the midvein of the pitcher may send anastomoses
joining it with lateral veins. The venation of the spur is made up
almost wholly of lateral veins derived from far down at the base of the
pitcher, swerving around from back to front, and then back again
below the rim. In specimens resembling the more adult type of
pitcher, veins appear in the lid which, though suggesting a midvein, are
really branches and anastomoses between the laterals and the midvein
(7 — 9; Text fig. 2, p. 63).

The mature leaf may in some species attain a length of one to

Chapter IV — 55 — Nepenthes

three feet. It consists of an expanded base, sometimes connate about
the supporting stem, and expands above into an elongate blade cor-
responding morphologically to the narrowed portion of the seedHng
leaf. At the apex it may sometimes be found to be peltate {N. clip-
eata), and this, as above said, is compared by Macfarlane to the
peltation observed by him of the two ventral ridges just below the
mouth of the pitcher. Beyond this there occurs a tendril which is
short and non-functional as such in soil rosettes {e.g. N. ampullaria),
but which in the climbing forms becomes long, stout and twining.
Sachs (1896, through Goebel) held that the tendril activity (the
actual winding) acts as a stimulant to the growth of the pitcher, but
the evidence is not convincing, for it is quite usual to find well de-
veloped pitchers when no winding has intervened (4 — 7, 8). Though
the tendrils wind about supports, they may wind even when supports
are not available; but it is not true, as Oudemans thought, that this
winding is a means of bringing the pitchers into the proper position.
The sensitive tissues which are responsible for this occur at the base
of the pitcher and neighboring portion of the tendril (Stern).

The Pitcher. — It is with the structure and behavior of the mature
pitcher that we are chiefly concerned. It shows a considerable variety
of form, from that of a cylinder (7\^. phyllamphora, N. gracilis), a
cylinder modified by a basal globular expansion {N. ventricosa, N.
Lowii), an open funnel, narrowest at the base {N. inermis, N. dubia),
to an oval sac slightly compressed laterally (N. ampullaria). All of
these forms have been illustrated by Danser (1928). In most species,
and this is especially noticeable in the approximately cylindrical ones,
the upper one-third, more or less, is somewhat constricted, correspond-
ing in extent to the waxy zone within (to be described beyond). From
some species this is absent {N. ventricosa, N. bicalcarata, N. ampul-
laria) or may be very narrow {N. intermedia). It is said to be ex-
ceptionally present in forms from which it is normally absent. The
size of the pitcher may reach in some species the length of a foot,
with a capacity great enough to accomodate small mammals, birds,
etc., e.g. N. rajah 25-30 cm. by 12 cm. (Hooker). The majority of
species have pitchers which range from 5 to 15 cm. in length.

The pitchers produced even in a single individual, this being a
character of the species, may be of two or even three different forms,
that is, they may be mono-, di-, or tri-morphic (Macfarlane). When
this occurs, the rosette leaves in contact with the soil differ from the
cauline, the uppermost of these being again different from those mid-
way of the plant. Thus N. ampullaria has rosette leaves with goblet
shaped pitchers, the cauHne ones being cylindrical; while in N. Bosch-
iana, N. maxima and A^. Vieillardii, the lowermost pitchers are globose,
the lower cauline tubular and the uppermost infundibuHform or cornu-
copioid. So different are they that different pieces of the same species
have been described as different species. In some cases the internal
structure differs, there being a wax zone in some pitchers and not in
others. In color the pitchers are usually green with more or less
splotchings of red, and when this occurs in the rim the color lies in
very definitely regular transverse stripes, obviously connected with the
regular, straight-rowed arrangement of the cells. Some species have,

Francis E. Lloyd — 56 ^ — Carnivorous Plants

according to Macfarlane, ''porcellaneous white" pitchers marked
with "deep crimson splotches" {N. Raffiesiana var. nivea, N. Bur-
bidgei). Others have uniform deep red color, even when growing in
the shade, or covered with a growth of moss, while the pitchers ex-
posed to greater illumination are less deeply colored, (N. Rajah, N.
Edwardsiana) . These relations, in perhaps less striking fashion,
are shown by N. ampullaria in which the soil pitchers are splotched
with red while the cauline pitchers are almost or entirely free of color.
Some species have pale green pitchers with no markings at all (N.
ventricosa) {4 — 7). On account of the frequently brilliant coloring, be-
lieved by Troll to be, in addition to the nectar, attractive to insects,
the pitchers are regarded by Malayans as "bungabunga" (flowers)
(Troll 1939). The glossy rim may be entirely red or trans-
versely striped with red, or devoid of color other than green. The
outer surface of the pitcher is usually clothed with a rough pubescence
of many branched hairs, each rising from a unicellular stalk with thin
walls, those of the rest of the cells forming the branching complex
being very thick {S — 4). There are also low sessile stellate hairs
which in some species {N . intermedia) stand in a pit {8 — 1-3)-
The four arms forming the star are each two-celled, but the whole
may be composed of eight to sixteen cells. They are regarded as
hydathodes by Stern (191 6). These trichomes are by no means con-
fined to the pitchers, however, the whole plant showing a marked de-
gree of the rough hairiness, especially along the tendrils and the backs
of the "phyllode."

Borne on a tendril, often hanging, the pitcher in order to function
must stand upright. This is accomphshed by tropisms resident in
the region between the pitcher base and the end of the tendril. Since
the tendril is positively geotropic, and the pitcher " geotropically con-
ditioned," though not simply negatively geotropic (Stern), the usual
position is a sharply upturned pitcher on the end of the vertically
hanging tendril.

In one species at least {N . hicalcarata) the portion of the tendril
near the pitcher is swollen and hollow to form a formicary, but the
space is separated from that of the pitcher by a partition and it re-
mains filled with air. Ants usually eat away an entrance into the in-
terior, as they do e.g. into the stems of Cecropia and the thorns of
Acacia sp. etc., and use the hollow as a nest.

The mouth of the pitcher is always more or less oblique, and dur-
ing development is hermetically sealed by the lid, which opens only
when the definitive size and shape of the pitcher is almost attained.
It is well known that, until this happens, the contained fluid, of which
there is a considerable amount, is kept in a bacteria-sterile condition.
The method by which the edge of the lid is kept hermetically sealed
during development is both interesting and unique. There is, it must
be observed, no concrescence or fusion of tissues (7 — 24; 8 — 19).
What happens is that the edge of the hd is in the first place tightly
applied. Then, whatever chink there may be left is tightly sealed by
a dense growth of branching hairs which clothe the outer face of the
pitcher mouth and the edge of the Hd (Macfarlane 1908). These
interweave so as to produce a firm wad of cottony stuff. As long as

Chapter IV —57— Nepenthes

the growth of the two parts is synchronous the sealing remains effec-
tive. During the last phase of development differences in growth cause
the Hd and pitcher mouth to separate and the former, as the result of the
growth of the isthmus of tissue between the hd and pitcher edge, is
hfted in many cases a considerable height above the mouth (7— 22, 23).
In its final position the lid may overhang the mouth, becoming a
more or less effective bar to the entrance of rain, especially in such
forms as N. Rajah Hook, in which the lid continues to grow and attains
a sufficiently large size to overshade the opening entirely. In other
species it remains small and narrow and turns completely back, fully
exposing the mouth of the pitcher (.V. ampidlaria Jack, .V. dubia
Dans.) (4 — 9), and though overhanging the mouth, is obviously
quite ineffective as a roof (A', incrmis Dans.). When the lid is large
and overhanging in position, it is thin, more or less emarginate, in-
dicating to Bower and to Macfaelane that the two halves of the lid
represent paired pinnae. In some species there is a median ridge on
the inner surface bearing numerous nectar glands (7 — 25), and in
other species there is a shallow invagination near the apex, the function
of which, if it has one, is not clear; or, as in .V. Ladenhurgii, there is a
short clavate projection. In .V. Tivcyi (and, says Macfarlane, in N.
maxima) there is a short, thick, glandular crest or ridge near the base
and near the apex a sharp thorn-like projection, hollow on its forward
surface (7 — 25).

The under surface of the Hd is the seat of numerous nectar glands
except in a few species {N . ampidlaria, N. inermis probably). In
N. Lowii Hook., it is suppKed with many small appendages or bristles,
as Danser calls them, with nectar glands on the general surface be-
tween their bases.

At or below the base of the hd on the outside of the pitcher stands
the spur. This, as may readily be ascertained by examining the young
pitcher during development, is the apical portion of the leaf (Hooker)
and it appears that the Hd is an outgrowth over the upper surface.
The spur is very small in some species and stands just at the base of
the lid {N. inermis). In N. bicalcarata, e.g., it becomes considerably
displaced downwardly, and stands out, quite suggesting a spur, from a
neck of tissue which raises the lid far above the opening {N. bical-
carata) (7 — 23). Sometimes the spur is compound and bears pinnae-
Uke laterals, suggesting lateral leaflets (Macfarlane) {N. ampidlaria,
N. phyllamphora) .

Special anatomy. — The edge of the mouth of the pitcher is of dis-
tinctly pecuhar structure. It appears to be a parapet standing on the
edge, sloping inwardly on the whole, but with the outer margin some-
times turned more or less down. In a section of it made transversely,
it is T-shaped with the arms of the T of various lengths, according to
the species. In the majority both arms are of some length, so that the
parapet in such cases overhangs as much on the outside as on the in-
side, and with a general slope as much away as toward the opening of
the pitcher. N. ventricosa may be cited as an example of this con-
dition (7 — 16). In others (7 — 15, 17) the inner arm is short,
the outer long, while in .V. inermis (7 — 20) both are very short, the
outer a trifle longer than the inner. In N. Veitchii the width of the

Francis E. Lloyd — 58 — Carnivorous Plants

rim towards the lid is so great (up to 60 mm, says Danser) as to bear
a likeness to a "Marie Stuart collar" (de Ruiter 1935). The greatest
reduction of the inner arm is found in N. Lowii (7 — 18), which has
been described as without a peristome (Danser). There is, however, a
row of glands embedded in tissues which project to form a slight,
interrupted shelf while the outer arm is of some width relatively. At
the other end of the series stand such forms as N. hicalcarata, N .
intermedia and A^. ampullaria (7 — 19), in which the outer arm is very
short and tightly reflexed and the inner very long; in these species the
peristome has a very pronounced funnel shape. In N. ampullaria,
which forms rosettes of pitchered leaves on the forest floor, the pitchers
partly buried on the humus, the whole constitutes a group of pitfalls,
each with a broad overhanging edge which would prevent escape quite
effectively in many cases.

Of the two arms of the T, one, the outer, represents the true
pitcher mouth edge, outwardly reflexed. The inner arm is an out-
growth from the inner wall near the edge. This is easily seen to be
the case in young pitchers during their development (Heide, 19 10)
{8 — 19). In any case it can be seen that the vascular tissues of the
inner arm are derived by sharp branching from the main trunks which
extend to and along the edge proper.

But although the peristome is composed as it were of two flanges,
an outer, the edge of the pitcher mouth, and an inner, growing out as
a ridge from the inner wall, the whole during late development is so
moulded that the two flanges are amalgamated to constitute a single
organ, the inner surface of the edging flange and the outer surface of
the side flange becoming a continuous uninterrupted surface. The
whole is mechanically very rigid, for it is strengthened by a very thick
cuticle and the surface is broken up into minute striae and coarser
corrugations (4 — 11). The latter give the peristome their ribbed
appearance, and their most pronounced expression is reached in N.
villosa Hook. On the inner edge of the peristome the corrugations
end in minute teeth, and between each pair of teeth (7 — 21) there is
an opening, the mouth of a large nectar gland which lies buried in the
tissues. The nectar oozes in a drop held between a pair of teeth, of
access to insects standing on the rim and reaching down. This ar-
rangement together with the nectar glands on the under side of the
lid constitute a lure, the ''attractive zone" of Hooker. The hard,
glossy surface of the peristome is not, as it may seem to the eye, a
smooth, slippery one, for as a matter of observation, small insects
(ants, etc.) can walk freely on it, using their footpads. When the
tissues below the base of the lid are considerably extended, as they
are in A^. hicalcarata and N. intermedia (7 — 22, 23), the peristome is
extended likewise, and in these two cases, but only in these, there is, at
its extreme upper ends which are separated by the base of the lid, a
very strong development of the last dozen or so corrugations to form
two long sharp thorns, resembling the canine teeth of a cat. In
A^^. hicalcarata, these are long, solid, curved, very sharp and distinctly
canine in appearance. A rather fanciful explanation of the use of these
was advanced by Burbidge (1880) who pointed out that the Tarsius
spectrum, a small, insectivorous, monkey-like mammal, "visits the

Chapter IV — 59 — Nepenthes

pitchers of N. Rafflesiana" (which is similar to N. hicalcarata in all
respects except that it lacks the canine-like thorns), "and empties
them of their prey, but not those of A^. hicalcarata, in which the very
sharp spurs are so arranged that the tarsius is certainly held and
pierced when he inserts his head to see what there is in the pitcher."
GoEBEL remarks of this idea that more study of the matter in the
habitat is required. In N. intermedia the spurs are interesting because
they are broad, and are quite obviously made up of a group of corru-
gations; they are not sharp and tooth-Hke, and could not act in the
manner described by Burbidge for N. hicalcarata. Yet so far as we
know, the latter shows no superiority over the former or over N.
Rafflesiana in the struggle for existence. /V. intermedia is a hybrid of
horneensis and Rafflesiana (the former parent is uncertain, Mac-
farlane). If this occurred in nature it would be doubtful if the
specialized tooth-Hke portion of the peristome could act adaptively as a
beginning for the condition seen in N. hicalcarata.

The several interpretations of the morphology of the Nepenthes
leaf, as resumed in part by Troll (1932, 1939), are the following:

1. The Hd is the lamina of the leaf, the rest is the petiole with
highly specialized regions, phyllodial at the base. This view is trace-
able to A. P. DE Candolle (1827). Among others Goebel took this
position in his earlier writings (1884). The recognition by Hooker
that the spur is the true organic apex of the leaf threw this out of
court. According to Bower, Goebel regarded the lid as only a part
of the lamina, the rest appearing in modified form as the pitcher,
tendril, etc.

2. Instead of regarding the laminar portion of the leaf as petiolar,
WuNSCHMANN (1872) preferred to see in it the "lower part of the leaf
blade", and therefore that the leaf is non-petiolate. The evidence
from development denies this.

J. The pitcher has arisen phylogenetically as an apical gland,
which through enlargement and specialization became the complex of
organs which we now know. This, Hooker's interpretation, was
based in part on embryological observations and by comparison with
such leaves as that of Flagellaria, Gloriosa which have a cirrhus, a
terminal tenuous apex serving as a tendril. Faivre held a somewhat
similar view that the pitcher arises in the elongated midrib. But the
spur is, as said above, the organic apex of the leaf (Hooker).

4. The leaf arises as a peltate one. According to this view the
pitcher is a peltate leaf in which the margin is contracted so that the
upper surface lines a hollow organ, the pitcher. Its outer surface is
the lower leaf surface. Dickson, receiving his impulse from Baillon's
examination of the embryology of the Sarracenia leaf, and impressed
by the analogy supplied by the interrupted leaf of Codiaeum sp., wrote
"it seems highly probable that in Nepenthes we have to deal with a
leaf, the lamina of which is interrupted in the middle of its course by
becoming reduced to a midrib and that, while the proximal portion of
the lamina retains its typical form of a flat expansion, the distal por-
tion becomes peltately expanded into a funnel or pitcher. " But
Troll, though conceding the outward resemblance, one which strikes
anyone who has made the comparison, even to the peltation of the

Francis E. Lloyd —60— Carnivorous Plant s

lower moiety of the blade with a similar condition found in N.^ clipeata
Dans., points out that the resemblance is but superficial, since the
Codiaeum leaf is petioled while the ''blade" of Nepenthes is more
probably an expansion of the leaf base (Blattgrund) to be compared
with the primary leaf of Pothos. Goebel also held the view that the
pitcher is a peltate leaf developed into a tubiform one, and compared
the pitcher of Nepenthes with that of Utricularia, which is also ter-
minal either to a single "leaf" {Polypompholyx, Utricularia Menziesii,
etc.), and has a lid (door) which springs laterally from the true apex of
the trap visible as such in some species, e.g. U. Welwitschii, or to a leaf

segment.

5. The leaf of Nepenthes is not simple but compound. According
to Bower the lid arises as a double organ, the two congenitally fused
(^_ 4) ^ and represents two leaflets. This was based on embryological
observations. Macfarlane went still further and claimed to be able
to analyze the whole leaf into "3 to 4 or 5 pair of leaflets", the basal
lamina, the wings on the ventral surface of the pitcher, the lobes of the
Kd (Bower), and one or two pairs of lateral appendages sometimes
occurring on the spur, which itself terminates the leaf. This idea goes
back to Ch. Morren (1838) (Goebel 1891) who regarded the leaf as
having fused foholes and the lid as a terminal leaflet. Goebel (1923)
remarked that this view might have been entertained if, in the circle of
relationship, plants with compound leaves were known.

6. Troll put forward the theory that the Nepenthes leaf is a com-
plete parallel to the ordinary foliage leaf consisting of a basal zone
(Blattgrund), a petiole, and blade which is the pitcher (Oberblatt)
disturbed, however, by a modification of the petiole whereby it is at-
tended by a displacement upwards of the edges of the leaf base to
become the wings of the pitcher. Such a displacement occurs in Syn-
gonium podophyllum, and I have shown (1914) that it occurs in Gos-
sypium in which the flower peduncle normally suffers displacement up
the internode above, bringing the flower into an unusual position.
More specifically, Troll sets forth that (7) the leaf base consists of a
clasping bottom leaf zone which is contracted briefly to reexpand to
form the conspicuous lamina, and which in some species extends at its
apex across the base of the tendril in total stipulation {N. clipeata,
and others). (2) The blade is differentiated into the petiole and true
leaf blade. The former takes the form of a tendril, the latter the
pitcher, the blade in peltate form. But here the relation between the
petiolar structure and the peltation does not behave so simply as in
simple peltate leaves, (j) The spur is unifacial (as in Pothos). Arber
(1941) questions this view. At its base, the edge of the blade grows
to form a transverse connection from which the lid arises. This again
is total stipulation.

The supporting evidence is now briefly stated. (/) In the first
place the tendril is of bifacial structure (Troll) {8 — 20), and not, as
C. P. de Candolle (1898) thought, unifacial. The arrangement of the
fibrovascular bundles is not concentric with respect to phloem and
xylem, since the wood faces ventrally in the ventral moiety of the
organ. I can confirm this. (2) Reexamining the embryology of the
leaf, it is clear that in the primary leaf (in seedlings) the thinned out

Chapter IV — 61 — Nepenthes

basal part is composed of two halves which separate above and now
appear as the wings on the adaxial pitcher wall to form a transverse
membrane below the rim (Hooker, Dickson, Macfarlane). When
the transverse connection is absent (which Goebel held to be the case),
there is often an indication of it in the presence of a row of gland-like
emergencies indicating such a connection (Stern observed such).
Macfarlane said that a transverse strand of the venation also is to be
taken as an indication, but I cannot substantiate this (7 — 7, 9).
Hooker's view that the pitcher is "the hollowed out upper half of the
petiole" is discarded, and Dickson's theory of contracted peltate leaf
blade accepted. The earlier embryological condition is now examined.
In an early stage, when the leaf appears as a low conical structure,
there is a pit just below the apex on the adaxial side. Just below it is a
transverse weal, the transverse connection of the edges of the leaf base.
The leaf blade, it is important to note, arises on the abaxial side of
the leaf base, the latter, as in Iris, presenting total stipulation. The
blade cannot therefore be an extension of the apex of the stipule, but
though near it must arise below, abaxially. If without further differ-
entiation this embryonic stage passes into permanent form, a primary
leaf results, in which the pitcher stands in a dorsal position. What
authors have designated the blade is therefore only the leaf base
showing total stipulation, of which the transverse sector, as already
said, may be suppressed. In support of this I may point out that the
extent of the pitcher wings is not commensurate with that of the
veins beneath them, the wings often extending beyond the venation,
which swerves away to pass around the mouth of the pitcher. This
in the adult leaf. In intermediate forms, the development of the
rudiments proceeds further, especially the tendril, by contraction be-
low the pitcher. Nevertheless the wings of the pitcher pass down
along the edges of the tendril. In purely adult forms the tendril be-
comes entirely wingless. Troll now asks: (7) May the tendril be re-
garded as the petiole of the leaf between the pitcher as blade and
the leaf base? (2) How are the wings of the pitcher to be understood?
To answer these he analyzes the embryonic condition. In this a peti-
ole is not recognizable as such, but assuming that it must be there, he
postulates a zone of tissue, broad abaxially and narrow or absent
adaxially, the narrow adaxial edge of this wedge of tissue impinging
on the leaf base at its transverse weal (Wulst). The elongation of
this petiolar zone meets, however, an impediment in the leaf base tis-
sues, which converge below the mouth depression. In consequence, the
leaf base is dragged out along with the petiole and adaxial side of the
pitcher up to the edge of the mouth (but not quite, it may be added).
The whole adaxial side of the young leaf from the leaf base to the
mouth (I should say not quite) belongs to the leaf base and one may
come to the view that the tendril is an extension of the leaf base as
Goebel showed to be the case for the fan-palms. Nevertheless Troll
insists that the tendril is a petiole, though it may in some instances
(such as ;V. clipeata) have an unifacial structure in the lower portion.
But the leaf base is never unifacial, always bifacial. But where the
tendril is bifacial it should be regarded not as entirely independent
indeed, but concrescent with the leaf base.

Francis E. Lloyd — 62 — Carnivorous Plants

As to the pitcher wings, which show a wide variety of definitive
development, they may be considered as secondary outgrowths, like
those of Cephalotus or, as Goebel held, hke the keel of Sarracenia.
Others have held them to be leaf margins. Troll comes to the con-
clusion that they are the edges of the leaf base dragged out (ver-
schleppte), while growing themselves, by the growing petiole and leaf
beneath. Concerning the lid, its interpretation, before Hooker rec-
ognized the spur as the true apex of the pitcher leaf, was easy, as
being the true apex. Stern 's suggestion that it arises by a longitudinal
splitting of the apical meristem is untenable in view of the anatomical
facts. The views of Bower, Macfarlane and Goebel are also dis-
carded. The key to the problem, says Troll, is to be found in the
structure of the spur, which is unifacial, from which it follows that
the edges of the leaf blade at its base run together and unite (total
stipulation). Important here is a fact, pointed out by Heide (1910)
that the inner (lower) face of the lid is anatomically identical with that
of the interior of the pitcher, and the upper (outer) with that of the
outer pitcher surface. The lid cannot therefore be an "outgrowth of
the upper surface" as Goebel held. It should here be noted that
Dickson stated (and truly) that the base of the lid in primary leaves
(as also in other juvenile leaves) is very broad, extending "around
fully one half of the orifice of the pitcher" (7 — 7). Troll's view as
just stated is certainly supported by an examination of the venation of
even old adult pitchers in which the isthmus between the orifice and
the lid is very narrow. A macerated preparation of A^. formosa demon-
strates this, by which it is seen that, as already indicated in discussing
primary leaves, the venation is but that of a totally stipulate leaf
blade, sharply constricted below the apex. The apical portion, the lid,
may in adult leaves be supplied with a midvein which is secondary
since in primary leaves such a midvein does not exist. And when pres-
ent, as it is in adult leaves, it is evidently smaller and is dominated
by the lateral veins.

A novel interpretation of the rim, lid and spur has been advanced
by Mrs. Arber (1941). In doing this she rejects all earlier views, that
of Troll included, which hold that the lid is a transversal pinna. If
Troll is right, she says, the veins of the fid should have their wood
upwards, not downwards. She questions also the statement of Troll
that the spur is unifacial, though admitting that the veins of the spur
"tend toward a radial arrangement." Had Troll selected A'^. intermedia
and/or A^. hicalcarata for study, his evidence would have been still
more convincing. Having disposed of the spur as the leaf apex, Mrs.
Arber argues that "both the lid and the median point are merely
localized expressions of collar-forming activity, which is responsible for
the double curve-over of the aperture edge .... the lid, which is
turned down in youth, corresponding to the inner curve-over, and the
median point to the outer curve-over."

"The relative hypertrophy of the lid and median point may be
correlated with the special character of the venation .... of the
parallel type as in other pitchers. The midrib passes directly to the
junction of the hd and median point, while the veins of the adaxial
part of the pitcher also show a strong tendency to converge upon the

Chapter IV

63

Nepenthes

apical region. The median point and the Hd can thus draw upon a
richer vascular supply than the rest of the collar, which is entered
only by minor lateral veins, and thus overgrowth of the median region
may be stimulated."

It may be answered (/) that the midrib vein enters and traverses
the spur to its tip (7 — 9, 10; Text fig. 2). (2) The Kd cannot be
regarded as the inner "curve-over" since the surface of the rim would
then be a part of the outer pitcher surface, which the histology of the
rim denies. The "inner curve-over" would then have to be sought as
an outgrowth of the under surface of the lid, and that does not exist,
(j) The vascular system of the lid, assuming its origin as a transverse
weal, along the pitcher edge (Troll), is as it should be. {4) The
anatomy of the spur shows it to be the organic apex of the leaf

Fig. 2. — Nepenthes (various species). — i, Venation of lid and spur of a pitcher
I cm long; 2, of a pitcher 2 cm long; 3, of a pitcher 2.5 cm long; 4, of a full sized pitcher;
the veins (dotted lines) lie at a different level and more ventral to the rest (solid lines);
5, Section of pitcher wall just below the insertion of the spur in N. intermedia; 6, Section
through the spur of N. bicalcarata.

(Hooker), this being supported by additional evidence here from
N. intermedia and N. bicalcarata (Text fig. 2). (5) The wide dis-
placement of Hd and spur in these and other species is not accounted
for.

Histology of the peristome or rim. — • If we examine into the minute
anatomy of the hard, glossy surface tissue of the peristome we find
that it is composed of straight rows of cells, running across following
the transverse curve. In each row the cells overlap very much, in one
direction, the tapering tail of one cell overlapping the next and forming
a sharp ridge along it {8 — 7). The rows being straight, the cells
not imbricated as in the other pitcher plants, the ridges of successive
cells overlap the one over the other, to form a single sharp ridge,
about 0.017 mm. from its parallel neighbor. The general surface is also
formed into sulci separated by sharp secondary ridges about 0.17 mm.

Francis E. Lloyd —64— Carnivorous Plants

apart, there being about 10-12 rows of cells to each sulcus (N. am-
pullaria) {8—17). Whether the very large ridges that occur in TV.
villosa are secondary, or of the third order I cannot say, as I have
had no opportunity of examining the plant. The pubhshed drawing of
Hooker (1859) suggests the former.

The epidermis seen in a transverse section is complicated and re-
quires elucidation. One may see a row of cells equal in size or larger
cells separated by a pair of smaller ones (8 — 18). The latter are the
backward extensions of two cells which straddle the large one between
them. Two small cells, one on each side of the larger one, are therefore
really the backward extensions of a single cell. Atop each large cell
there is a central projection of various dimensions. This is the over-
lapping point of another neighbor cell, and appears as a solid mass of
cellulose, or with a lumen, according to the position of the section.
It is evident that the ridge is composed of the continuity of overlaps
(Heide 1910). N. Lowii presents a different appearance {8 — 12). The
overlapping spur is not lengthened so that no sharp ridge can be seen in
transverse sections. Only where the secondary ridges occur do the cells
give indication of striae; these not as well marked as in N. ampullaria.
With regard to these details Macfarlane's account (1908) is inadequate.

The ridges of the second order of magnitude, those readily seen
by the naked eye, end at the inner edge of the peristome in more or less
prominent teeth. When these are definite and prominent there can
be seen between them re-entrant bays marking the marginal pits, at
the bottom of which lie the flask-shaped glands first observed by Hunt
(1874), further studied by Dickson (1883) and called by him "mar-
ginal glands." The conformation of the bays is such as to afford a
seat for sustaining a large drop of nectar in position to attract insects
to the peril of falling into the pitcher.

The secondary ridges of N. Lowii are very low and not conspicuous
enough to catch the unaided eye except where, at their inner extrem-
ities, they become more elevated and end in a tooth beneath which
rests the large nectar gland. In N. ijiermis a few low ridges converging
on the broad tooth overhanging the gland may be seen. That it is
true that the general surface of the peristome affords a precarious foot-
hold for insects, ants at least, is as I have already said, doubtful.
Knoll found that they can use their footpads, for which, in spite of
the minute ridges, the surface is sufficiently smooth.

Histology of the glaitds. — Brongniart (1824) was the first to notice
the glandular character of the inner surface of the Nepenthes pitcher.
Treviranus, Meyen (1837) and Korthals (1839) recognized the glands
but thought that they were subepidermal, an error corrected by
Oudemans (1864).

The pitchers of Nepenthes are conspicuously supplied with glands,
those which serve to attract prey, the alluring glands, and those which
secrete the fluid of the pitcher, which is digestive. The alluring glands
are to be found on the under surface of the lid {8 — 8) and
between the teeth of the inner edge of the peristome {8 — 13). The
former are usually dished, biscuit-shaped, sessile glands resting in deep-
ish depressions. Some of these glands, in shallower depressions, are
to be found in the invagination near the apex of the hd in N. Tiveyi,

Chapter IV — 65 — Nepenthes

suggesting that the pocket may serve to hold a drop of nectar when
the pitcher is in active condition. In this species also, and in others
perhaps, in which a strong ridge stands on the median line on the
under surface of the lid, there occur on this ridge a number of nectar
glands, deeply enough sunken so that the surrounding rim makes a
distinct duct (8 — i6). The gland tissues are limited by a course of cells
with suberized radial walls. The most strikingly developed alluring glands
are to be found, as Macfarlane showed, distributed here and there
on the other leaf parts (midrib, tendril) serving to attract a wander-
ing population of ants which sooner or later find their way to the
pitcher. These glands are among the most highly developed struc-
turally in the plant kingdom, notably because of the deep duct {8 — 15).

Digestive glands occur on the inner surface of the pitcher wall in
great numbers — as many as 6000 per cm. in A^. stenophylla, as few as
100 in N. gracillinia (Danser).

Both nectar and digestive glands have the same structure. They
consist of a single course of deep columnar cells resting on two courses
of rounded cells, these lying in turn on a single course of cells having
their radial walls suberized, called by Macfarlane the "Hmiting"
layer, and being in strict continuity with the surrounding epidermis.
This indicates their origin which, according to Oudemans, Macfarlane
and Stern, is wholly epidermal, though Fenner has asserted that they
involve also the underlying parenchyma. His drawing is not convinc-
ing. As to the origin of the marginal nectar glands, these too have
been regarded by Macfarlane as of epidermal origin, but Stern has
maintained that they have two centers of origin, the deeper portion
of the gland being of mesophyll, and only the upper portion of epider-
mal origin. I have examined N. ampullaria {8 — 13), the species that
Stern worked with, and the evidence favors a doctrine of uniformity,
that they are of wholly epidermal origin. The presence of the limiting
layer seems to be decisive evidence.

Anatomy of the pitcher wall. — The wall of the pitcher is thin but
of great strength, attributable chiefly to the thick- walled epidermis
both within and without, supported by the veins which have a gen-
erous supply of sclerenchyma. The most interesting feature of the
wall anatomy is the occurrence of large idioblasts with spirally thick-
ened walls first seen by Unger in Nepenthes (according to Man-
gust 1882). These are very large spindle- or rod-shaped cells with
clear contents, apparently merely sap, and multispiral wall thicken-
ings. These, when the tissues are cut or torn, are drawn out as long
cottony conspicuous thread. The natural expectation that these pe-
culiar cells are connected with the vascular tissue system is not real-
ized (GiLBURT 1 881) as they do not stand in any relation to, and are
not at any point in contact with it.

Similar cells occur in some if not all species of Crinum (Mangin);
also in some orchids {Pleurothallus, Bulbophyllufn) (Trecul, through
Mangin); and in Salicornia (Duval-Jouve 1868). Mangin con-
sidered them as organs of support; and it is quite possible that they
contribute to the walls of the pitcher a considerable degree of mechani-
cal strength which they certainly display. In Dischidia the walls of the
pitchers have in analogous situations sclerenchyma fibers. Duval-

Francis E. Lloyd —66— Carnivorous Plants

JouvE thought them to be organs of aeration, and that they were al-
ways in contact with sub-stomatal cavities, which is surely not the
case. I have satisfied myself that they are quite independent of all
other cells than those of the parenchyma in which they lie. They
occur elsewhere than in the pitchers. It is probable that they are more
properly to be regarded as water-reservoirs (Kny and Zimmermann

1885).

The vascular system. — The course of the vascular strands is such as
to indicate that the pitcher is produced by the expansion chiefly of the
abaxial moiety of the leaf, and this is also indicated by the mutual
approximation of the wings along the edges of the ventral surface (Mac-
farlane). The finer endings of the vascular tissue often but not
always (Macfarlane) abut on the under side of the surface
glands found on the interior surface of the pitcher and of the lid. The
fact that unopened pitchers which have been removed from the plant
soon lose their juice (invariably found in young pitchers before open-
ing) observed by de Zeeuw (1934) seems to be related to this fact.

Surface anatomy. — By this we mean the anatomy of the epidermis,
that of the interior surface of the pitcher being of primary interest to
us. Examination of the interior of the pitcher {4 — 6) will show that,
with some exceptions (A^. ampullaria, hicalcarata, ventricosa, inermis)
there is a broad zone, beginning just beneath the rim, having a glau-
cous, opalescent appearance caused by an ample waxy secretion with
a pebbly surface. The epidermal cells here are simply polygonal with
the exception of a large number of slightly projecting lunate ones, so
placed that their concave edges are turned downwards {8 — 5). They
have the appearance, at once perceived, of half stomata, each in itself
looking like a guard cell. Oudemans (1864) thought them to be wax-
secreting glands. WuNSCHMANN would have none of this (1872) and
pronounced them to be squat hairs, broader than long. Dickson
(1883) was the first to arrive at the correct interpretation: "I have
here to note that each crescentic ledge consists of a semilunar cell
which overlaps a lower and smaller one. Occasionally these two cells
puzzlingly resemble deformed stomata," he wrote. His sometime
associate Macfarlane confirmed this, as did Haberlandt, independ-
ently, and later Bobisut (1910) showed that they are completely non-
functional stomata, having no pore, though Macfarlane had thought
otherwise. Macfarlane thought, too, that they exude water; and
Goebel that they might serve for gas exchange (1891), neither of
which can be true in the absence of a pore. I (19336) have confirmed
Bobisut's observations. The lunate cell is one guard cell, projecting
somewhat above the general level of the surface, hiding beneath itself
the second guard cell {8 — 6) , the whole having been rotated on the
longer axis. The whole waxy zone is a "conductive" (Hooker) or
slippery surface (Gleitzone, Goebel) on which insects such as ants
can find no foothold.

It is interesting to note in this connection that Macbrlde, in 181 7,
made the suggestion that the inabiUty of insects to cling to the surface
of the pitcher of Sarracenia adunca might be due to the presence of
an impalpable powder, or to the breaking away of fine hairs. To this
question in relation to Nepenthes Knoll (1914) has directed some
painstaking experimentation.

Chapter IV — 67 — Nepenthes

Knoll found that if he placed an ant on a cleaned surface of an
iris leaf {Iris pallida), the waxy secretion thus being locally removed,
and then placed the leaf in a vertical position, the ant could not get
away from the smooth, clean part. It seems that the ant clings to
smooth surfaces by means of its foot-pads, not by its claws, since there
is no roughness available. It cannot cling to the glaucous surface of
the Iris leaf, however, because the waxy secretion is loose and pulls off,
cumbering the foot-pads so that the ant must stop to clean them be-
fore they are again useful. This Knoll proved experimentally by
seeing if an ant can walk on a smooth surface as of glass when it has
been coated with a thin layer of a powder such as talc or carbon and
found that it cannot do so. Since the ant can walk on clean glass or
a smooth wax surface (beeswax melted onto a glass plate) it is quite
evident that the difficulty for the ant lies in the particles which come
off on his pads and prevent him from clinging. Experiments with the
loose waxy covering of the iris leaf first removing it and then applying
it again, showed the same result. Coming to the waxy zone of most
Nepenthes pitchers, Bobisut had already experimented and believed
to have found that ants could not climb the surface when in the verti-
cal position; even after he had (as he thought) removed the waxy
surface. Believing that he had failed to remove the waxy covering
perfectly, Knoll continued his experiments in the same sense as be-
fore with Iris, etc. He removed the wax thoroughly with chloroform,
rubbing downwards to avoid breaking the lunate cells and produced a
smooth green surface showing clearly the red markings, and upon this
he found that the insects could climb and run in any direction. When
now he scattered talc powder or wax powder obtained from the pitchers
themselves, they failed, showing that their ability to climb on the
smooth surface was due to the absence of a deterrent to the use of their
pads. He observed, however, that ants could readily negotiate the
gliding zone of older pitchers in greenhouses, and thought that this is
due to the removal of the wax by the vigorous sprinkling with water
which the plants usually receive, just as rain is known to remove the
waxy covering from plants like Cotyledon, etc. Knoll's observations
on the walking behavior of ants and the effectiveness of the waxy
zone as a precipitating mechanism have been repeated by my friend
Prof. W. KuppER and myself. The plant was a vigorously growing
one of N. gracillima (aff.?), one which is evidently very attractive to
ants as they are always to be seen in numbers rapidly walking hither
and yon especially about the tops of the pitchers. We observed that
they persistently visit the lid and the rim. They run no risk of capture
on the lid. On the rim, however, it is supposed that they do. As a
matter of fact, however, they do not, for they can walk on it in any
direction with rapidity, and they frequently stop to take the nectar
from the marginal glands. They even passed underneath the rim and
back several times in one excursion without danger. If, however, they
venture on to the waxy zone they at once display a quite different
behavior. They cannot then by any chance move rapidly forward.
If they progress at all, it is very slowly and with much groping with
the legs as if searching for a hold. Usually this ends in a complete
loss of foothold, and the ant falls into the abyss. One pitcher I ex-

Francis E. Lloyd — 68 — Carnivorous Plants

amined held a collection of ants which must have run into the thou-
sands. With regard to the ability of flies (houseflies and blue-bottles)
to retain a foothold on the rim, my friend Professor A. H. Reginald
BuLLER repeatedly observed many years ago that, in trying to
straddle the rim, they promptly fell into the pitcher, in N. Master-
si ana.

BoBisuT further thought that the curious deformed stomata could
furnish a foothold for the claws of the ants, etc. but Knoll showed
that the conformation, position and size of the ant's claws and of the
apparently available points for grasping with claws make them un-
available. From the ant's point of view the projecting guard cells
should have been turned the other way. Haberlandt thought that
they helped an insect to crawl downward but not upward, since they
afforded no foothold for the claws, but since the claws are not used,
but the pads only (Knoll), and since ants cannot climb downwards
any better than upwards on the surface, Knoll, not being able to
avoid the impression that the stomata are in some way connected
with trapping of insects, has advanced the following suggestion, namely,
that the numerous projecting guard cells serve, when the waxy surface
has more or less been removed by various means (rain, much traffic of
insects), as a means of joggling the body of the ant by the slipping of
a foot over them, somewhat as when, on climbing on a steep, precarious
rocky surface, a hand should slip from its hold of a ledge and slap the
rock surface just below. " Riitteleinrichtungen " Knoll calls the
projecting half-moon shaped cells, and regards them, briefly, as an
arrangement for hindering the climbing of the walls of the slipping
zone (Hooker's conducting zone). It must be remembered that an
ant uses its footpads and not the claws in trying to climb a smooth
surface. The frequent irregularities in form of the surface make it the
more perilous, according to Knoll. The theory is ingenious and may
very well represent the facts, which to Knoll are such in view of his
observations.

Below the slide or conducting zone, when present, the whole of the
remaining surface constitutes the detentive or digestive zone {4 — 6) .
It is a glossy green or red (A^. ventricosa) in color, and stands out in
sharp contrast with the glaucous color of the waxy zone above. The
surface is richly supplied with glands. Each gland stands in a slight
depression, the upper edge of which projects and overhangs the gland like
an eave, sometimes slightly, more often covering at least half the gland
{8 — 10, 11), or in the case of N. Pervillei (7 — 14) forming a deep pit.
In the depths of the pitcher, the glands often become more or less ir-
regular in shape and are devoid of any overhang (Macfarlane, Stern).

There seems to be every reason to regard these glands as both diges-
tive (or peptic as Macfarlane called them) and absorptive. Their ac-
tivity becomes evident long before the pitcher reaches its maturity,
young unopened pitchers always having the cavity half-filled with
fluid. Later a plentiful additional secretion occurs when organic, but
not so plentiful if inorganic materials are placed in the pitcher
(Hooker). That they are capable of reabsorbing the fluid is evident
in the fact that in a rather short time (24 hours or so) the fluid may
entirely disappear from unopened pitchers (de Zeeuw), and Goebel

Chapter IV — 69 — Nepenthes

showed that nitrogen, as ammonia and peptone, is rapidly reabsorbed
(1891).

Concerning the overhanging eave-like coverings of the glands,
Knoll argued that they serve to prevent the use of the gland for foot-
hold by insects, but incidentally prevent also damage by their claws to
the glands themselves.

Digestion. — The students of digestion in Nepenthes (as in other
insectivorous plants) have been divided into two camps {a) of those
who argued that it is a function of the plant itself carried out by the
secretion of an appropriate enzyme and {b) of those who have believed
it to be the result of bacterial action (decay or rotting, Dubois). If
the latter only takes place (as seems to be true in Darlingtonia, Hel-
iamphora, and perhaps some spp. of Sarracenia) this fact does not
disqualify these as carnivorous; bacterial action is an invariable ac-
companiment of some animal digestion {e.g. of cellulose in herbivores).
Bacterial action is often unavoidable in open pitchers and it has not always
been possible to separate the different digestive processes. Nepenthes
offers a special condition in that the pitchers secrete a quantity of fluid
before they open. The nature of this fluid was investigated by Voel-
KER (1849). He described it as hmpid and colorless, with a slight
agreeable odor and taste, and containing a non-volatile acid. The
total solids in percentage of the fluid ranged from 0.27 to 0.92 of
which 63.94% to 74.14% was non-volatile substances. Potassium,
sodium, magnesium, calcium, chlorine (as hydrochloric acid) and
organic acids were found, chiefly malic, with a little citric. Tait
found that pitcher fluid from unopened pitchers was sometimes acid,
but frequently not. When flies had found their way into open pitchers
the fluid became much more acid as well as more viscid. According
to VON GoRUP and Will (1876) the fluid is colorless, clear or slightly
opalescent, odorless, tasteless and of various consistency. After stimu-
lation the fluid changes from being neutral or only slightly acid, to
decidedly acid. "Miss R. Bok found that carefully washed beakers
of Nepenthes filled with distilled water did not show any acid pro-
duction while the addition of 2o/mgr./liter NH4CI would cause prompt
acid production. The pH went down to about 3.0 in 24 hours".
(Baas Becking, in ep.).

It is an important and well attested fact that the fluid of unopened
pitchers is above all free of bacteria, owing in part to the tight sealing
around the edge of the lid by interwoven branching hairs, a precursor
in Nature of the cotton plug used in bacteriological technique.

The pioneer work, constituting a prime stimulus to the investiga-
tion of digestion in carnivorous plants, was done by J. D. Hooker,
announced in his address before the Biological Section of the British
Association for the Advancement of Science in August 1874. Hooker
was in touch with Charles Darwin, and his interest was a natural
outcome of this contact; for Darwin was finishing his book on car-
nivorous plants at the time. Hooker found that bits of egg-white,
meat, fibrin and cartilage, when placed in the pitchers, showed un-
mistakable evidence in 24 hours of disintegration, but that this action
was by no means so pronounced in fluid placed in test tubes. From
this Hooker inferred that the digestion depends not on the first fluid

Francis E . Lloyd —70— Carnivorous Plants

secreted by the glands, but that there is a direct response to the
presence of the material to be digested. He saw evidence also of
antiseptic action in that odor was not developed so rapidly in the
pitcher fluid as in water. His general conclusion may be stated in his

own words: " it would appear probable that a substance acting

as a pepsine does is given off from the inner wall of the pitcher, but

chiefly after placing the animal matter in the acid fluid; " In

the following year (1875) Lawson Tate announced that he had suc-
ceeded in separating a substance "closely resembling pepsin" from
the secretion of Drosera dichotoma and a little later he obtained a sim-
ilar substance from the fluid taken from the pitchers of several species
of Nepenthes, but did not subject these extracts to the appropriate
tests. The preparations seem to have been glycerin extracts, in which
both were soluble. At the same time Rees and Will of Erlangen
(1875) made preparations of Drosera, drying the leaves with absolute
alcohol and extracting the ground material with glycerin. Such ex-
tracts, but only when slightly acidified with HCl (.2%), caused the
disappearance of swollen fibrin at 40 degrees in 18 hours, peptones
being produced, thus confirming the work of Darwin on Drosera. At
about the same time von Gorup-Besanez (1874) studied the fluid of
Nepenthes pitchers, and found that when he subjected shreds of fibrin
to the naturally acid secretion, they were nearly digested in an hour at
40 degrees, peptones then being present. Additional acid as above
accelerated the action.

Von Gorup and Will (1876) investigated further. They compared
the behavior of the fluid from stimulated pitchers (to which insects
had had access) with that from unstimulated pitchers. The former
was filtered and tested with acidulated fibrin, raw meat, coagulated
egg-white, legumin and gelatin, obtaining positive evidence in all cases
with the Biuret reaction, the gelatin excepted. This yielded a non-
jelHng gelatin-peptone. The fluid of unstimulated pitchers was found
to fail to act unless acidified, but responded in the presence of HCl,
formic, malic, citric, acetic and propionic acids. The efficiency of these
was various, formic acid being very active ("fast momentan"), followed
by malic, citric, acetic and propionic in the order named. The length
of time in which positive results were obtained, as indicated by the
Biuret reaction, varied from 10 minutes to three hours or more accord-
ing to the activity of the acid and the temperature.

Vines was busy at the same time. Following the method of Rees
and Will, he (1877) alcohol-dried pitcher walls bearing the glands of
Nepenthes and ground and extracted them with glycerin. In testing
his extracts he used the following method. In each of three test tubes
he placed (i) extract acidified; (2) extract only and (3) acid only, and
added a bit of swollen fibrin and kept the tubes at 40 degrees. Only
the first of the preparations gave a positive result and a peptone reac-
tion could be detected; the other two were negative. Vines noticed
that the pitcher fluid in von Gorup-Besanez' experiments appeared to
be more active than his own extracts. Following the lead of Ebstein
and Gruetzner and of Haidenhain (through Vlnes), who had ob-
tained more active extracts of animal glands by previous treat-
ment with acid. Vines then treated the pitcher wafls bearing glands

Chapter IV — 71 — Nepenthes

with 1% acetic acid for 24 hours, before extracting with glycerin, and
found that this extract was more powerful than that of the control
prepared without previous acid treatment. This indicated that, as in
the case of animal glands (Haidenhain) , the ferment exists in the
glands as a zymogen, a basic substance from which the ferment is
derived by acidification. The facts seemed to bring the whole phe-
nomenon of plant digestion into line with that in animals. This was
the beginning of a sustained investigation on the part of Vines on this
subject. Dubois and Tischutkin held that there is no digestion
proper to the Nepenthes pitcher, and that such digestion as takes place
is bacterial. Goebel's examination of the matter, however, afforded
experimental evidence in agreement with that of Vines (1877), who
now, however, repeated and extended his earlier work and drew the
conclusion that settled the matter to all appearances. For instance,
he showed that digestion goes on in the fluid of (unopened) pitchers
in the presence of poisons deadly to bacteria (HCN, thymol, KCN,
chloroform); but as opened pitchers were used the possibility is not
excluded that a bacterial ferment had already accumulated. Vines
concluded that the ferment present in the pitchers is secreted by the
pitcher glands, is not a product of bacteria, but is tryptic in na-
ture, like that of certain seeds (Green 1899) not producing pep-
tones, or if it does, these are broken down at once into other bodies
(leucine, etc.). It is remarkably stable and has an antiseptic action.
The pitcher liquid is usually distinctly acid, contrary to the prevaihng
views, the acidity therefore not depending on the supposed stimula-
tion by foreign bodies. In his third paper (1898) Vines showed more
in detail that the enzyme is unusually stable towards heat and alkalis,
for while exposure to these agencies does reduce its activity, "it re-
tains a sort of residual digestive power which asserts itself in a very
slow and prolonged digestion, and which can only be destroyed by
very strong measures." The enzyme exists in the tissues as a zymo-
gen, is essentially tryptic in character, and among its products of di-
gestion true peptones are present. In his last paper published in 1901,
Vines proposed the name "nepenthin" for the proteolytic ferment
which he had previously studied and made further tests of the action
of the pitcher fluid on fibrin and on Witte peptone, exposing them to
action for several days at 38.5 degrees C. with the addition of HCl
or citric acid. The results showed the presence of tryptophane, char-
acteristic of tryptic digestion.

The detail of Vines' general conclusions, that the digestion is
rather of the tryptic kind, was later called in question by Abderhalden
and Teruuchi (1906). From data obtained by experiments in which
glycyl-1-tyrosin was used, which gave negative results, they concluded
that the Nepenthes protease is not a trypsin, though they did not as-
sert certainty in view of the lack of sufficient material for further work
{See Stern and Stern, beyond).

Quite opposite conclusions were drawn by Tischutkin (1891),
Dubois (1890) and Couvreur (1900). Tischutkin placed small cubes
of egg-albumin in unopened pitchers by passing them through a small
window cut in the wall under sterile conditions, and saw no digestion.
When the test material was placed in pitcher fluid in vitro, digestion

Francis E. Lloyd —12— Carnivorous Plants

occurred after some days during which bacteria had accumulated.
Dubois (1890) found the sterile fluid from unopened pitchers without
action, but that from recently opened pitchers, while still clear, acted
vigorously on egg-albumin. Dubois voted for the bacterial action
theory. Couvreur (1900) argued that Vines' results were due to the
interaction of the reagents on one another. This totally negative
attitude had been combatted by Goebel (1893). In a prehminary ex-
periment, he took a pitcher of A^. paradisiaca (a hybrid) which contains
a "clear, odorless and tasteless fluid" and in it placed a bit of flbrin,
with one in water as control. In six days the fibrin was broken up
and bacteria were plentiful, and the fluid showed a neutral or sHghtly
alkaline reaction. A yellow reaction was obtained in the water but not
in the pitcher, by which the products had been resorbed. No peptone
had been produced. Cultures showed the presence of Bacterium
fiiiorescens liquejaciens. This result admittedly agreed with those of
Dubois and Tischutkin. But Goebel pointed out that the plant was
not normal. When he took a strong, wefl grown plant he found other-
wise. It had three pitchers, an old one, a strong vigorous one and an
unopened one. In the old one, a wasp was attacked and digested. In
three days the fluid was alkaline and bacteria and infusoria were plenti-
ful. In the open but vigorous pitcher a fly had been caught. A bit of
fibrin was introduced and was attacked in one hour. In 3 hours pep-
tone was demonstrable. Another bit of fibrin together with 0.2% HCl
were introduced, and this was digested in 40 minutes, and no bacteria
could be found. The fluid of the unopened pitcher was neutral. In
its fluid fibrin accompanied by 1% formic acid was digested in 12
hours, and no bacteria were detected after 8 days. He concluded
therefore that a peptone forming ferment was present in the fully
normal pitchers. He further showed that normal pitchers, when stim-
ulated by the presence of an insect, secrete formic acid. By way of
further control he tried to see if fibrin might be digested by the secre-
tions of the lid, with negative results. To do this he fastened a bit of
fibrin on the under side of the lid with moist filter paper. Thus
Goebel confirmed Vines' conclusions. In general support of the view
that the bacteria of decay have nothing to do with the digestion of
insects in normal plants in their native habitats Goebel quoted
Wallace who wrote in The Malay Archipelago as foflows: "We had
been told that we should find water at Padangbatu, but we looked for
it in vain, as we were exceedingly thirsty. At last we turned to the
pitcher plants, but the water contained in the pitchers (about half a
pint in each) was full of insects and otherwise uninviting. On tasting
it, however, we found it very palatable, though rather warm, and we
all quenched our thirst from these natural jugs." And stiU earlier
Hermann Nicolaus Grimm recorded (in 1682) the discovery of "aqua
dulcis, limpida, amabihs, confortans et frigida" in the pitchers, and the
fluid from six to eight of them was sufficient to satisfy a thirsty person.
That our greenhouse cultivated plants, because of their compara-
tively feeble vitality as compared with plants in their native habitats,
may often behave abnormally, is indicated by the observation of
MoHNiKE, whom Goebel cites, who said that the pitcher almost al-
ways contains a mass of dead insects including even large beetles.

Chapter IV —73— Nepenthes

The larvae of Apogonia spherica were found entire but quite digested
internally. Insects die in the pitcher fluid much more quickly than in
distilled water. In 48 hours or so, insects are disintegrated, only their
chitinous skeletons remaining. Such statements, encountered in other
writings, indicate a very vigorous action. Goebel ventured the state-
ment that of all the pitchered carnivorous plants Nepenthes is the most
vigorous in these matters.

Clautriau (i 899-1 900) took the opportunity of studying Nepen-
thes in its habitat in Java. His results fully corroborate in general
Goebel and Vines. He observes:

While the fluid in unstimulated pitchers is neutral, it becomes acid
on the introduction of foreign bodies. Even shaking has this effect,
the strongest acidity obtained being equal to that of a Uter of water
acidified with 2 cc. of fuming HCl. In the fluid there is a thermolabile
substance which acts as a wetting agent, so that insects are quickly
drowned but are not killed by any poison. Insects are digested with-
out any putrefaction. Antiseptics such as formaldehyde, chloroform,
etc. inhibit both the secretion of acid and digestion, and the pitchers
presently die. On the introduction of egg-white, both digestion and
absorption occurred. If a small quantity was used absorption equalled
digestion in rate; if a too large quantity was used, the products re-
mained in quantity sufficient to afford a culture medium for bacteria.
Quantitative experiments showed that 5 cc. of egg-white (10 cc. to 90
cc. water) is completely digested in vigorous pitchers in 2 days. If a
pitcher were separated from the plant, digestion was inhibited, and
Clautriau usually found that in vitro experiments gave negative re-
sults. At home in Brussels he showed by refined methods that al-
bumin is completely digested to peptone. This is readily absorbed by
the pitcher walls, so that he was able to give successive doses of food
(albumin) and see that they were digested perfectly by the pitchers of
N. M aster siana.

Clautriau concluded that the enzyme is a true pepsin as it acts
only in an acid medium and produces true peptone as an end result.
No other products could be found. No amylase was detected. The
evidence indicated that an ample secretion of both enzyme and acids
required stimulation, and, on microchemical evidence, that peptone is
absorbed by the glands and stored as protein. A superabundance of
food may allow the play of bacteria, and the products of their activity
(amino acids and ammonia) may be used by the plant. These do not
necessarily damage the pitcher itself.

Fenner has (1904) advanced an interesting presentation of what
he believes goes on in natural conditions. The original pitcher fluid is
slightly acid (formic acid, Goebel). If a few gnats are introduced,
they float on top of the fluid. If alive they endeavor to escape by
cKmbing up the wall, and in this way they come in contact with the
glands below their overhanging eaves, which, Haberlandt has sug-
gested, serve the purpose of retaining fluid by capillarity. ^ The body
of an insect wet with pitcher fluid thus applied serves to stimulate the
glands to action, when they secrete a highly viscid, active fluid which
attacks the insect so vigorously that it is digested in 5-8 hours.
Tenner tested this view experimentally by taking an opened pitcher

Francis E. Lloyd —74— Carnivorous Plants

and placing an insect (a gnat) on an area of the wall which had been
dried. A slight amount of secretion then occurs which is insufficient to
act and readtly dries up. But if an insect wet with pitcher fluid is used,
an ample secretion from the gland ensues and the insect is digested in
the time indicated above. It would appear according to Fenner's in-
terpretation that the pitcher fluid acts as a stimulant to secretion. In
this way the body of a smaU insect comes into contact with a more
vigorous secretion. The greater activity, therefore, is not within the
depths of the pitcher fluid but in the films of fluid by which the bodies
of the insect adhere to the glands. Into this position they come nat-
urally enough since they float towards the walls, and the fluid level,
by shaking (as by the wind), is moved so that insects stick on the

walls above it.

The collection of Nepenthes accumulated at the University of
Pennsylvania by Professor Macfarlane, furnished an abundant amount
of material for the study of proteolysis by Hepburn (191 9), who car-
ried out his experiments with unopened pitchers, and opened pitchers
from which insects were excluded by means of cotton wool plugs. Some
of these were stimulated by the introduction of glass beads after shak-
ing. A distinction between "stimulated" and "unstimulated" pitchers
became evident: Their fluid was found to differ in its activity. Bac-
teria were carefully excluded by means of active bactericides, and all
experiments were controlled. With various substrates (ovalbumin,
fibrin, ovomucoid, Heyden's nutrient and Witte peptone) and by
means of formol titration (Sorensen) he found that the fluid from
stimulated pitchers digested all of them; but not that of unstimulated.
In the presence of very dilute HCl edestin was also acted upon by
fluid of stimulated but not by that of unstimulated pitchers. Carmine
fibrin in the presence of acid was digested by both, but not by that of
unstimulated pitchers in the absence of acid. Protean (from the globu-
lin of the seed of castor bean, Ricinus communis) and ricin were
attacked by the fluid of both stimulated and unstimulated pitchers if
in the presence of very dilute acid. With sufficiently long exposure,
glycyltryptophane was "apparently" hydrolysed by the fluid of
stimulated pitchers. It appeared that the fluid of stimulated pitchers
possessed proteolytic power in the absence of acid (as weU as with
acid) while that of unstimulated pitchers always required the ad-
dition of acid. It is not clear how the stimulation acts: whether by a
change of acidity creating a favorable medium for an enzyme already
present, or by the activation of a zymogen already present or by an
increase in the secretion of the protease of the glands.

In 1932 Stern and Stern reopened the question. They chose a
series of substrates (gelatin, casein, edestin, ovalbumin and serum
protein), and tested the effect of the pitcher secretion on them through-
out the whole physiological range of pH, and found that they obtained
two maxima, one at pH 4.7 and 7.0 for gelatin, pH 3 and 8 for edes-
tin, pH 4 and 8 for ovalbumin. Serum protein was not measurably
attacked between pH 1.5 and 8.4. The behavior of casein is anoma-
lous. The curve shows two maxima, at pH 3 and 5.5, with a deep dip
between, due probably to the flocking of the protein at the isoelectric
point and the binding of the enzyme. The tryptic optimum was not

Chapter IV —75— Nepenthes

evident, due possibly to the inhibiting effect of the glycerine present.
These results were obtained on pitcher secretion preserved with 50%
glycerine, from N. Hibherdii and N. mixta. The secretion from open
pitchers containing insects, mostly ants, was used. In order to exclude
the effect of microbes and the enzymes of insect bodies, the authors
also took the glandular walls, comminuted and extracted them with acetic-
glycerine. The extract they found active on gelatine at pH 8, and on
ovalbumin only in the region of pH 3-3.5, thus supplying evidence that
a tryptic ferment is secreted by the glands of the Nepenthes pitcher.
In order to compare the enzymes of Nepenthes with those of animals
they made tests of the effect on them of certain activators, known to
affect other proteinases, \\ath negative results. Neither HCN, H2S or
cystein have any effect on the proteinase, nor does enterokinase on the
tryptase; the latter Stern had shown for the proteinase of white
blood cells.

The conclusions of Stern and Stern, that there are two enzymes
present, a catheptic and a tryptic, and that the latter is not attribut-
able to the presence of bacteria, led W. de Kramer (1932) in Baas
Becking's laboratory at Leiden to re-examine the question. He came
to the conclusion that the opinion that the tryptic action is due to
bacteria is justified. De Zeeuw, who quotes de Kramer's unpubhshed
results, attacked this question. Both catheptic and tr>T>tic action was
found by them. De Zeeuw experimented with unopened pitchers
which were allowed to open under sterile conditions, using bromine
water and sterile cotton for insurance against bacterial infection, and
with unopened ones, which were always found sterile.

The fluid of unopened pitchers does not digest fibrin until an acid
is added, an enzyme is therefore present. It becomes active within
the pH range of 3.4 to 4.4, phosphoric, malic and citric acid having
been used, and a phosphate buffer. That from an aseptically opened
pitcher acted at pH 3.6 in phosphoric acid, while that from normally
opened pitchers was active at pH 3.2 with phosphoric acid and from
7.2 and 8.6 with phosphate buffer. The last named was not sterile.
Bacterium fluorescens liquefaciens, B. prodigiosum and two others were
present, and all of these were found to exert tryptic action. By way of
control the fluid of a pitcher, opened under sterile conditions, of
N. Morganiana, was tested and found to digest fibrin at pH 4.4 to 5.5,
the pH increasing steadily during 15 days. An acetic acid-glycerine
extract was found to digest fibrin at pH 2.3 to 4.2, in direct contra-
diction to the results of Stern and Stern (1932) who also believed
their extract to be bacteria-free.

Open pitchers display a wide range of pH (3.0-7.2), S3% reacting
neutral or basic, 36 pitchers being examined. When completely di-
gested insect cadavers were present, the fluid was neutral or weakly
basic; when digestion was in its early stages, acid. Into a pitcher
which showed an acid reaction (pH 3.0) the acid was neutralized by
means of hme water, and a pH of 8.2 established. Since digestion was
proceeding, the next morning the fluid was found to be at pH 3.0 again.
Pitchers after being washed out thoroughly with distilled water were
then supplied with distilled water (pH 7). When fibrin was added, the
pH dropped to 3.5, as also when egg-albumin (such as used by Clautriau)

Francis E. Lloyd —76— Carnivorous Plants

was used. This is interpreted as demonstrating that the addition of a
protein to the fluid stimulates the secretion of acid; but de Zeeuw was
unable to bring this about by mechanical stimulation, the contrary
having been reported by Hepburn {see above). The secretion of un-
opened pitchers had been found by de Kramer to be always neutral,
and this was re-examined by de Zeeuw who found the pH ranging
from 4.2 to 7 (ave. 6.6 ± 1.2) in October and from 4.2 to 4.8 (ave.
4.5 + 0.3) in November and December, a difference possibly attribut-
able to the time of year, with a lower temperature prevailing (in the
greenhouse?). The fluid of pitchers opened under sterile conditions,
therefore without chemical stimulation, always reacted acid (pH 4.2 to
5.8) but required additional acid to secure digestion. When acidified to
pH 3.0 to 3.5 with certain acids (phosphoric, HCl, formic, malic,
and succinic acid), and kept sterile with toluene, digestion proceeded,
but not with the others tried, which probably destroyed the enzyme.
What kind of acid is secreted by the pitcher, aside from the fact that it
is not a volatile one, was not determined. But the acid reaction of the
glands indicated that these are responsible. De Zeeuw therefore
reached the conclusion that the enzymes present are catheptic and
tryptic, but that the former only is present in sterile pitcher fluid, the
latter occurring only in opened pitchers to which bacteria had had
access. Acid is secreted by the gland when stimulated by chemical
but not by mechanical means.

As the matter stands at the present, therefore, the positive evidence
that a catheptic proteinase is secreted by the pitchers of Nepenthes is
conclusive. That tryptic digestion in the absence of bacteria takes
place there seems little doubt, but this cannot yet be said to be com-
pletely proven.

Antisepsis of pitcher fluid. — Reference has been made to the fact,
usually accepted as such, that the pitcher fluid of normal actively di-
gesting pitchers is free of bacterial action. Wallace has already been
quoted as testifying to this in the natural habitat in Borneo. Goebel
atributed this, in the experiments he conducted, to the presence of for-
mic acid secreted by the pitcher glands. Robinson (1908) observed
that meat extract might remain in the pitchers of N. destillatoria for
two weeks without the odor of foulness. Although they confirmed the
generally accepted belief that the fluid of unopened pitchers is sterile,
Hepburn et al. (1919, 1927) found that opened pitchers, whether
containing insects or not, invariably contained bacteria in large num-
bers, whose activity in digesting proteins they found was low, and that
they play only a secondary role in the digestion of insects, the leading
role being played by the protease proper to the pitcher itself. They
argued that the bacteria five in symbiosis with the plant, assisting some-
what in the digestion of insects, thereby drawing nutrition therefrom.
Since the plants they experimented with were under cultivation, the
argument that their results do not reflect the conditions found in
nature, as indicated by such experiences as Wallace, seems justified.
Testimony is, however, not uniform on this point. Jensen (1910) speaks
twice of the horrible stench arising from pitchers loaded with centi-
pedes, cockroaches, butterflies and a huge scorpion found in pitchers
near Tjibodas, Java. This may mean merely that the pitchers were

Chapter IV — 77 — Nepenthes

overloaded beyond the limits at which the antiseptic effect could be
expected to work. On the basis of experiments, Jensen regards it as
sure that certain larvae which live on the debris in pitchers have an
antiferment which is not possessed by the same kind of larvae when
inhabiting water in pools.

Under the title, ''The animal world of Nepenthes pitchers", August
Thienemann (1932) brought together all that at the time of publi-
cation was known about the fauna to be found in the pitchers of
Nepenthes. Long ago, as early as 1747, G. E. Rumphius, the renowned
explorer, remarked in his Herbarium Aniboinense (pt. 5, p. 122): —

"In aperto varii repunt vermicuK et insecta, quae in hoc moriuntur,
excepta parva quadam squilla gibba, quae aliquando in hoc reperitur

et vivit " Since that time innumerable observations have been

made and it would scarcely be profitable to detail them.

The first question which will occur to one interested in this fact is
one which Jensen (1910) asked, namely, how can animals live in the
digestive fluids of the pitchers. In answer he said that he beheved
there was indicated the presence of an antipepsin formed by the
animals in question. Dover (1928) agreed with him, but did not go so
far as to assert the presence of an antipepsin, though he beheved that
mosquito larvae do possess such, and suggested that the "presence of
neutral salts in the tissues of the larvae might possibly retard peptic
digestion;" Thienemann, however, maintained that there is no bind-
ing evidence that there is an antipepsin and goes further in saying that
he sees no special problem to be involved. The numerous internal
parasites of the animal body hve in body fluids rich in ferments.
Dover, himself, observed that the larvae of Megarhinus acaudatus can
remain alive in a very weak iodine and in a strong pepsin solution and
in the latter Kved some days, pupated and hatched out. Are we then
to expect if an antipepsin is present that there is also an antiiodine?
We may recall here that Hepburn and Jones (191 9) believe that they
demonstrated the presence of antiproteases in the larvae of Sarcophaga
which inhabit the pitchers of Sarracenia flava.

The inhabitants of the pitchers are divided by Thienemann into
three classes, (a) those which are occasionally found, but which belong
properly in other places (nepenthexene) ; {h) those which occur, find in
the pitcher suitable conditions and can pass their watery fives there
but which are not confined to them (nepenthephile) and thirdly those
which five only in the pitchers and are not found elsewhere (nepen-
thebionts). Since the pitchers are commonly only partly filled with
fluid, namely, ca. up to the waxy zone, there is a "terrestrial fauna"
as well as an aquatic fauna.

Of the former, aside from 2 species of leaf miners (which, however,
have been claimed to behave in relation to the water level) which are
questionably peculiar to Nepenthes pitchers, there are four spiders,
three of which are claimed to be nepenthebiont. The 4 species are
Misumenops nepenthicola, M. Thienemannii, Thomisus callidus and Th.
nepenthephilus. Th. callidus is nepenthephile; the others have been
found up tin the present only in pitchers of Nepenthes, but are not
confined to any one species. But they are excluded from .V. ampul-
laria because there is no w^axy zone, states Thienemann; they should

Francis E. Lloyd

78 — Carnivorous Plants

also be absent from N. ventricosa. Since the spiders above named find
their food in insects attracted to the pitchers, they may be regarded
as commensal. The case is somewhat if not quite the same as that of
the spider-plant combination of Roridula (Lloyd, 1934)-

The "aquatic fauna" nepenthexene forms include protozoa, myx-
ophyceae, desmids and diatoms, rotatoria, Oligochaetes, crustaceae
and also' larvae of various Diptera and a very occasional tadpole.
Such forms occur relatively infrequently, but are most abundant in
those pitchers of N. ampullaria which stand half buried in the sub-
stratum, as would be expected. The nepenthephile animals occur in
only very small numbers; only three known in fact. It is interesting
to know that of these one is represented by two races, one of which
lives in hollows of bamboos. The nepenthebionts include the remark-
able number of 26 species; of the Phoridae 6, Chironomidae i, and of
the Culicidae 19. All these are Diptera, 19 of which are mosquitos.
It is admitted that further research may reduce or enlarge this number
somewhat, but it can hardly alter the general weight of the evidence
that there is a strikingly large number of animals which habitually live
in the pitchers of Nepenthes and nowhere else. They feed on the ani-
mal detritus found there. To account for this large number of forms
adapted only to Nepenthes as commensals, Thienemann points out
that Danser refers the origin of the genus to a time earlier than
the beginning of the Tertiary, in the Chalk, but Danser thinks of the
genus as a young one.

Folklore, uses. — It is inevitable that such an unusual and curious
plant as Nepenthes should figure in the folklore of the peoples in con-
tact with it. In this connection I quote an interesting passage from
RuMPHius {Herbarium Amboinense 5:123) containing notes made
about 1660 in the Far East. This was kindly translated for me by
Prof. Baas Becking, who indeed drew my attention to it.

"Uses. This remarkable plant mostly serves as a curiosity, to keep its pitchers amongst other
strange objects which are worth keeping to show the nice playfulness of nature. To this end open
pitchers are preferred. They are emptied and wind-dried, filled with cotton or other fine material
in order that the natural form may be preserved. Or the dried pitchers are placed in a book and
pressed flat. However, to show the curiosity more completely, one should have the leaf still at-
tached.

"The natives are unwilling to bring them to us from the mountains, because of an old super-
stition according to which if one cuts off the pitchers and pours out the water one will meet with a
heavy rain before reaching home. As this happened a few times when I had ordered them to fetch
me the largest species from the mountains of Mamalo, they were strengthened in their superstition,
notwithstanding the fact that I convinced them that it had rained on the two days previous to this
expedition. Others go to the mountains when the rain has not fallen for a long time, and empty
all pitchers which they can reach with a stupid zeal as they want to bring rain to the land in this
way; but the converted natives do not dare to perform such tricks, out of respect to our and to the
Mohammedan priests.

"Now listen to the contrary effect. If children often wet the bed, the native goes to the moun-
tains and fetches a few of the filled and still unopened {sic) pitchers, the water of which he pours over
the head of the children and makes them drink of it, as they also do to adults who are unable to
keep their water.

"As it seems, one or the other must be a lie or a great miracle, if one could by means of this
little pitcher draw the water from the heavens and also keep it in the children's bellies."

At a guess, the virtue attributed by the natives to the open pitchers,
out of which water can be poured, and the unopened pitcher, lies fun-
damentally in the fact that the latter holds its water. The symboHsm
appears evident.

Chapter IV — 79 — Nepenthes

B. H. Danser (1927) remarks that no trace of these superstitions
is to be found nowadays, but that the Malayans from Malacca and the
Riouw Archipelago use the fluid from the unopened pitchers to wash
their eyes or put it on inflamed skin until the new skin is formed.

He points out also that the long viney stems (lianas) of N. ampul-
laria are used as ropes for slinging foot-bridges. Possibly other species
are similarly used.

Literature Cited:

Abderhalden & Teruuchi, Zeitschr. f. Physiol. Chem. 49:21-25, igo6.

Arber (see under Cephalolus).

BiscHOFF, G. W., Lehrbuch der allgemeinen Botanik. Berlin, 1834 {through Troll).

BoBisuT, O., tjber den Functionswechsel der Spaltoffnungen in der Gleitzone der Nepenthes-

Kannen. Sitzungsber. d. K. Akad. d. Wiss., Math.-Naturwiss. Klasse 119:3-10, 1910.
Bower, F. O., On the pitcher of Nepenthes. Ann. Bot. 3:239-252, 1889.
Bower, F. O., On Dr. Macfarlane's observations on pitchered insectivorous plants. Ann.

Bot. 4:165, 1889.
Brongniart, Ad., Ann. Sci. Nat. 1:29, 1824.
Burbidge, F. W. T., Card. Chron. 1880:201.

DE Candolle, a. p., Organographie vegetale. Paris, 1827. {through Goebel and Troll).
DE Candolle, C. P., Sur les feuilles peltees. Bull. Trav. Soc. Bot. Geneve. 9:1, 1898/9.

Clautriau, G., Memoires couronnes Acad. R. d. Sci. etc. Belg. 59:1-55, 1900.

CouvREUR, C. R. 130:848-849, 1900.

Curtis, W., Botanical Magazine 53 (no. 2629), 1826.

CzAPEK, F., Biochemie der Pflanzen, 3 vols., 825 pp., Jena, 1925.

Danser, B. H., Indische bekerplanten. De Tropische Natuur 1927:197-205.

Danser, B. H., The Nepenthaceae of the Netherlands Indies. Bull. Jard. bot. Buit. ser. 3,

9:249-438, 1928.
Dickson, A., On the structure of the pitcher in the seedling of Nepenthes as compared

with that of the adult plant. Proc. R. Soc. Edin. 1883/4:381-385.
Dickson, A., Gard. Chron. n. s. 20:812, 1883.

Dover, Cedric, Fauna of pitcher plants. Journ. Malayan Br. R. As. Soc. 6:1-27, 1928.
Dubois, R., Sur le pretendu pouvoir digestif du liquide de I'urne des Nepenthes. C. R. in:

315-317, 1890.
Duval-Jouve, J., {re Spirally thickened cells in Salicornia). Bull. Soc. Bot. France 15,

1868 (through Mangin).
Faivre, E., Recherches sur la structure des urnes chez Nepenthes. Mem. de I'acad.

d. sci., belles lettres et arts de Lyon 22:173-211, 1877.
Fenner, C. a., Beitrag zur Kenntnis der Anatomie, Entwickelungsgeschichte und Biologie

der Laubblatter und Drusen einiger Insectivoren. Flora 93:335-434, 1904-
GiLBURT, W. H., Notes on the histology of pitcher plants. Quekett Micr. Journ. 6:151-

164, i88r.
GoEBEL, K., Pflanzenbiologische Schilderungen II. Marburg 189 1. {Nepenthes: -p^.g^)^ 186).
Goebel, K., Organographie, 2d. ed., pt. 3, Jena 1923.
Gorup-Besanez E. von, tJber das Vorkommen eines diastatischen und peptonbildenden

Ferments in den Wickensamen. Ber. d. deutsch. Chem. Gesellsch. 7:1478-1480, 1874

(Not directly concerned with carnivorous plants).
Gorup-Besanez, E. von, Weitere Beobachtungen iiber diastatische und peptonbildende

Fermente im Pflanzenreiche. Ber. d. deutsch. Chem. Gesellsch. 8:1510-1514, 1875

(Not directly concerned with carnivorous plants).
Gorup-Besanez, E. von & H. Will, Fortgesetzte Beobachtungen iiber peptonbildende

Fermente im Pflanzenreiche. Ber. d. deutsch. Chem. Gesellsch. 9:673-678, 1876.
Green, J. R., The soluble ferments and fermentation, 1899.
Grimm, H. N., De planta mirabili destillatoria in Miscell. nat. curios. Dec. II, ann. I,

1682 {through Wunschmann).
Guenther, Die lebenden Bewohner der Kannen der insectenfressenden Pflanzen, N. des-
tillatoria auf Ceylon. Zeitschr. wiss. Insectenbiologie 9:123, 1913.
Haberlandt, G., Physiologische Pflanzenanatomie, 2. Aufl., Leipzig 1924 {Nepenthes, p. 442).
Heide, F., Observations on the corrugated rim of Nepenthes. Bot. Tidsskrift 30:133-

147, 1 9 10 (The cover is dated 1909).
Heinricher, E., Zur Biologie von Nepenthes, etc. Ann. Jard. Buit. 20:277-298, 1906.
Hepburn, J. S., Biochemical studies of the pitcher liquor of Nepenthes. Proc. km.

Phil. Soc. 57:112-129, 1918.
Hepburn, J. S., E. Q. St. John & F. M. Jones, Biochemical studies of insectivorous plants.

Contr. Bot. Lab. U. of Penna. 4:419-463, 1919. — See also p. 39.
Hooker, J. D., On the origin and development of the pitcher of Nepenthes, with an ac-
count of some new Bornean plants of the genus. Trans. Linn. Soc. 22:415-424, 1859.
Hooker, J. D., Address to the Department of Zoology and Botany, B. A. A. S. Report of

the forty-fourth meeting, 1874:102-116, 1875.

Francis E. Lloyd — 80 — Carnivorous Plants

Hunt, J. Gibbons, (A minute in) Proc. Acad. Nat. Sci. Phila. 26:144, 1874.

Jensen, H., NepenlheS'TieTe, II. Biologische Notizen. Ann. du jard. Buitenzorg, Sup.

Ill: 941-946, 1910.
Knoll, F., tJber die Ursache des Ausgleitens der Insektenbeine an Wachsbedeckten Pflan-

zentheilen. Jahrb. wiss. Botan. 54:448-497, 1914.
Kny, L. and A. Zimmermann, Die Bedeutung der Spiralzellen von Nepenthes. Bet. d. d.

bot. Gesellsch. 3, 1885.
KoRTHALS, p. W., Nepenthes, in Verb. d. nat. Geschiedenis der Nederl. Overzeesche

Bezittingen (Botanie), Leiden, 1839/1842, Ed. by C. J. Temminck.
Lloyd, F. E., Abscission. Ottawa Naturalist 38:41-52; 61-75, iQM-
Lloyd, 1933&, see p. 269; 1934, see p. 8.
Macbride, J. M., On the power of Sarracenia adimca to entrap insects. (Read in 1815)

Trans. Linn. Soc. London 12:48-52, 181 7.
Macfarlane, J. M., Nature, Dec. 25, 1884.
Macfarlane, J. M., Observations on the pitchered insectivorous plants, I. Ann. Bot.

3:253-266, 1889/90; IL 7:403-458, 1893.
Macfarlane, J. M., Nepenthaceae. Das Pflanzenreich, Leipzig, 1908.
Mangin, L., Sur le developpement des cellules spiralees. Bull. Soc. bot. France 29:14-17, 1882.
Menzel, R., Beitrage zur Kenntnis der Mikroflora vom Niederlandischen Ost-Indien; II.

tJber den tierischen Inhalt der Kannen von N. melamphora Reinw. mit bes. Beriicksich-

tigung der Nematoden. Treubia 3:116-122 (Doubts that the pitchers are a mere

"Luxus-Anpassung"). Harpacticiden als Bromeliaceen-Bewohner. Ihid. 3:122-126, 1923.
Meyen, F. J. F., tJber die Sekretionsorganen der Pflanzen. Berlin, 1837.
Mohnike, Blicke auf das Pflanzen- und Thierleben in den niederlandischen Malaienlan-

dern. 1883 (p. 148).
Morren, Ch., Morphologic des acidies. Bull. R. Acad. Brux. 5:430, 1838.
Morren, Ch., Criticism of Bower's Review of above. Ann. Bot. 7:420, 1893.
Oudemans, C. a. J. A., De Bekerplanten. Amsterdam 1864.
Oye, p. van, Zur Biologic der Kanne von Nepenthes melamphora. Biol. Zentralblatt 41:

529-534, 1921-
Rees, M. & H. Will, Einige Bemerkungen iiber " fleischf ressende " Pflanzen. Bot. Zeit.

33:713-718, 1875 {also see Sitzungsber. d. phys.-med. Soz. Erlangen 8:13, 1875).
Robinson, W. J., Torreya 8:181-194, 1908.
RmxER, C. DE, Op zoek naar de bekerplant met de "Marie-Stuart kraag", Nepenthes

Veitchil Hook. f. De Trop. Natuur 24:195, 1935. {through Troll).
ScHMiTZ, P. H. S. J. & J. V. DE Janti, Contribution a I'etude de la faune nepenthicole.

Natuurhist. Maanblad I, 21 (9):ii6-ii7, 1932; II, 21 (12), 1932; III (by A. Starke),

22 (3):29-3i, 1933; IV (by E. O. Engel, Beitrag zur Morphologic der Larva von

Wilhelmina nepenthicola Villeneuve), 22 (4) :46-48, 1933; V (by Schmitz), 23 (3) :26, 1934;

VI (by S. L. Brug, Culicidae collected from iVe/>eM/Aej in Borneo), 23 (ii):i49-i5o, 1934-
Sms, John, Nepenthes phyllamphora, Ventricose Pitcher Plant. Curtis's Botanical Magazine,

53: plate 2629, 1826.
Stern, K., Beitrage zur Kenntnis der Nepenthaceen. Diss., Jena 1916, Flora 109:213-283, 1917.
Stern, K. G. and E. Stern, Ueber die Proteinasen insektivorer Pflanzen. Bioch. Zeitschr.

252:81-96, 1932.
Tate, Lawson, Nature 12:251-252, 1875.
Thienemann, a.. Die Tierwelt der N epenthes-Y^avmen. Archiv f. Hydrobiologie, Suppl. 11,

1932. Tropische Binnengewasser 3:1-54.
TiscHUTKiN, N., tJber die RoUe der Mikroorganismen bei der Ernahrung der insectenfres-

senden Pflanzen. Arbeiten d. St. Petersb. Naturfor. Gesellsch., Abt. f. Bot., 1891:33-37.

{Digest in Bot. Centralb. 50:304-305, 1892).
Tr£cul {through Mangin).
Treviranus, Zeitschr. f. Physiologic 3:78, 1829.

Troll, W., Morphologic der schildformigen Blatter. Planta 17:153-314, 1932.
Troll, W., Vergleichende Morphologic der hoheren Pflanzen. Berlin, 1939 (A rich source

of literature citations.).
Vines, S. H., On the digestive ferment of Nepenthes. Journ. Linn. Soc. 15:427-431,

1877. The proteolytic enzyme of Nepenthes I: Ann. Bot. 11:563-584, 1897; II: Ajin.

Bot. 12:545-555, 1898; III: Ann. Bot. 15:563-573, 1901.
Vines, S. H., Jour. Anat. a. Physiol. 11: 124-127, 1876/1877.
VoELKER, A., On the chemical composition of the fluid in the ascidia of Nepenthes. Ann.

and Mag. Nat. Hist. II, 4:128-136, 1849.
VoUK, v., Physiologischer Beitrag zur Kenntnis der Entwickelung des Nepenthes-Blditte?,.

Bot. -Physiol. Inst. d. K. Univ. Zagreb, 22 Jan. 1918 (In support of the Goebel inter-
pretation of the leaf morphology, based on growth localization).
Wallace, A. R., The Malav Archipelago. loth ed., p. 24, London, 1913.
WuNSCHMANN, E., Ubcr die Gattung Nepenthes. Diss. Berhn, 1872, 46 S.
Zacharias, E., iiber die Anatomic des Stammes der Gattung Nepenthes. Inaug. Diss.

Strassburg, 1877.
Zeeuw, J. de, Versuche uber die Verdauung in N epentheskanntn. Biochem. Zeitschr.

269:187-195, 1934.

Chapter V
CEPHALOTUS FOLLICULARIS

Distribution. — Habit. — Habitat. — Foliage leaf. — Pitcher leaf. — Development of
pitcher leaf. — Morphology. — Anatomy. — Digestion.

The West Australian Pitcher Plant is a unique form and, though
related to Sarracenia and Nepenthes, diverges from them in many de-
tails of form and structure. It occurs in a lunate area, in extreme
S. W. Austraha, one horn of the crescent lying about 150 miles S. of
Perth, the other at the Fitzgerald River, the southern rim of the area
passing through Albany. Its first collector was probably Archibald
Menzies, naturalist of the Vancouver's Expedition of 1791. Menzies
"landed at King George's Sound and made large collections." But as
these were not studied till much later by Robert Brown, the plant,
if actually found, did not become known. In the following year, 1792,
came the expedition under d'Entrecasteau (" Voyage a la recherche de
la Perouse''). The naturahst was La Billardiere. He landed first "on
one of the islands of Esperance Bay and then on the mainland" (Gard-
ner 1926). Here the naturalist of the expedition found the plant
which he later (1806) described under the name Cephalotus follicularis.

The plant is of rosette habit, the rosette, where primary, surmount-
ing a tap-root (La Billardiere). and in older plants ending branches of
a freely forking rootstock. These branches when small produce for
some time only minute leaves and pitchers; more massive branches
produce at once larger or even normal sized organs. The flowers, in a
short panicle, and borne on a very long slender scape, triangular at its
base, are small, apetalous, have a six-parted calyx and twelve stamens

The habitat is the drier parts of peaty swamps. The leaves are,
as has been known since the publication of La Billardiere's descrip-
tion, of two very distinct kinds: the fohage leaves, or " non-ascidif orm "
(Dickson 1878) (9 — 6) and the pitcher or ascidiform leaves {g — 1-3).

The fohage leaves attain a length of about 13 or 14 cm. when of
large size. The blade is ovate and acute, about the length of the peti-
ole, which, as Troll has shown, is of unifacial structure. Two of the
vascular strands, dorsal and ventral, facing each other wood to wood,
enter and extend up into the blade, thus indicating, according to Troll,
the peltate structure of the leaf. The ventral strand enters and sup-
phes vascular tissues to the hd of the pitcher when this develops in
place of a flat leaf. The blade is furthermore inchned to transverse
thickening above the petiole (9 — 6 at left). This becomes very pro-
nounced in intergrade forms between pitchers and foliage leaves which
in this plant occur very frequently, and will be described below.

The leaf is thick, coriaceous and supplied with nectar glands, and
its surface smooth and glassy. The margins are ciliated with the
pecuHar hairs mentioned by Dickson (1878). Their pecuUarity con-

Francis E. Lloyd

— 82 — Carnivorous Plants

sists in the secondary filling in of the lumen by callus (or callus-like
substance), the protoplast withdrawing toward the base. A central
thread-like core of protoplasmic substance with more or less con-
tinuity can be traced through the otherwise soUd mass of callus.

FoHage leaves are produced in seasonal rhythm. Of this A. G.
Hamilton (1904) wrote, ''I believe that the ordinary leaves develop
in the autumn, reaching their full maturity in the spring, and then
gradually going off, while the pitchers grow in winter and spring and
are fully formed and functional in summer when the insects which they
capture are most plentiful." This seems to be a true account. I can
only add that at the time of my visit to the classical ground of Albany
in the spring (Oct. 1936), the pitchers were in full representation, and
foHage leaves much less conspicuous.

The pitchers when full sized measure in length about 5 cm. or
slightly more. The majority measure less, say about 3 cm. in length,
and about 2 cm. in transverse measurement, somewhat compressed
from front to back. The orifice is oval in form, wider transversely
than from back to front, measuring in a pitcher 5 cm. long, i X 1.5 cm.
(inside measurement). Hamilton well compares the form of the
pitcher with that of a loose slipper, with the heel turned over to form
a lid. Its stalk (the petiole) stands approximately at right angles with
the axis of the pitcher, and in this at once we see a marked divergence
from the morphology of Sarracenia and Nepenthes, in that the mouth
of the pitcher faces the base of the petiole in Cephalotus, while the
opposite occurs in the other two genera.

This is best understood by examining the development, as did
EicHLER (1881) and Goebel, or by comparing the various aberrant
pitchers which in this species are rather common and have been re-
marked by Dickson, Goebel, and Hamilton. The true orientation
is clearly seen in a pitcher only i to 3 mm. long. Such a one in longi-
tudinal section is seen in the figure {10 — 6), in which it is evident
that the Hd does not terminate the leaf (Dickson), but is an outgrowth
from the upper surface of the petiole below the pitcher proper while
the pitcher has been produced by a ventro-dorsal invagination of the
upper, more distal region. Abnormal leaves, which occur in sizes be-
tween I mm. to a few cm., bear out the above interpretation, and fur-
ther show that the hd represents the transverse extension of the leaf
margins across the basal zone of the blade, that therefore the pitcher is
a peltate leaf highly differentiated into the compHcated apparatus that
it is {10 — 13-18). The orientation and course of the vascular bundles
are in accord with this interpretation, though Arber argues that the
absence of a median ventral vein in the petiole does not agree with
Troll's description, and that this raises doubt as to the truly peltate
condition. With Arber, I find no median ventral bundle [lo — 2).

The mouth of the pitcher is surrounded by a corrugated rim, each
corrugation forming a claw-hke tooth extending inward and downward,
much as in Nepenthes, except here the teeth are coarser and are not
provided with glands. There are about 24 such teeth, the numbers on
each side not being always symmetrical. I counted 12 on one side and
II on the other in a particular specimen. They are largest in front
(the ventral aspect of the opening) and are smaller and smaller as one

Chapter V — 83 — Cephalotus

swings around the curve towards the lid, and are longitudinally ridged.
The largest teeth, however, are opposite the median and lateral ridges
{lo — 2). The purse of the pitcher externally has three strong ridges,
one a ventral one, T-shaped in transverse section, extending along the
front of the pitcher along its whole length, along and below the midrib
of the pouched leaf, the other two lateral and obliquely placed. These,
too, are T-shaped though less obviously so. The lateral and median
wings are connected by a low ridge, readily discernible only in strongly
developed pitchers. From each lateral wing there runs a similar but
more vague ridge toward the petiole {10 — i, 11). All three bear
strong cilia, chiefly on the edges of the lateral wings {10 — 7). These
cilia develop early, so that a young pitcher looks, as Hamilton put it,
like a "vegetable hedgehog." These ridges must be regarded as ena-
tions from the ventral and subventral surfaces of the leaf (Goebel
1 89 1, Troll 1932). In addition to these there are low but quite evi-
dent ridges between them, especially evident near the toothed rim, and
which may pass for mere rugosities, but which are probably more than
that. The rest of the frontal (appearing ventral but really dorsal)
surface presents low rugosities. The lid overhangs the opening more or
less closely according to age, and is nicely arched, but is not, as once
believed (Woolls), moveable. It is traversed radially by narrow
patches of green ciHated tissue, often forking once or twice toward the
margin of the lid; and lying between them are clear patches devoid of
chlorophyll, which present window-like areas framed in green, or in
nature usually bright red mullions. Whatever their purpose is, they are
evidently analogous to the fenestrations in Sarracenia and Darling-
tonia: they are to insects apparently open spaces and the insects are
thus tempted to escape through them, to rebound into the depths of
the pitcher. The lid is emarginate, a feature which can be seen in
abnormal intergrade forms, in which the transverse pad at the base of
the blade betrays itself as a bilobed structure {10 — 11, etc.). In the
unopened pitcher the apical notch of the Hd lies beneath the end of the
median enation and straddles its wing, the lid margin inclosing the
teeth. The edge of the lid is ciHated, the hairs becoming reduced and
more or less contorted along the frontal region. It is devoid of a mid-
vein, being supplied by two pairs of veins from the ventral moiety of
the petiole bundles {10 — 4, 8). The veins traverse the green strips of
the lid between the white patches. From each of the angles of the lid
and mouth edge runs a low ridge (scarcely "wings" as Arber puts it)
demarking, according to Dickson, the ventral aspect of the pitcher
(70 — 1, 11).

Mrs. Arber (1941) has advanced the suggestion that "the lid ... .
may be interpreted as a hypertrophy of the collar region", that it is
"essentially of the same nature as the collar" being "indicated by the
fact that the cornice continues unaltered below both the collar and the
lid. It is possible that the thickened ribs of the expanded hd are
equivalent to the hooks of the collar". To this it may be replied that
the teeth of the rim are developed from the margin of the abaxial, distal
part of the leaf and that the lid is the whole adaxial, basal part of the
leaf which, as the teratological evidence shows (/o — 13-18), arises as
two lobular extensions that fuse (concrescence), the indication of this

Francis E. Lloyd

84 — Carnivorous Plants

fusion being found in the emargination of the lid, and in its ''dual"
structure, to be expected in peltate leaves.

The whole of the pitcher, "slipper" shaped as already said, has a
gentle forwardly concave curvature. The under side is the thinnest
region, and rests, in nature, on the surface of the soil, in such a manner
that the pitcher stands more or less obHquely {9 — 1, 2).

The interior of the pitcher is divisible into two distinct zones, the
upper of which forms a collar with, at its lower edge, an overhanging
eave. The epidermis of this collar ("conducting shelf", Dickson)
forms a surface of low pointed trichomes which are downwardly di-
rected, supplying a smooth, glistening, chalk- white face. This surface
is continuous with that of the lid, where the trichomes point in the
same sense but here they are very low and appear as imbricated.
Among them are numerous nectar glands.

The jutting eave overhangs, like the entrance of a lobster pot, the
far interior of the pitcher. Here the surface is smooth, and the epider-
mal cells are wavy-walled, the radial synclinal walls supported by
numerous buttresses from the angles of the undulations. There are in
the upper region (ca. one half of the surface) extending further down
in front than behind, many glands, which are smaller above, becoming
larger below. These are, it may be fairly argued, digestive in function.
In the lower half there is on each side an obliquely placed kidney-
shaped mass, in reaUty a thickened bolster of tissue, called by Dick-
son the "lateral coloured patch," since it is usually deeply red colored,
and which Hamilton preferred to call the "lateral gland mass." The
upper zone of this bolster is the seat of a number of very large glands
though they are not wholly confined to it (9 — 4)- Its lower half has
a very peculiar feature in the presence of numerous immobile stomata
with widely open mouths first observed but not properly understood
by Dickson (/o — 23). The function of the glands also is digestive,
the general evidence for which was offered by Dakin {see below). The
lower portion of the general surface of the pitcher interior is entirely free
of glands. Hamilton thinks that normally only the lower portion of
the pitcher holds fluid and the obliquity of the distribution of the
glands in the upper zone is correlated therewith, since the pitchers
usually lie somewhat obliquely on the ground. My own observations
lead me to doubt this as a matter of fact; particularly it is difficult to
agree that the quantity of fluid is so definitely restricted. While it is a
curious enough fact that the distribution of the glands is as described
above, there may very well be another explanation, for the glands of
the lateral patches in any event would, according to the Hamllton
view, be submersed.

Slender rhizomes produce very small pitchers, having a slightly
different aspect in detail from that of the normally larger pitchers.
They attain a size in general of about i cm. in length, often less, with
tissues correspondingly thinner and more delicate. A major difference
is in the development of the teeth surrounding the mouth: there
are fewer of them and all arise from an external low ridge, and stand
freely independent of the actual edge of the mouth (70 — 9-12), one
opposite each of the three wings, and two further back on each side.
A further important difference is the relatively greater width of the

Chapter V — 85 — Cephalotus

collar, as will be clear in the figures. Correlated with size are the
simpler venation and small number of glands. Hamilton drew atten-
tion to this condition (1904). In the large pitchers the teeth are
concrescent with the rim and overhang it inwardly. Another feature
of juvenile pitchers is the large size of the lid, which is strongly arched
and widely overhangs the opening, so that it more efifectually pre-
vents the entrance of rain water, or appears to. As I have observed
these small pitchers are efficient in catching correspondingly small
insects.

Transition forms between the large and small pitchers have not
been observed. When a relatively large pitcher appears after a number
of small ones have been produced, the passage from the small to the
large form is made at once in a jump.

We turn to the anatomy of the pitcher {10 — 3, 6, 12). The
venation is derived from two systems of bundles in the petiole, a dorsal,
of three veins, and a ventral of two, these splitting near the pitcher
into four, then six and branching further in spreading. Referring to
the figures in which the veins are numbered, we see that of the ventral
system, Vi, (median ventral pair) passes into the lid, right and left
of the midline; V2 sends veins into the sides of the hd and into the
collar; V3 goes entirely to the upper part of the digestive cavity,
anastomosing with the dorsal veins. It seems quite doubtful that
Arber's statement about "the relatively high development of the
ventral system of the pitcher's venation" corresponds with the facts,
since one third of it is not connected with the lid at all, and only a
small part of it with the collar. Of the three dorsal veins of the
petiole, the median is the midvein of the pitcher, passing entirely
around it, and ending, not in the point of the median ridge, as Dickson
claimed, who therefore regarded it as the leaf apex, but in the collar,
opposite the middle tooth, there branching. Of the laterals (D2) each
runs obliquely down the side of the pitcher toward the upper end of
the glandular patch, having just before reaching it sent a branch into a
lateral ridge, whence it emerges in the collar. Traversing the glandular
patch obliquely it leaves it near the middle point and then runs up the
wall parallel to the midvein, and ending in the collar. The midvein
(Di) sends branches right and left into the lower part of the pitcher.
This basic arrangement of the vascular system of the pitcher can be
most clearly seen in a very young one, 3 mm. long {10 — 5).

The external surface of the pitcher is covered with an epidermis of
isodiametric cells with thick walls, and is supplied with stomata and
nectar glands. On the lid the epidermis of the green patches is of
more or less wavy-walled cells, with glands and stomata interspersed,
while in the fenestrations the cells are isodiametric and straight walled,
with glands but no stomata.

The epidermis of the interior surface of the lid and collar has been
already described above. That of the far interior is of wavy- walled
cells, the walls thick and buttressed at the angles. Scattered through-
out the surface, except along a narrow strip beneath the eave of the
collar, and the deeper portion of the pitcher demarked by an oblique
fine running downward and forward from about the middle point of
the back surface across the top of the glandular bolster, there are

Francis E. Lloyd — 86 — Carnivorous Plants

numerous glands. These are smaller above and become increasingly-
larger below. In the bolster itself the glands attain the maximum size,
and occupy chiefly the upper half of it, though not entirely excluded
from the lower half {9 — 4). This latter is covered with a wavy-
walled epidermis supplied with extremely numerous stomata.

The small glands which occur on the outer surface and on the inner
surface of the Hd, have, as Goebel (1891) pointed out, essentially the
same structure as those of Sarracenia, but are directly comparable
rather to those of the outer surface of the pitcher in that genus. In
these there is only one course of cells, six in number, surmounting a
single parenchyma cell (2 — 16). The same is true of Cephalotus, with
the difference that, while in Sarracenia the "cover" cells are inwardly
drawn out to a point, those in the Cephalotus gland reach inwardly as
far as do the four surrounding cells {10 — 21). The glands are very
small, indeed no bigger in transverse section than the stomata with
which they are interspersed, and are no deeper than the surrounding
epidermal cells. The outer walls are all suberized, except both outer
and inner walls of the basal cell, derived from the parenchyma. The
inner wall of this cell, in contact with the six other cells, is not, as in
Sarracenia, reticulated. Whether more than one cell at the base of the
gland may be regarded as part of the gland is questionable but possible.
I have seen some indications that such is the case, as Goebel (1891)
seems to have thought. As he did not afford a drawing (nor has any-
one else since) of this particular gland, it is difficult to decide what
precisely was Goebel's meaning.

These glands are found on all green parts, and appear to have the
same function as analogous glands in Nepenthes, Sarracenia. Hamil-
ton observed insects feeding on the outer surface of the pitcher, but
could not satisfy himself that nectar was present. It is possible as
Goebel suggested that they secrete something else attractive to in-
sects.

The glands of the inner surface of the lid have the same structure
as those above described.

In the far interior of the pitcher the glands are of various sizes,
smaller above and increasingly larger the deeper they are placed till
the maximum size is reached in the glandular patches. They are flask-
shaped, with a broad neck lying in the plane of the epidermis, made
up of a greater or smaller number of columnar cells (neck cells) whose
outer walls are very much thickened and pitted. The wafls lying
against the epidermis around the neck of the gland are also thickened
and suberized, and, forming an investment of the whole gland there is
a single layer of flatfish cells (similar to the flat cell below the small
gland of the outer pitcher surface) which are strongly cuticularized in
their radial waUs only, not, as Goebel thought, wholly. Each of these
sheathing ceUs is therefore a window, or better a double cellulose
window framed in mullions of suberized walls. The body of the gland
is made up of rounded thin-walled cells, evidently the active glandular
secreting cells, as indicated by the richness of the protoplasmic con-
tent {10 — 20).

When the neck ceUs are examined as part of the epidermis in face
view, the outer walls, in the case of the smallest glands, are arranged

Chapter V — 87 — Cephalotus

in the typical manner — two cover cells surrounded by four others. In
slightly larger glands additional cells are intercalated. Their outer
walls are seen now to be thick and pitted. The surrounding epidermal
cells overlap the shoulder of the flask, and the strong buttress thicken-
ings of their radial cell walls stand out {lo — 20).

The largest glands are to be found in the areas in the "colored
patches" as Dickson called them, on account of their deep red color-
ing. They differ in no respect beyond that of size from the others.
They are spherical in form, with a thick neck and the central mass of
something like 150 to 200 cells is surrounded as seen above by a single
layer of flat cells with their radial walls suberized, the periclinal walls
being of cellulose, thus ensuring a path of diffusion {10 — 19).

The colored or glandular patches, of which there are two, one on
each side of the pitcher, are the most remarkable feature of this species.
They are reniform, thickened regions of the wall, the outline being sharp
and well marked below and more or less crenate along the upper
edge (p — 4). It is a "bolster" (Goebel) of tissue in which the
mesophyll is more developed than otherwhere, and projecting in-
wardly, showing no sign of its presence on the outer surface. The
glands just mentioned are more numerous on the upper moiety, but are
by no means confined to it. The epidermis between the glands offers
the most remarkable appearance of all in that it is supphed with in-
numerable stomata. Dickson (1878) described them as small oval
bodies surrounded by two to four other cells. Hamilton remarked
that they are "remarkably Kke stomates" but that there is always a
wide opening between the guard cells. Dakin (191 8), at that time a
member of the staff of the University of Western Australia, visited
Albany and there obtained material for study. He saw clearly that
these structures are stomata, confirming Goebel's earlier description
(1891). It is clear that the guard cells are immobile and that these
stomata do not function as such. Goebel called them water pores,
pointing out that the pore is plugged by the cellulose membrane of a
parenchyma cell underlying it, which would not, of course, prevent the
excretion of water. Dakin found that the membrane closing the pore
is locally thickened to form a "pad" which he thought acted as a
torus that, with changing turgidity of the cell, would open and close
the pore, the whole acting as a regulating mechanism. He further
thought that the function of the stomata is absorption and suggested
that the glandular patches be called lateral absorbing areas {10 — 12).

I found (19336) that Dakin is correct in his claim that the wall of
the underlying parenchyma cell is thickened beneath the pore; but
that the thickening is so definitely torus-like as he showed in his figure
(191 8, Fig. 11), and especially his interpretations are certainly to be
doubted. There is some evidence that the plugging membrane is the
result of hydrolysis of the occluding wall and that there is given off a
mucilage-hice secretion (Lloyd 19336), but further study on fresh ma-
terial obtained at Albany does not strengthen this idea of the matter.
The more ready staining of the torus-like thickening is due to the fact
that it is not cuticularized as are the guard cells and the epidermis in
general, so that cellulose stain (such as methylene blue) attacks the
thickening quickly. That, however, these structures are important

Francis E. Lloyd — 88 — Carnivorous Plants

physiologically is hard to resist in view of their number and general
relations. Goebel's idea that they are water pores seems the most
acceptable, that is, that they pour a fluid into the pitcher cavity; but
this fluid may contain substances in solution, more Hkely enzymes,
possibly one or more enzymes different from those of the glands. I
did a simple experiment with living pitchers to test Goebel's idea.
Halving a pitcher longitudinally, and cleaning it out thoroughly, I
placed it in contact, by its outer surface, with water in a closed damp
chamber. In the course of some hours beads of moisture appeared
from the mouths of the glands, larger ones from the larger glands but
none from the stomata, at least in appreciable quantities. This seems
to indicate that water excretion by the "water pores" plays a minor
role, if any, and that Dakin's suggestion that their function is that of ab-
sorption cannot be dismissed without further examination. Any in-
terpretation of the activity of these stomata must take into account
the constant presence of a large amount of starch in large grains in the
mesophyll of the glandular patches.

The problem of digestion by the pitchers has been examined in any
thoroughgoing way only by Daken (191 8), who spent a vacation at
Albany, W. A., making as careful a study as he could, under laboratory
conditions. To be sure, Dickson (1878) had reported that Lawson
Tate had examined into the matter somewhat and had found that
"fluid taken from virgin or unopened pitchers" showed "that it ex-
erted a similar digestive action on animal substances to that exhibited
by the Nepenthes pitcher, etc." Dakin made use only of the fluid
from opened pitchers, which did not surprise me when on careful ex-
amination of all the unopened pitchers which I could come by on my
own visit to Albany, I found no one of them containing any sign of
fluid, a matter of disappointment as I had intended to conduct experi-
mentation on such fluid if it could be found. Dakin's results are
as follows: He found that the pitchers capture many insects, notably
ants, as others had found. They are represented usually by fragmen-
tary remains of the chitinous parts. Even the very small pitchers, as I
have previously said, catch small insects. That the soft parts undergo
dissolution in some sort is at once evident. But, Dakin asked, is this
the result of digestion by enzymes secreted by the pitcher glands, or of
bacterial action, or of both ? Fibrin was his test substrate. The ex-
periments were conducted with pitcher fluid with an antiseptic (HCN)
and with and without weak acid (HCl) or alkaK. The specific
results which he records showed that pitcher fluid in vitro in the
presence of added acid does digest fibrin, and that it contains a di-
gestive ferment which will break up proteins into peptone-like bodies
in the presence of acid. Since non-acidulated pitcher fluid does not act
thus, it cannot be concluded that this process actually takes place in
the pitchers under normal circumstances. Pitcher fluid alone procures
dissolution of fibrin with the odor of putrefaction. Dakin admits the
possibility that digestion by pitcher fluid may, however, take place
very slowly in the pitchers.

He raises, however, the question of the usefulness or necessity of
this to the plant. He kept plants under his eye in the laboratory
where they grew thriftily and flowered without having been supplied

Chapter V — 89 — Cephalotus

with insects. In view of the work of Busgen (Utricularia) and of F.
Darwin (Drosera) he does not exclude a "carnivorous tendency."
On the whole, therefore, at the present moment, the evidence favors
the view that both the secretions of the pitcher and the action of
bacteria contribute to the breaking down of proteins making the
products available to the plant. Experiments with starch showed no
evidence of the presence of diastase.

Literature Cited:

Arber, Agnzs, On the morphology of the pitcher-leaves in Heliamphora, Sarracenia,

Cephalotus, and Nepenthes. Ann. Bot. n.s. 5:563-578, 1941.
Brown, Robert, General remarks on the botany of Terra Australis. Miscellaneous

Botanical Works i :76-78, 1866.
Dakin, W. J., The West Australian pitcher plant (Cephalotus follicular is), and its physi-
ology. Journ. Roy. Soc. W. Austr. 4:37-53. 1917/1918.
Dickson, A., The structure of the pitcher of Cephalotus follicularis. Journ. of Bot. 16:1-5,

1878.
Dickson, A., On the morphology of the pitcher of Cephalotus follicularis. Trans, and Proc.

Bot. Soc. Edin. 14:172-181, 1882.
EiCHLER, A. W., tJber die Schlauchblatter von Cephalotus. Jahrb. des Berliner Bot. Gart.

1:193-197, 1881 {through Engler u. Prantl).
Gardner, C. A., The history of botanical investigation in Western Australia. Handbook

for B. A. A. S. i8th meeting, Perth, W. A., 1926 (Pp. 40-52).
GiLBURT, W. H., Notes on the histology of pitcher plants. Quekett Microscopic Journal

6:151-164, 1881.
Goebel, K., Pflanzenbiologische Schilderungen, Pt. 2. Marburg 1891. {Cephalotus: pp.

110-115; 170-173).
Hamilton, A. G., Notes on the West AustraHan pitcher plant {Cephalotus follicularis La

Bill.). Proc. Linn. Soc. N. S. W. 29:36-53, 1904.
Lloyd, F. E., The carnivorous plants — a review with contributions (Presidential Address).

Trans. Roy. Soc. Can. HI, 27:1-67, 1933.
Maury, Paul, Note sur I'acidie du Cephalotus follicularis La Bill. Bull. Soc. Bot. France

34:164-168, 1887.
Tate, Lawson, Phil. Trans. Birmingham 1878 {through Hamilton).
Troll {see under Nepenthes).
Woolls, W., Lectures on the vegetable kingdom, p. icx) {through Hamilton).

Chapter VI
GENLISEA

Discovery. — Early studies. — Two kinds of leaves. — Anatomy of trap-leaf.

The specimens on which the genus Genlisea is based were discovered
by AuGUSTE DE Saint-Hilaire in Brazil in 1833. Most of the species
are found in the New World in Brazil, the Guianas and Cuba, while
two are known from west tropical Africa. The Cuban species, found
many years ago by C. Wright at the time he found Biovularia olivacea,
has never again been collected.

For our information about these plants we are indebted first of all
to Warming (1874) and to Goebel (1891). All the species are small
plants which inhabit swampy places and apparently live mostly sub-
mersed in shallow water, only the inflorescence, as in Utricularia, pro-
jecting above the surface. This is to be inferred from the absence of
stomata and from the fact that colonies of algae have been observed by
me attached to the surfaces of the leaves. Benjamin in the Flora Bra-
siliensis says merely "herbae paludosae." The close relationship to
Utricularia is shown by the fact that the structure of the flower is the
same in the two genera, that of Genlisea differing in having a five-
parted calyx instead of the two-parted calyx of Utricularia. All are
rosette plants with two kinds of leaves, foliage and trapping, arising
from a vertical or sometimes nearly prostrate rootstock. Like Utric-
ularia, there are no roots, though the trap leaves look superficially
much like them and have been mistakenly so regarded by some (p — 7 ;

The first thorough description, though lacking in an important de-
tail, was published by Warming in 1874. This work was known to
Darwin, whose son Francis repeated Warming's observations and
afforded the description given by Darwin in his Insectivorous
Plants (P. 360, 2nd ed. of 1875). Goebel's description of 1891, though
incorrect in certain details, leaves otherwise little to be desired. The
plant which these authors studied was Genlisea ornata, the largest
known species. The present account is based on herbarium specimens
(British Museum of Natural History and Kew) but more particularly
on alcohol material kindly sent to me by Dr. F. C. Hoehne, collected
in Butantan, Brazil. As far as the anatomy is concerned the genus is
very homogeneous. Darwin, it is true, described G. filifor?nis ^ as
bearing bladders like those of Utricularia and being devoid of ''utricu-
Hferous leaves" characteristic of the other species. I examined all
the specimens of Genlisea filiformis at Kew, which was the source of
Darwin's material, but could find no evidence to corroborate him.
It seems quite certain that he examined a plant which had been grow-
ing with a Utricularia whose stolons had intermixed with the Genlisea
leaves. Indeed, I saw a case of this.

There are two kinds of leaves, true foliage leaves, linear or spatulate

Chapter VI — 91 — Genlisea

in form, and trap leaves, all arising densely crowded and without trace-
able order, from a slender rhizome, very much as the leaves and stolons
arise from the radially symmetrical corm-like stem of the seedling of
Utricularia. There are no axillary buds, again as in Utricularia,
but the rhizome produces a few branches toward the apex, which is
the widest part. The trap leaves arise like the stolons of Utricularia
and at first look like them. At first cyHndrical with a tapering grow-
ing point, they grow out for some distance (i cm. or more or less) be-
fore any further differentation takes place. In structure this portion
consists of epidermis inclosing a very extensive intercellular air space
of lysigenous origin. In the dorsal sector lies a cord of relatively few
parenchyma cells surrounding the vascular tissue, again quite hke a
Utricularia stolon. This portion may be called the foot stalk, but not
petiole since this leaf region is produced by intercalary extension be-
tween the leaf base and the apex, while, as Goebel pointed out, the
base at the foot stalk is the oldest portion of the trap leaf, which ex-
tends solely by apical growth. At length the end of the footstalk be-
gins to widen and an invagination takes place just behind the tip and
on the ventral (upper) side. The basal portion of the invagination be-
comes a subspherical hollow bulb. The neck of this bulb extends for
some distance to form a tube, which toward the mouth gradually
widens to right and left, so that the opening becomes a transverse slit,
with the lips dorsal and ventral, the latter being shorter, and the for-
mer being more or less arched over the opening. The angles of the
mouth develop into two long arms with circinate apices, the slit being
on the outer curve of the crook (// — 3, 4, 7). During elongation and
resulting from rotatory growth, the arms become twisted, the one on
the right, clockwise, the other counter clockwise {11 — 8, 10). In
consequence one lip of the mouth of the arm, which extends through-
out its length, becomes longer than the other, so that, if an arm be
laid open it takes the form of a spiral ribbon {11 — 9). The arms may
be likened to two ribbons folded longitudinally and twisted on the long
axis so that the two edges form spirals roughly parallel to each other.
One edge becomes the inner, and in the plant is the shorter. In the
actual trap, the two edges are anchored to each other at short inter-
vals. This is accomplished by large marginal cells, cystid-like in ap-
pearance, which during growth become pressed into, and adherent to,
the tissue of the apposed edge. These large cells we may with Goebel
term prop-cells. They were first described by Goebel (1891) but not

quite correctly. He wrote " the funnel shaped entrances are

formed by the occurrence at certain distances apart of two large clear
cells which He the one upon the other, and which may be called prop-
cells. They are merely the end cells of the rows of trapping hairs in
which, however, the hair itself is merely one-celled, while the cell be-
neath is swollen to a giant size." By making a paper model it will be
seen, continues Goebel, "that in order that the two prop-cells shall
really meet each other it is necessary that the shorter edge of the arm

shall be bent outwardly. One can see the two prop-cells "

This passage is quoted to indicate that Goebel thought that there is
a row of prop-cells along each margin of the arm entrance, and that
during development these meet and adhere in pairs, the one prop-cell

Francis E. Lloyd — 92 — Carnivorous Plants

to the other. The facts are otherwise. There is a row of prop-cells
on only one edge, and the prop-cell is Only the middle cell of a three-
celled hypertrophied trichome, the basal cell of which is much enlarged,
while above it the middle cell is enormously large and ends in a small
knob-shaped cell terminating the trichome (72^15). In structure
they are, therefore, not at all different from the neighboring trapping
hairs, except for relative sizes of the component cells. The size of the
basal and middle cells is so large that, in sections which are bound to
be pretty complicated to the eye, they appear as two apposed and ad-
herent cells. GoEBEL represented them thus in his figures (7a and 76,
plate 16, 1 891). It is significant that Goebel showed a terminal cell
on only one of each pair of prop-cells, as he regarded them to be (Figs.
6 and 76). In this detail Goebel was correct. What he took for the
prop-cells along the shorter border of the ribbon-like arm are the scar-
like depressions, optically suggesting raised surfaces, which are really
dished out surfaces against which the prop-cells of the longer border
lay and to which they were attached {11 — 6). When the two
margins of the arms are torn apart in dissection, it happens more
frequently than otherwise that the whole of the prop-hair is torn away
from its moorings, leaving bare the depressed surface to which it was
attached. The depression so caused is spoon-shaped, the bottom being
formed of cells which have been more or less distorted by the pressure
of the prop-cell during growth {12 — 18). On the other hand, the prop-
cell is sometimes torn away from its basal cell, and remains on the
wrong margin, a perfidious witness whose evidence is hereby impeached.

A striking analog of the prop-cells is to be found in the cystidia in
CopHnus atramentarius in which they serve to keep the slender gills
at a certain distance apart, allowing the free dispersal of spores, as
described by Buller (1922), in his Researches on Fungi, where he in-
troduces the engineering term "distance pieces" for the cystids.
Protruding from one gill, from which they arise, their free ends are
attached to the surface of the next gill.

The size of the trap leaf in Genlisea repens, one of the smallest
species, is as follows. The footstalk is about i cm. in length support-
ing the bulb-shaped flask which is about i mm. long and 0.7 mm.
broad. The surmounting tube is about i cm. long, and 0.27 mm. in
outside diameter. The arms extend i cm. beyond the transverse mouth
and are little more than 0.5 mm. in width. In a large African species,
the traps are about three to five times the foregoing dimensions, the
tube being relatively shorter. The footstalk may be 5 cm. long, the
tube two and the arms 3 to 5 cm. long. The bulb is about 4 mm. long
and 2 mm. in diameter. The turns of the arms are looser and make a
larger angle with a transverse plane.

The outer surface of the plant is supplied with a large number of
sessile globular, glandular trichomes, similar to those of Utricularia,
and which secrete mucilage {12 — 11). The trap, whose inner surface
is most complicated, has excited the wonder of all who have busied
themselves with this object. Darwin referred to it as "a contrivance
resembling an eel-trap though more complex." Goebel (1891) re-
marked of it that "it is in the highest degree remarkable; one might say
of it that it is constructed with over-weening care and anxiety so as

Chapter VI — 93 — Genlisea

to allow only very small animals to enter and then to hold them ir-
revocably". This remarkable structure is as follows.

In form, the bulb and tubular neck (the tube) may be compared to
a chianti flask. Within the flask there are two ridges (if we were
speaking of an ovary they would be called placentae), one ventral and
one dorsal, extending from the base up the sides about two-thirds the
distance to the neck above {12 — 10). Within the tissue of the ridges
runs in each a branch of the single vascular strand arriving from the
footstalk, while the surface bears numerous glands, which may be pre-
sumed to be digestive and absorptive, either but probably both. A
few additional glands are to be found on the rest of the surface. The
two vascular strands, each of a single spiral vessel accompanied by a
thin strand of phloem, the one dorsal and the other ventral, pass up-
ward from the bulb into the walls of the tube without change of di-
rection. Near the mouth of the tube they divide, a branch from each
supplying each arm, which then has two vascular strands quite as if it
were a closed tube branched from the main tube. The inner surface
of the tube is broken up into a series of some forty transverse ridges
each formed of a transverse row of radially thickened cells, each of
which sends downward toward the flask a stiff curved trichome {g —
8; 12 — 8, 9, 13). Of these cells there are about 50, so that there are
that number of slender stiff bristles projecting inward and downward
from each ridge. Each section of the tube below and including a
transverse ridge is therefore of the form of the entrance to an eel trap,
or lobster pot, if you will. The whole tube, 0.13 to 0.42 mm. inside
diameter, is a series of such traps, some forty to fifty in number, each
with its funnel extending into the next below. In addition to these
downwardly directed hairs, and just below the ridges in each section
there are one or two transverse rows of glandular trichomes {12 — 8,
9, 16). The zone where these occur broadens toward the outer end
of the tube and is composed of wavy-walled cells, while the bristle
bearing cells are conspicuously straight and narrow, lengthwise the
tube.

On approaching the open end, the tube widens somewhat, and
spreads out to form the arms. The open end is formed of the upper
and lower sides to form two lips, the upper (ventral) somewhat shorter
than the lower, and fixed in a position a little distance apart by the
ballooned cells above mentioned (prop-cells) (77 — 1-4; 12 — 2, 3).
These are closely enough placed so that in between, alternating with
them, a series of funnels, guarded by inward pointing hairs, is formed.
This is repeated along the open slit of the arm (77 — ^11) quite to the
apex.

In passing up into the arms, the same general structure described
for the tube is repeated {12 — i, 4), but the ridges are now curved
obhquely, comformably with the directions and amounts of growth
(77 — 6, 9). Along the edges of the arm, as one inspects it if laid
open, the ridges run almost parallel thereto, each ridge beginning in a
prop-cell. Passing obliquely inward and forward they gradually ap-
proach the other edge in a harmonic curve. When past the middle of
the arm they bend rather sharply back and approach a direction again
parallel to the other edge and then end at the scar-like depression

Francis E. Lloyd — 84 — Carnivorous Plants

formed by its prop-cell at the other end {ii — 6). In consequence of
this development, the trapping hairs stand approximately at right
angles to the edges of the funnel formed by the prop-cells, so that al-
though oblique, the ridges with their trapping hairs function as in the
straight tube, although no two hairs on the same ridge have pre-
cisely the same direction. The whole structure is one to arouse won-
der in the observer.

The inner surface, except that occupied by the bristle ridges, is
made up of wavy-walled cells with scattered glandular hairs, repeating
again the structure of the tube (// — 6). The funnel shaped mouths
of the tube and arms are guarded, outside the level of the prop-cells,
by shorter stiff er hairs, claw-like in shape {12 — 12), allowing some
room for the entrance of prey, but nevertheless inveigling them to-
ward the interior. The captures consist of copepods, and the like,
small water spiders, nematodes and plenty of other forms, many of
which I have seen in the Brazilian material studied.

In both species examined, the structure is the same, with the slight
difference that the large African species structures are not so crowded
and in consequence are easier to decipher.

The glands are all of the same type, that common to this genus and
Utricularia, consisting of a basal cell anchored in the epidermis, a short
neck cell, and the capital of two to eight cells. It is wholly a matter
of speculation as to the function of these glands. They may supply
only mucilage to lubricate the interior and facilitate the movements of
prey downwards through the arms and neck, or they may secrete
digestive enzymes or both. I have observed that prey only half way
down the tubular neck shows signs of a far degree of disintegration,
but, as bacterial action cannot at the moment be excluded, it boots
nothing to do more than indicate the possibilities. The goal of prey is
the flask at the bottom of the neck. Here one finds various remnants
of the animals, copepods, spiders, nematodes, together with algae.

According to Goebel, the twisting growth of the arms facilitates
their penetration of the substrate which, being filled with water, is
quite loose. This explanation does not help for the trap leaf up till the
time when the arms begin to form, which is a good deal more than half
the time of its growth activity. If teleological interpretation be of any
use, one might venture that the twisted form of the arms results in the
presentation in all direction of entrances to the interior so that prey
find openings in whatever direction they may approach.

Literature Cited:

Benjamin, L., Flora Brasiliensis, 10:252, 1847.

BuLLER, A. H. Reginald, Researches on Fungi, Vol. 2, 1922.

Darwin, C, Insectivorous Plants. 2d. ed., London 1875.

Goebel, K., Pflanzenbiologische Schilderungen, 1891. Zur Biologic von Genlisea. Flora

77:208-212, 1893.
St. Hilaire, A. de, Voyage au district des Diamans, 11:428, 1833.
St. Hilaire & F. de Girard, Monographie des Primulacees et des Lentibulariees du Bresil

meridional et de la Republique Argentine. Mem. Soc. roy. des Sci. etc. d'Orleans 5,

1840.
TuTiN, T. G., New Species from British Guiana, Cambridge University Expedition, 1933.

Journ. Bot. 1934:306-341.
Warming, Eug., Contribution a la connaissance des Lentibulariaceae, I. Genlisea ornata

Mart.; II. Germination des graines de VUtricidaria vulgaris. Vidensk. Medd. f.

Naturhist. For. Kj0benhavn 1874:33-58. Resume in French (appendi.x 8).
Wright, C, in Grisebach's Catalogus plantarum Cubensium. Leipzig, 1866.

Chapter VII
BYBLIS

Occurrence. — Appearance and systematic position. — Habitat. — Structure. — Func-
tions of the glands.

Byhlis is a genus confined to western Australia, where it is endemic.
There are two species, B. linifolia Salisb. and B. gigantea Lindl., the
latter being much the larger plant, one about 50 cm. tall. It is a half-
shrub in habit, consisting of a woody rhizome bearing in any one
season the dying parts of the previous and the growing ones of the
present season (zj - — i). These consist usually of a single chief stem
with one to three branches from near the base, all bearing long (1-2
dm.) linear leaves, clothed with numerous stalked mucilage glands.
The color, a yellow-green, is characteristic, and the surface is charged with
numerous ghstening mucilage droplets. The flowers, raised on axillary
peduncles, are violet or rose colored, have a deeply five lobed rotate
corolla, which appears superficially as pol}^etalous, the lobes alter-
nating with five oval attenuate sepals and with the five stamens. The
systematic position of this plant has not been at all clear. Planchon
(1848) and Bentham (Flora australiensis 2:469) believed that it is re-
lated to the Pittosporaceae rather than to the Droseraceae. Later
Lang, stressing too much its sympetaly, advanced reasons for its re-
lation to Pinguicula and its inclusion within the Lentibulariaceae, while
more recently Domin (1922) has placed it in a new family, the Bybli-
daceae, of which B. linifolia is the type.

Byhlis gigantea was found growing abundantly in sandy, swampy
places in the Swan River district not far from Perth, where also are
to be found very characteristic species of Polypompholyx, (P. tenella
and multifida) and the peculiar Australian species of Utricularia, U.
Menziesii, Hookeri, etc., and all, except Polypompholyx tenella, confined
to W. Australia. Byhlis gigantea is, however, to be found in drier and
better drained parts of such swamps, as for example at Cannington
where it grows around the base of a low hillock on which stood a house,
and not, as Ross suggests, in very wet places on the banks of streams.
The substrate was a coarse quartz sand with some admixture of fine
white or yellow clay, and little humus. Specimens of Byhlis linifolia
were received from N. E. Arnhem Land where it was found growing
"around rocky pools in the bed of a river".

The stem arises from a slender rhizome with triarch (Lang) or, as
I have observed, diarch roots often showing a considerable degree of
secondary thickening with a thick cortex loaded with starch and tannin-
emulsion-colloid (Lloyd 191 i). Both of these may be regarded as
storage material. From the perennating rootstock arises the new an-
nual stem with its appendages, which are secondary branches, leaves
and long peduncled flowers. All these parts are clothed with two kinds
of glands, sessile and stalked. In aU parts except the sepals, the
epidermis is composed of elongated straight-walled cells, all of which

Francis E. Lloyd —96— Carnivorous Plants

in young organs lie at the same level. With maturity, the epidermis
becomes ribbed with sunken furrows between the ribs. The floor of
the furrow is composed of a double row of shorter cells, each pair
bearing a sessile gland (74 — 10, 13). In scattered positions occur
stalked glands which secrete abundant mucilage. In the sepals the
epidermal cells are wavy-walled on both surfaces, less so on the outer
(lower) surface toward the base. On the outer, dorsal face of the leaf
occur both sessile and stalked glands, the latter very numerous, on the
inner face only sessile glands occur (p — 9). Stalked glands are to be
found even on the ovary wall. Stomata occur on both faces of the
sepals, and on the leaves and stem they are to be found interrupting
the rows of sessile glands {14 — 13). They are somewhat raised and
extend considerably above the ditch bottom. In this way according to
Fenner the stomatal pore does not become clogged with the secretion
of the sessile glands, which probably fills that reach of the ditch oc-
cupied by them.

The leaves are long, slender and linear in form, tapering toward
the apex. When in the bud they display, in the case of B. gigantea,
no circination, the apices showing only a very meagre outward curva-
ture, if any. In B. Hnifolia, however, the leaves are outwardly cir-
cinate, as in Drosophyllum. This somewhat surprising fact was clearly
seen in the material from Arnhem Land sent me by my friend Mr.
Charles Barrett, and figured in 14 —7. B. gigantea is seen in 14 —
8. Of this DiELS (1930) says merely that the leaves are spirally in-
rolled at the tip.

In transverse section the leaves are triangular with round angles
{14 — 11). Toward the tip they become nearly cylindrical and the tip
itself is somewhat enlarged to form a knob, properly interpreted to be
a hydathode (Lang, Fenner). Its interior is occupied by a large
mass of tracheidal tissue in contact with and ending the vascular
strand which reaches thereto. One or two protuberant stomata are to
be found at the apical surface, not by any means always at the extreme
apex, together with both stalked and sessile glands. The rigidity of
the leaf, which is very slender for its length, is attained by the very
thick-walled epidermis and the strands of mechanical tissue accom-
panying the vascular bundles. Beneath the epidermis on all sides
there is a thick layer of chlorenchyma in which there is no sharp de-
markation between palisade and spongy tissue. All of the cells are
oval rather than columnar and lie in three courses. Beneath the
epidermis the palisade cells have expanding ends in contact with it
{14 — 9, 10). This, Fenner explains, ensures a contact for lively
diffusion between the glands and the vascular system. The upper leaf
surface is rather flat, with a very shallow depression along the middle.
On this surface there are very few stalked glands, which on the lower
surface are very numerous. Sessile glands are as numerous here as
elsewhere {g — 9) .

The sessile gland {14 — 10, 13) stands upon a pair of epidermal
cells, and consists of a capital of eight radially disposed cells, supported
on a single very short stalk cell, this resting on two short epidermal
cells, which according to Fenner originate from a single basal cell of
the very young trichome. The furrow in which the sessile glands stand

Chapter VII —97— Byblis

is sufficiently deep and narrow so that the sides of the glands lean
against and are supported by the sides of the furrow.

The stalked gland (14 — 12-15) has a capital of usually 32 cells
radiating from the centre and standing out like an umbrella top. These
cells all abut on a central short cell resting on the top of the long stalk
cell. This in turn stands on a group of basal cells which may be as
many as eight in number, or as few as two in the case of a small
stalked gland. The latter may also have as few as four cells in the
capital, the mature glands showing no great degree of uniformity in
this regard. The stalk cells of the larger glands have strongly striated
thick cellulose walls, the striations reaching deeply, as far as the cuticle.
These striations run obliquely (as in the cotton fiber) and when the
gland dries (in air or alcohol), the stalk cells twist, as noted by Dar-
win {14 — 14). Fenner regards this arrangement as one to allow
bending of the trichome without collapse.

While the gland capitals are covered with a thin cuticle there is
access by diffusion through pores, mentioned but not described by
Fenner. I found them (19336) to be rather large oval openings ar-
ranged in a circle about and some distance away from the centre of the
capital. They become evident on treatment of the stalked glands with
sulfuric acid {14 — 12). Both the sessile and stalked glands are
readily penetrable by dyes (methylene blue).

Our earlier knowledge of the function of the glands bearing on the
question of the carnivorous habit of the plant we have at the hands of
A. NiNiAN Bruce (1905). Her work is clearly indicative of this, but
the question needs further investigation, which in this type of plant is
not easy. Bruce placed minute cubes of coagulated egg-albumen in
contact with the sessile glands, and after a period of some days (two
to eight) the whiteness has completely disappeared. During the
progress of digestion the round white core of the cube of albumen could
be observed to suffer gradual reduction in size. This material placed
in contact with the heads of stalked glands failed to show any evi-
dence of digestion, but when removed and placed in contact with the
sessile glands promptly did so. This seems to indicate that bacterial
action does not supervene. Some observations by Fenner justify
Bruce's results. When insects are caught, he says, and come in contact
with the sessile glands, a secretion is thrown out by them which is
much less viscous than that of the stalked glands. After four to six
hours, the group of glands affected again become dry and an examina-
tion of them shows that the contents of the gland cells and even of
the stalk cells betray evidence of absorption in the presence of a greater
density of the protoplasm and the presence of large rounded dark
masses. These changes are not to be observed in the stalked glands,
which do nothing else than secrete mucilage. I attempted to prove
the matter for myself at Perth, W. AustraHa. Byblis appeared late in
the season, during the latter part of my visit, so that I had only lim-
ited time at my disposal. My method consisted in placing minute
fragments of carmine fibrin in contact with the glands of the living
leaf, on the plant, and in a small vial with a dozen short pieces of leaf
with and without a little added water, with and without added weak
HCl, and with and without ammonium nitrate. The results were en-

Francis E. Lloyd —98— Carnivorous Plants

tirely negative, even after two weeks, though there was at length a dis-
tinct and unpleasant odor emitted.

I first learned from Mr. A. G. Hamilton that Byblis harbours a
small insect which he called a ''buttner". In Perth I received the
same information from Mr. H. Stedman, who kindly took me to a
locality at some distance north of Perth where we found a lot
of plants growing. All of these were infested with a small wingless
capsid which turns out to be a new genus and will be described by
Dr. W. R. China of the British Museum (Natural History) {13 — i).
While small insects in general are caught by the mucilage secreted by
the stalked glands, this capsid moves about freely without difficulty,
just as do similar insects, also capsids, over the surface of Drosera
leaves in Australia, and of the African genus Roridula, once thought
to be carnivorous. How the insect manages this is a bit puzzling. It
is noticeable that it prefers to walk on the upper leaf surface where
there are very few and usually smaller glands but when alarmed it
progresses rapidly in any direction without becoming entangled with
the mucilage. Full sized insects are perhaps too big to be readily en-
cumbered, but the smaller ones move about just as freely. Their food
consists of freshly captured flies, the juices of which they suck, the re-
lation of insect and plant affording a sort of commensalism, but this
term could hardly be used in the case of Roridula (non-carnivorous)
the secretion from whose glands is resinous (Lloyd 1934).

Literature Cited:

Bruce, A. Ninian, On the activity of the glands of Byblis gigantea. Notes Roy. Bot. Gar-
den Edin. 16:9-14, 1905, also 17:83, 1907.
DiELS, L., Byblidaceae, Nat. Pfianzenfamilien. i8a. 1930.
DoMiN, K., Byblidaceae, a new archichlamydeous family. Contr. to the Australian flora,

undated, but about 1920. Extracted from MS. and published separately in Acta Bot.

Bohem. 1:3-4, 1922.
Fenner, see under Nepenthes.
Hamilton, A. G., Notes on Byblis gigantea. Proc. Linn. Soc. New South Wales 28:680-

684, 1903.
Lang, F. X., Untersuchungen iiber Morphologie, Anatomic und Samenentwickelung von

Polypompholyx und Bvblis gigantea. Flora 88:3-60, 1901.
Llo\T), F. E., The tannin-colloid complexes in the fruit of the persimmon. Biochem. Journ.

1:7-41 (pi. 1-3), 191 1.
Lloyd, 1933 (see under E eliamphora) .
Lloyd, 1934 {see under Introduction).
Planchon, J. E., Sur la famille des Droseracees. Ann. sci. nat. bot., 3 ser., 9:79-99, 1848.

(Contains also descriptions of Drosera carpels bearing tentacles, these being intergrades

between normal leaves and carpels).
Ross, H., Byblis gigantea. Gartenflora 51 :337-339 (pl- 15°°), 1902.

Chapter VIII

DROSOPHYLLUM LUSITANICUM

Drosophyllum lusilanicum Lk. (jj — 2) is a plant with much the ap-
pearance of Byblis, but it is larger and shrubbier (1-1.6 m. tall) and is
unusual, for the carnivorous plants, in growing not in a wet, but in a
very dry habitat in Morocco and nearby Portugal and Spain. Harsh-
BERGER visited a locahty in Sra. de Valongo near Oporto, where he
found Drosophyllum growing in open formations, scattered over the
quartz-rocky soil. He observed its leaves to be crowded with small
gnats. Its flowers are bright sulphur yellow, are i-i>^ inches in
diameter, and have convolute aestivation. It is called locally "herba
piniera orvalhada" (dewy pine) in allusion to its bedewed appearance
due to the numerous glands carrying large droplets of clear mucilage.
The base is strongly woody, and its abundant roots penetrate deeply
into the dry soil. "Mr. W. C. Tait informs me that it grows plenti-
fully on the sides of dry hills near Oporto, and that vast numbers of
flies adhere to the leaves. The latter fact is well known to the vil-
lagers, who call the plant the 'fly-catcher,' and hang it up in their
cottages for this purpose" wrote Darwin (1875). Inquiry by corre-
spondence with Dr. QuiNTANiLHA has elicited doubt of the correctness
of Tait's statement as to the use by the paisanos of the plant as a
fly-catcher, though it seems reasonable enough.

The leaf is linear with a deep furrow along the upper side. It
is traversed by three vascular bundles, a median and two lateral, arising
from a single bundle entering at the base {14 — 5).

A peculiar feature is found in the reverse circination (14 — 4) the
rolled leaf-tip facing outwardly while in Drosera very generally the
opposite holds. Although in Byblis gigantea the leaves are nearly
straight, showing no evident circination, in Byblis linifolia the be-
havior is like that of Drosophyllum. Fenner expresses the opinion
that this arrangement has its significance in permitting the free de-
velopment of the stalked glands, but he overlooks the fact that the
circination of Drosera is in the opposite sense without any prejudice
to the development of the tentacles. The case of Byblis linifolia was
not known to him. In any event, in the tight coils the dorsal and
ventral leaf surfaces are mutually compressed; and assuming that the
tentacles (hairs in the case of Byblis) develop after uncoiling, the ven-
tral (upper) surface is freer than the dorsal, where the most of the
tentacles or hairs are to be found.

Another characteristic behavior of the leaves is their marcescence.
Instead of falling away as they die, they remain attached, forming a
grass-skirt about the stem. Franca (1922) regarded this as a symptom
of a condition which he regarded as pathological, due to overnutrition
and the inability, because of the absence of an excretory apparatus, to
throw off waste. Quintanilha, however, disagrees with this and, in
our opinion, justly.

Francis E. Lloyd — 100 — Carnivorous Plants

In the seedling, the cotyledons withdraw from the seed during
germination and develop into broadly linear tapering members, sup-
plied with glands enabling them to capture prey (Franca).

The leaf bears two kinds of glands, stalked mucilage glands and
sessile digestive glands {14 — i, 2, 6). Their position is determined if
at all only to a sHght extent by the three vascular bundles, from
which, however, they receive branchlets ending at the bases of the
glandular tissues. There are three double files of stalked glands, one
along each leaf margin, roughly speaking, and two rows along the
under leaf surface, one on each side of the midvein. The sessile glands
are more scattered, and apparently only in some degree determined in
position by the vascular tissues. Sessile glands occur on both upper
and under leaf surface, stalked glands only on the under surface and
along the margins.

Structure of the glands. — Drosophyllum differs from Byblis in that
the glands, instead of being trichomes, are emergences, and, as Darwin
pointed out, have much the same structure as those of Drosera, with-
out, however, being endowed with the power of movement. This
refers of course to the stalked glands. These have a stout stalk sur-
mounted by a large nearly hemispherical capital, and, as Darw^in put
it, have the "appearance of miniature mushrooms."

The capital {14 — 2) consists of three courses of cells running
parallel with the outer surface. The outer of these, the epidermis, is of
rather thick, wavy-walled cells, with strong buttress thickenings, stiffen-
ing the angles of the radial walls {14 — 3). The dense protoplasmic
contents and prominent nuclei speak for their glandular activity.
These are covered with a cuticle, which according to Fenner is finely
porous, thus permitting the exudation of the mucilaginous secretion.
I have not succeeded in convincing myself that the pores are optically
demonstrable, but it is certain that the cuticle offers no impediment to
the diffusion of methylene blue, for less than a minute's exposure to a
watery solution of this dye results in the deep staining of the whole
capital while the dye does not penetrate the remaining epidermis at
all. Meyer and Dewevre also failed to see the pores but demon-
strated on kiUing the escape through the cuticle of the pigment which
renders the gland conspicuous. They found also that lithium nitrate
taken up through the roots is found 12 hours later in the mucilaginous
secretion. The cells of the second course underlying the epidermis are
somewhat more irregular in form, but are likewise provided with but-
tress-thickenings in the radial walls, though they are not so numerous
and prominent as in the epidermal cells. The general character of
these two courses is the same; they were called, by Penzig (1877), the
secretion-layer. Underlying these two courses is a third, of flat cells,
of greater size in the transverse direction (with reference to the axis of
the gland) with their contiguous radial walls strongly cuticularized, so
that in a cleared preparation when suitably stained, as with congo red,
one sees a strong network lying within the capital. Contrary to an
earlier view (Solereder 1899, p. 367) not the entire but only the
radial walls are cuticularized, thus (Goebel 1891) leaving a free
diffusion passage. This feature is held in common with other gland-
ular structures described elsewhere.

Chapter VIII — 101 — DrosophyUum

The third layer (limiting layer of Penzig) caps a mass of short
irregular tracheids constituting the expanded end of a strand of vas-
cular tissue extending through the stalk and communicating with the
vascular tissues of the leaf. This strand consists of both xylem and
phloem elements (Fenner contra Meyer and Dew-eyre) affording,
according to Fenner, not only a pathway for water but, in the case of
the phloem, for the transmission of stimuli to the neighboring sessile
glands, which have been shown to show secretory activity in response
to such stimulus received from the stalked glands. The stalk itself is
made up of the epidermis and an underlying course of parenchyma, sur-
rounding the vascular strand. The capping secreting cells contain
brilHant red coloring matter, interpreted as an optical lure for insects,
and when the capital bears its shining droplet of clear mucilage, which
acts as a Hght collecting lens, the glands appear as brilUiant red dots.
The sessile glands have no such coloring matter. These {14 —6) have
the same structure as the stalked glands, differing only in the absence
of the stalk. Occasionally an intergrading condition is met with;
Goebel found one such, with a very short stalk. The sessile glands
are usually oval, generally smaller, and have a less extensive contact
with the vascular system. Each gland, however, is underlaid by a
group of cavernous looking tracheidal cells, with no protoplasmic con-
tent, evidently an important part of the gland but with what function
we do not know. If Fenner saw this feature, he regarded it as the
end of the tracheidal system. For there is also to be found at the base
of each gland the end of a branch of the vascular system. These
glands are devoid of a mucilaginous secretion, as of coloring pigment
and even of chlorophyll, for they appear whitish.

The mucilage secreted by the stalked glands is peculiar, in that it
is not readily drawn out into slender viscous threads, but is easily
pulled off the gland by a touch of even a needle point as Darwin ob-
served. "From this peculiarity, when a small insect alights on a leaf
of DrosophyUum, the drops adhere to its wings, feet or body, and are
drawn from the gland; the insect then crawls onward and other drops
adhere to it; so that at last, bathed by the viscid secretion it sinks
down and dies, resting on the small sessile glands with which the sur-
face of the leaf is thickly covered" (Darwin, 1875, 2nd ed., p. 271).
The secretion of mucilage continues after removal and Darwin
found that when a plant is kept under a bell glass to prevent evap-
oration the secretion is produced in such quantities as to run down
the leaf surface in droplets; and further that the secretion shows an
acid reaction. Goebel found that among the possible acids pres-
ent formic acid is one, and believed that this is effective in pre-
venting bacterial action. Emanating from these glands, probably, is
an odor which Goebel likened to that of honey, which would be at-
tractive to insects and thus act as a lure.

In the case of many carnivorous plants "overfeeding" usually re-
sults in the damage and death of the leaf wholly or locally, notably in
the pitcher plants. This has not been observed to occur in Drosophyl-
lum, and may be accounted for by the inhibition of bacterial action as
just indicated.

The sessile glands do not exude a secretion unless stimulated

Francis E. Lloyd — 102 — Carnivorous Plants

(Darwin, Goebel, Fenner, Quintanilha). The secretion appears
normally when the mucilage glands are stimulated by the catching of
prey, but not merely mechanically, as by placing on them sand grains,
bits of paper, etc. Fenner showed in considerable detail by appropri-
ate experiments that the maximum activity of the sessile glands is ob-
tained when, after the stalk glands nearby have received prey, both
prey and mucilage secretion are brought into contact with them. But
in the presence of mucilage removed from the stalked glands and mixed
with the juices of prey, leaving the stalked glands unstimulated, the
sessile glands work only slightly if at all. Fenner concluded that the
maximum activity of the sessile glands is called forth by something
passing through the tissues by way of the vascular elements (phloem).
The sessile and stalked glands must, therefore, be considered as a single
mechanism in which one part is dependent on the other.

There is a general agreement on the part of the authors mentioned
that Drosophyllum exercises its own proper power of digestion, and that
this is not the result of bacterial activity. As mentioned already,
Goebel regarded digestion as too rapid for bacterial action, and that the
presence of formic acid excludes such activity, and though he was
unable to state the concentration of acid present, he supports his in-
ference by inoculating nutrient gelatine plates with negative results.
The activity of formic acid may not, however, Goebel adds, be con-
fined to that of an antiseptic, but it may consist in an initial dis-
solution of the proteins of the body of the prey, with the escape of
materials v/hich then affect the sessile glands and stimulate them to
greater activity.

Darwln found that fragments of egg albumen, fibrin, were acted
upon rapidly when they came in contact with the sessile glands. If
only in contact with the stalked glands, they were not attacked. If
then placed on the sessile glands, there was a copious secretion, and
the albumen was completely dissolved in 7 to 22 hours. "We may
therefore conclude, either that the secretion from the tall glands has
Httle power of digestion, though strongly acid, or that the amount
poured forth from a single gland is insufficient to dissolve a particle of
albumen which within the same time would have been dissolved by the
secretion from several of the sessile glands." Fibrin likewise, when
placed on the stalked glands, was not attacked, though, as in the case
of albumen, the secretion was absorbed (together with whatever es-
caped into it from the fibrin). But when the fibrin was slipped onto
the sessile glands, digestion proceeded rapidly (17 to 21 hours) with an
abundant exudation of fluid from the glands. Darwin thought the
digestion more rapid than in Drosera. He had not excluded the action
of bacteria, which, however, as above said, Goebel did by suitable
culture experiments. He observed a more rapid action than did
Darwin. A fibrin flock i cm. long and one-fourth the width of the
leaf was noticeably attacked in a half-hour on a warm summer day,
and in an hour no trace could be seen, though the spot had been care-
fuUy marked by a bit of paper. With a lens small fragments could
still be seen. A true digestion, he concluded, is therefore present. The
enzyme is secreted in response to a special stimulus, and chiefly, if not
exclusively, by the sessile glands. The stalked glands are chiefly a
trapping apparatus (Goebel 1891).

Chapter VIII — 103 — Drosophyllum

In 1894 came Meyer and Dewevre. They managed to collect 1.6
grams of mucilage and investigated it. It was stiff, clear, had the odor
of honey and was strongly acid. It contained no free reducing sugar,
but on heating with HCl it reduced Fehling, and gave a weak red
color with thymol and H2SO4 (indicating polysaccharides). The
presence of a sugar was indicated by a yellow coloration with chlor-
zinc-iodide as also its precipitation by lead acetate, by barium hydrox-
ide and by alcohol. No proteins were present. It was poor in salts,
only Ca being present. No K, phosphates or nitrates were found.
The acidity was due to a non-volatile acid, and not to formic acid, as
Goebel had held. These authors verified Darwin's observation that
the sessile glands secrete only on stimulus by a protein. Insects are
attracted both by the odor and by the gHstening of the droplets of
mucilage. They recorded observations also which indicate that there
are two periods of activity in the plant, (i) From the beginning of
vegetative activity to the beginning of seed ripening (from Jan. 15 to
May 15, in the greenhouse). During fruit ripening the leaves begin
dying from apex to base and the glands do not secrete vigorously. The
soil must be kept "dry" during this period. (2) After fruit ripening is
complete (Aug. i to Oct. 15) secretion and odor are both strong, es-
pecially in sunny weather. The experience of Darwin was again sub-
stantiated in finding that coagulated egg albumen, meat and fibrin
were acted upon, especially if well smeared with mucilage and placed
on the sessile glands. The time necessary for complete digestion was
about the same as in Darwin's experience. Goebel's figures were
criticised as being too low but an error of proofreading may have
crept in, rather than, as is suggested, incorrect observation. It was
found that very small fragments of albumen were attacked in the
mucilage of a stalked gland and completely digested in 7 days. If large
bits were imposed, they absorbed the mucilage, and damage might re-
sult to the gland in consequence. No diastase was found. Bacteria
were never found and it was clear that the mucilage, as Goebel said,
is antiseptic.

Franca (1925) gave a general account of the plant, and studied
especially the cytological changes which he observed in the glands
during digestion and absorption, using both Hving material viewed
microscopically, and fixed material stained with iron haematoxyhn and
fuchsine, etc. He found evidence that the two courses of the glandular
cells (the outer and second) have different functions, that the outer
course is secretory only, the inner both secretory and absorptive. The
sessile glands have only the power of secretion. This evidence con-
sists in the cytological appearances observed during digestion and ab-
sorption. Changes of the bright red color in the glands to a deeper,
much darker shade, had been noted by Darwin. When such glands
are examined, the cells of the outer layer of the gland are seen to have
remained unchanged, while those of the second layer are now charged
with large black granules. These rise to a maximum some hours after
the glands have been suppKed with muscle fiber. Some similar granules
are found also in the more distal short cells of the stalk of the gland
and finally in approximate leaf tissues. Such dark granules are seen
when an insect has been captured, but only in the deeper gland cells.

Francis E. Lloyd — 104 — Carnivorous Plants

With neutral red the protoplasm displays a great number of small red
granules, considered to be granules of secretion. On the other hand
the deeper cells are filled with voluminous granulations of dark red
color. In this way it is supposed that the power of absorption is
demonstrated, for when prey has been captured, the superficial cells
show only the small granules, while the deeper cells and the distal
stalk cells are at the same time found crowded with grey or black
granules in addition to the secretory granules seen also in the outer cell
layer. The recounted facts are held to support Fran^a's conclusion
that the outer glandular layer of cells is secretory only and the inner
layer both secretory and absorptive. Additional and supporting evi-
dence is found by Franca in the presence of canalicuh. Some occur
"in the thickness" of the membranes between the cells of the two
layers and open on the outside of the gland by means of minute oval
mouths. It is these which permit the entrance of absorbed substances
to the deeper cell layer. Others occur in "the thickness" of the but-
tresses of the epidermal cells (between which fingers of the protoplasm
project, as described by Haberlandt for Drosera), and these "without
doubt" permit the escape of secretion. I have carefully examined
preparations after treatment with II2S04, followed by Sudan III and
have been unable to find any evidence of pores. The evidence in the
form of a drawing in his plate has httle convincing effect.

A critical study of Drosophyllum was undertaken by Quintanilha
at Coimbra. His results, pubHshed in 1926, briefly stated are as fol-
lows. Drosophyllum is indeed a carnivorous plant, acting by means of
a proteolytic ferment of the type, of animal pepsin. A mosquito can
be completely digested in 24 hours. Bacterial digestion does not
enter into the picture (in this agreeing with Goebel, whose experi-
ments were repeated and verified). The stalked glands are essentially
organs of capture but at the same time they are "signales d'alarme";
that is, on capture of an insect, they send a stimulus to the sessile
glands and provoke their activity. These are exclusively organs of di-
gestion and absorption, but they act only on stimulation. Experimen-
tally and under favorable conditions, the stalked glands may digest
very small amounts of albumen without the intervention of the sessile
glands, but the proteolytic properties of the mucilage are always in-
significant. Experimentally it was shown also that the sessile glands,
when previously excited, can digest and absorb albumen without the
intervention of the stalked glands which had been removed by ampu-
tation, and in this condition the absorption is as rapid, or even more
rapid, than it would be in collaboration with the stalked glands be-
cause of dilution of the secretion. In the normal state the stalked
glands act as traps and furnish stimuli to the sessile glands.

Excitation of the sessile glands can be procured directly by chemi-
cal but not by mechanical means and indirectly by both means. Simple
pressure or friction of the stalked glands does not procure excitation
of the sessile. However, the cutting off of the glands from the stalks
can excite indirectly and mechanically the sessile glands. The ex-
citation is slow of transmission and is limited to an area of about i cm.
from the tentacle stimulated. On anatomical grounds Quintanilha
inclines to believe with Fenner that the phloem of the vascular system

Chapter VIII — 105 — Drosophyllum

serves to transmit the stimulus. The digestive capacity of the plant
is reduced after fructification.

The same author studied the cytological concomitants of the di-
gestive and absorptive activity, and his findings are of interest in con-
nection with those of Homes and others on aggregation in Drosera
and with those of Franca above given. He found that the state of
aggregation can be procured independently of digestion. During
digestion there occur "black concretions" in the inner course of cells
of the stalked glands. These are not the "spherules alimentaires " of
Franca, but intravacuolar precipitation of anthocyanin compounds.
In the sessile glands, however, concretions appear in the cells derived
from the absorption of albumins impregnated with melanin. The
chondriome of the glandular cells does not act directly in the elabora-
tion of proteolytic enz3rmes and it does not present alterations which
allow us to attribute to them an important role in the phenomenon of
digestion. Only in the internal secretory layer of the sessile gland
the elements of the chondriome are considerably reduced in volume
during intracellular digestion. The number of the chondrioconts is also
reduced and that of the mitochondria is increased proportionally.

On the other hand, the vacuome appears to be the seat of the
elaboration of ferments and certainly has an important role in the
processes of digestion and secretion.

Pathological conditions in the plant due to overfeeding have not
been observed. It is clear that Drosophyllum profits by food materials
supplied by animals, and that this compensates for an insufficient
mineral nutrition, Quintanilha says in general conclusion.

Literature Cited:

Darwin, C, Insectivorous Plants. 2d. ed., London 1875. 1908 reprint.

Fenner, C. a., Beitrage zur Kenntnis der Anatomic, Entwickelungsgeschichte und Biologic

der Laubblatter und Driisen einiger Inscktivoren. Flora 93:335-434, i904-
Fernandes, Ab{lio, Morphologia e biologia das plantas carnlvoras. Anuario da Sociedade

Brotcriana 6:14-46, 1940; 7:16-52, 1941. A third part appeared later in 1941; the

whole was issued as a brochure, repaged, in 1941. Good photographs of Drosophyllum

and of its habitat.
FR-A-Nf A, C, La question des plantes carnivores dans le passe at dans le present. Bol. Soc.

Broteriana I (2 ser.):38-57, 1922.
FRANfA, C, Recherches sur le "Drosophyllum lusitanicum" et remarques sur les plantes

carnivores. Arch, portug. d. Sci. biol. 1:1-30, 1925.
GoEBEL, K., Pfianzenbiologische Schilderungen. Marburg 1889-1891.
Harshberger, J. W., Notes on the Portuguese insectivorous plant, Drosophyllum lusitani-

cum. Proc. Amer. Philosoph. Soc. 64:51-54, 1925.
Meyer, A. & A. Dewevre, tjber Drosophyllum- lusilanicum. Bot. Centralbl. 60:33-41,

1894.
Penzig, O., Untersuchungen iiber Drosophyllum lusitanicum. Diss. Breslau, 1877.
Quintanilha, A., O problema das plantas carnivoras. Dissertation Coimbra, 1926, 88 pp

(Contains a very full bibliography of the Uterature pertinent to carnivorous plants.

French resume). Extr. from Bol. Soc. Brot. 4.
Solereder 1899 {see under Dionaea).

Chapter IX
PINGUICULA, BUTTERWORT

Distribution. — General appearance. — Habitat. — The leaves. — Two kinds of glands
(Points of structure. Early work of Darwin: movements. Secretion and digestion). —
Popular uses.

The genus Pinguicula consists of about 30 species distributed
throughout the northern hemisphere in temperate or cool temperate
regions. Although making use of a far different mode of capture of prey,
it is closely related to Utricularia and Genlisea, and is one of the three
lentibulariaceous genera, as shown by the flower structure. The
personate corolla is blue, purple or yellow, and differs from that of
Utricularia principally in the five-parted cal30c.

All the species are of very uniform character. The plant consists
of a short vertical stem giving rise to a compact rosette of leaves which
usually lie flat on the ground, or in some species {P. gypsicola) are
directed obHquely upward also. They exhale a distinct fungus-like
odor. The tissue tensions in the leaves are such that when a plant is
uprooted from the soil they become at once strongly reflexed, as
Darwin observed; but this is a feature common to rosette plants.
The leaves are entire, usually ovate {P. vulgaris) or broadly ovate {P.
cuneata), with upcurled margins. In color they are a pale faded green,
yellowish in bright light (Batalin), deeper green in the shade, in P.
vulgaris pale purple due to the presence of pigment in the lower epider-
mis. They are very soft and yielding, easily bruised and torn; and,
being "greasy" to touch, the name, derived from the Latin pinguis,
fat, was suggested, according to accounts. The dorsal surface is quite
smooth and shiny, the ventral ghstening with myriads of minute
mucilage glands. In addition to their glands, both surfaces bear num-
erous stomata peculiar in having no chlorophyfl, though there is pres-
ent according to Batalin a pale yellow pigment. The flowers are
borne singly on slender, glandular, pubescent scapes, have a five-parted
corolla, with a slender spur, so large and showy in some species of the
genus that they are found in glasshouse cultivation. Although the
peduncles also have glandular hairs, Darwin thought them devoid
of digestive function. The seedlings have a short taproot possessed
of a few root hairs, but this does not persist and soon gives way to
adventitious roots arising from the stem above. In possessing a tap-
root, even though fugacious, this genus differs from the others of the
family, in which there is none. There is but one cotyledon, which
arises as a semicircular ridge around the plumule, and when fully
developed is strongly folded lengthwise and may in longitudinal
sections be easily interpreted as two, as Goebel pointed out.

Pinguicula grows in wet places, with mosses, etc., in chinks of wet,
dripping rocks, on hummocks in swamps {ij — 4) and similar situa-
tions, in general conformity with the majority of carnivorous plants.
Towards the end of the growing season the plant produces very com-

Chapter IX —107 — Pinguicula

pact buds of various sizes (brood-buds) which can reproduce the plant
in the following growing season (Hovelacque).

The entire leaves and peduncles are provided with two kinds of
glands, stalked and sessile (jj — lo, ii), densely scattered on the
upper surface, with a much smaller number of sessile glands with four-
celled capitals on the lower dorsal surface (75 — 2-4). According to
Fenner the latter are hydathodic in character for he observed a
minute droplet of fluid water, presumably, on each gland. They may
safely be excluded from taking part in the capture and digestion of
prey. Goebel had thought their secretion to be mucilaginous but
this seems not to be the case. All these glands are of epidermal origin
(Gressner, 1877; Fenner, 1904). The stalked glands of the upper
surface stand on an epidermal cell, the stalk cell displaying a marked
entasis, ending in a single short domed cell supporting the capital com-
posed of 16 radiating cells. This secretes and supports a globule of
stiff mucilage which serves to entrap and smother the prey, which
must be small — only small insects are effectively caught — such as
aphides, minute flies of various kinds, etc. The sessile glands have a
similar structure, but the stalk cell is not cut off from the foundation
epidermis cell, and there is no elongation of it. The base of the gland,
therefore, lies flush with the general surface. The capital has only
eight cells. The sessile glands of the under surface have capitals with
only four cells. All these have been described by Fenner. This in-
vestigator further adds that some four rows of cells along the very
thin leaf margin are also glandular, and that these secrete mucilage.
The margin is of only three cells in thickness, a single course of cells
being embraced between the two epidermes (75 — 5). It is always
curled upwards through about 180 degrees, and this has been inter-
preted as an adaptation for conserving the digestive fluids which escape
from the glands on stimulation. Fenner believes also that the escape
of secretions from the glands is made possible by the occurrence of
pores in the cuticle. I have not been able to see them, but treatment
with methylene blue proves the easy passage of solutes, for if a leaf is
plunged into a solution of medium strength the capitals of the glands
are almost immediately and deeply stained. The capitals of the stalked
glands are also stained but not so quickly as those of the sessile glands,
perhaps because of the presence of mucilage. With regard to the struc-
ture of the cells along the margin of the upper surface, I can see no
cytological evidence, such as claimed by Fenner, that they are glandu-
lar, nor have I seen a band of mucilage as described by him. J. R.
Green (1899, p. 214) cites Darwin to the effect that Pinguicula
secretes a digestive fluid on the edges of the upper surface of the leaf
which folds over to enclose its captive. On perusing Darwin's account
I am unable to subscribe to Green's statement. True, Darwin does
use the expression ''placed among one margin" or "on one margin" but
this was not meant to indicate that when secretion occurred it was
confined to the margin, but that the nearby stalked glands contributed.
Drops of meat infusions could not be confined to the margin without
coming into contact with nearby glands. Darwin in his first set of
experiments was concerned with the possibility of leaf movement which
he demonstrated to his own satisfaction. In his experiment on di-

Francis E. Lloyd —108— Carnivorous Plants

gestion he invariably placed the substrate to be acted on "on the
leaf", and I think it is quite evident from the context that Darwin
did not think of the margin of the leaf as having a localized digestive

action.

Pmguicula was first studied and shown to be carnivorous by Dar-
win. "I was led to investigate the habits of this plant by being told
by Mr. W. Marshall that on the mountains of Cumberland many
insects adhere to the leaves" {Insectivorous Plants, p. 297). He noted
the presence of two kinds of glands, sessile and stalked, later studied
carefully by Fenner. Having studied Drosera extensively Darwin
first looked for and discovered movements of the leaves. In a se-
ries of 17 experiments small flies, or portions of larger flies, smaller
and larger fragments of meat, meat juice stabilized in small bits of
sponge, even fragments of glass were placed in various positions in
rows parallel to the margin, near the apex, and along the midrib, and
he found curvatures of the leaf margin to occur within periods of a few
(2-4) hours, to increase for some hours and finally to disappear. The
apex of the leaf never shows motion, this being confined to the margins.
He found evidence leading him to believe that the stimulus could be
transmitted to a distance of about 6 mm. (his exp. 13). A weak so-
lution of ammonium carbonate caused marked incurvation of the leaf
margin in 3.5 hrs., a stronger solution (i to 218 H2O) causing no move-
ment, probably due to damage. Mechanical irritation of the leaf
surface either before or after the apphcation of meat juice, thus im-
itating the actions of dying prey, did not hasten or increase the re-
sponse. The effect produced by fragments of glass was as rapid as
that following the application of nitrogenous substances, but the de-
gree of curvature was less. The substances used other than glass in-
cited a more or less copious flow of secretion.

Darwin commented on the brevity of the response action, there
being a complete restoration of form within 24 hours. He was thus
prompted to doubt the usefulness of the behavior, but ventured the
idea that the infolded margin could prevent the washing away of prey,
as in fact was observed by a friend of Darwin in Wales. If the prey
is large the infolding leaf margin pushed it further toward the middle
of the midrib, thus bringing it into contact with more glands, an effect
comparable to the action of the tentacles in Drosera. The margins of
the leaf are always curved up, and this Darwin thought to help to
conserve the fluids from loss, keeping them on the leaf surface to be
absorbed. Goebel could not substantiate Darwin's conclusions about
the sensitivity of the Pmguicula leaf, his experimental results being
mostly negative. On the other hand, Fenner, one of Goebel's
students, did find sHght movements on the application of fragments
of glass, followed by quick recovery. The secretion of mucilage is
thereby excited. When an insect falls on or near the leaf margin, an
abundant secretion foflows, overwhelming it. This escape of fluids
from the leaf alters the tensions and this results in the inrolhng of
the leaf margin which does not occur in older mature leaves. When the
insects sink down to the leaf surface and come into contact with the
sessile glands (75 — 11), an acid secretion of greater viscosity and con-
taining a digestive enzyme escapes from these. Goebel had shown

Chapter IX —109— Pinguicula

that the abundant mucilaginous secretion following application of
granules of sugar is without digestive power.

Having cultivated material of P. vulgaris collected in the mountains
of California east of Crescent City, I repeated such experiments as
done by Darwin, Goebel and Fenner on about a dozen leaves, with
definitely positive results. I cite only one as typical, this being il-
lustrated in 15 — I, see also 13 — 6. The total activity extended over
more than six days. Four minute flies were observed caught in a row
parallel to one margin and two similarly placed with respect to the
other margin. Already the one margin was slightly curved upwards
on Oct. 2, the other showed no motion until the night of Oct. 3-4. On
the morning of Oct. 4 both margins were well curved, enough to hide
all the flies. On Oct. 6, the inward rolHng of the margins was well
developed, and next day it had begun to recede, again exposing the
flies to one's vision. This behavior was typical of the whole series of
cases. This and a number of other cases observed seem to throw doubt
on the vaHdity of Darwin's statement that the time leaves remain
incurved, even though the exciting objects remain in position, is but
short, i.e., not more than twenty-four hours. It is further well known
that the contact of an insect with the leaf at a point removed from the
margin, i.e., near the midrib, results in the dishing of the leaf below
the insect (Darwin, Batalin). This, as Batalin suggests, is the
same phenomenon as observed in Drosera, and must be attributed to
growth and not to injury as Darwin supposed. When flies are ar-
ranged along and more or less parallel to the leaf margin the growth
results in the rolling of it. There is Httle doubt of the correctness of
this explanation; and moreover it agrees with our knowledge of the
procedure in Drosera and Dionaea.

Movement in Pinguicula is then an undoubted fact. How much sig-
nificance may be attached to it is a question. Goebel attached Kttle.
Darwin thought that the rolling of the leaf margin brings more glands
into contact with the prey, and in some cases pushes it into new posi-
tions further away from the margin. Darwin probably underesti-
mated the persistence of the change in movement, and therefore its
importance. The upward curved leaf margins help to hold the se-
cretion in place. This is probably as much as we can say about the
matter.

Darwin then turned his attention to the question of secretion and
digestion. He found that when he placed prey (small flies), fragments
of meat, cartilage, fibrin, albumen (egg-white, coagulated), gluten and
gelatin, etc., on the leaf surface, there was an increase of secretion,
often copious, and that this was acid. Evidence of digestion was
clear: insects fell apart readily, and other substances showed the ex-
pected signs of disintegration. Objects not containing soluble nitrog-
enous matter, or other soluble matter do not excite secretion. Non-
nitrogenous fluids can cause free flow of the secretion, but this remains
neutral (non-acid). Among the substances or objects which incite
acid secretion were small leaves {Erica tetralix), pollen and various
seeds, all often seen to adhere to leaves in the open, aU, of course, con-
taining nitrogen from which Darwin argued that these objects also, as
well as animal prey, help to nourish the plant. Since the peduncles are

Francis E. Lloyd —HO— Carnivorous Plants

equally glandular with the leaf, and since the life of a peduncle is fully
a month or more, whatever benefit may be derived from prey caught
by leaves may also be said to accrue from that caught by the peduncles.

Cytological changes. — Darwin examined the condition of the glan-
dular cells after being in contact for some time with sources of matter
which was plainly absorbed, and found evidence of change in structure
and appearance of the protoplasm and its content, usually in the ap-
pearance of granular matter colored brownish, or in the cell contents,
at first limpid, being aggregated into slowly moving masses of proto-
plasm. The difficulties of observation and inference are obviously
great, a great deal more so than in the case of Drosera. Darwin re-
ferred the appearances to the absorption of food materials.

NicoLOSi-RoNCATi (1912) endeavored to relate cytoplasmic changes
observed in fixed and stained material to secretive activity, in P.
hirtifiora. In actively secreting glands (mucilage glands presumably),
the cytoplasm is vacuolated and contains many fuchsinophile granules
scattered toward the periphery of the cell with moniliform bodies in
the vicinity of the nucleus. The nucleolus, large and intensely fuch-
sinophile at the beginning of secretion, diminishes notably in volume
and in capacity for staining in evidently secreting cells. The author
concluded that the first impulse to secretion comes from the nucleolus,
the primary granules of secretion being formed by the chromatin.
These diffuse throughout the body of the cell definitively elaborating
secretory substance. This work, while affording a beginning, does not
lead us very definitely forward, as at this time we are unable to dis-
tinguish the kind of secretion dealt with, whether of mucilage or en-
zymes.

TiscHUTKiN (1889) carried out experiments similar to those of
Darwin, and worked also with glycerin extracts of leaves and mix-
tures of leaf secretion, withdrawn by means of a pipette, with glycerin,
acidified variously (HCl, formic, mahc acids). Both glycerin extracts
and mixtures gave only negative results. Albumen, gelatin and
fibrin placed on the leaves gave results for him much the same as for
Darwin. Tischutkin states then that in Pinguicula insects which
are caught call forth a secretion of acid sap which can procure a cer-
tain alteration of their substance. Examining the work of Rees,
Gorup and Will (later substantiated) he sees in its deficiencies ev-
idence of bacterial action and he comes to the conviction that the role
of the plant is the secretion of a medium which is suited to the hfe
and activity of microorganisms (bacteria), and concludes without fur-
ther experimental evidence that in Pinguicula we are deaHng with
bacterial action, in this agreeing with Morren (1875).

Somewhat later Goebel also attacked the problem of digestion in
Pinguicula. When he put particles of fibrin on the leaves, the secre-
tion was intensified, and the smallest particles digested in 24 hours.
The secretion was weakly acid. When insects (those, as already said,
are always small, for Pinguicula is adapted to the capture and di-
gestion of only small ones) are found in an advanced stage of digestion,
the glands are found to contain droplets of fat. Large insects or fibrin
fragments are overcome by decay. By ad hoc culture experiments
Goebel showed that Tischutkin's views are not justified. He showed

Chapter IX — 111 — Pinguicula

that when even small flies, partly digested, were transferred to nutrient
gelatin plates, no evidence of bacterial activity was forthcoming. He
convinced himself, on experimental evidence, that Pinguicula secretes
an antiseptic substance which prevents bacterial action, and, while his
procedure cannot be regarded as beyond criticism, yet it is to be
noted that later Loew and Aso (1907) claimed to have found benzoic
acid in the leaves. Naturally the amount present is not sufficient to
meet all conditions, since in nature the Pinguicula catches only minute
flies, and only small amounts of the antiseptic agent are called for. In
Tischutkin's experiments, says Goebel, he used too large masses of
material with erroneous results. The capacity of the stomach to digest
cheese, he added, cannot fairly be judged by feeding a kilo of cheese at
one time.

Goebel made an experiment which seems to distinguish between
the action of the sessile and stalked glands, substantiating Darwin's
findings that "non-nitrogenous fluids if dense cause the glands to pour
forth a large supply of viscid fluid, but this is not in the least acid.
On the other hand the secretion from glands excited by contact with ni-
trogenous soHds and fluids is invariably acid ". Tischutkin had

tried to extract leaves by placing them in glycerin, with negative results.
By strewing granular cane sugar on the leaf surface of some 70 plants,
Goebel collected about i cc. of secretion which was neutral and after
the addition of 0.2% formic acid a particle of fibrin remained in it
undigested at 35° C. From this it appears that an abundant se-
cretion (probably from the stalked glands) is not necessarily correlated
with digestive activity. On the other hand if leaves are stimulated by
strewing particles of fibrin, smeared with meat juice and finally placed
in meat juice, with 1.5 Tc formic acid added and allowed to stand for
18 hours, a fluid was obtained which digested swollen fibrin in 25 hours.
No bacteria were present, due to the hindering action of the formic
acid. In any event, the amount of enzyme obtainable is smafl.

In the foregoing it will be seen that the conclusion that Pinguicula
is a true carnivorous plant rests on the evidence that fragments of
nitrogenous matters and insects are disintegrated by the secreted
juices, and that this takes place in the absence of bacteria (Goebel).
Dernby (191 7) pushed the matter further, and by means of glycerin
extracts obtained a true tryptase, not observed elsewhere among
plants. There is also a weak and incomplete pepsidase effect, as small
amounts of amino-compounds are set free at pH 8. The tryptase at-
tacked caseinogen at pH 8-9.

But this has not gone without further challenge.

MiRiMANOFF (1938) found that the gland ceUs of both stalked
and sessile glands, where an insect was attached, showed aggregation.
His description of this agrees with that of Darwin and others. He
could not induce it, however, with other substances (egg-white, cheese,
meat extract). It appeared to him that only certain products of
deamination were responsible for disturbing the osmotic equihbrium
of the cefl, inducing the changes leading to aggregation. It is revers-
ible, and different from those irreversible changes observed on the ap-
phcation of neutral red, though by some observers they have been
regarded as similar or ahke. Incidentally, pointing out the total con-

Francis E. Lloyd — 112 — Carnivorous Plants

tradiction between the results of Tischutkin, who denied the role of
other than bacterial digestion, and Colla, who argued the opposite,
he states his belief that digestion by the leaf is extraordinarily feeble,
and it seemed to Mirimanoff that Pinguicula would better be re-
garded as a "semi-carnivore". Following up this hint Olivet and
Mirimanoff (1940) re-examined the matter by a new method. They
applied {a) a sterilized insect (Drosophila) to a bacteria-sterile leaf,
and (b) one to a non-sterile leaf; and (c) a non-sterile insect to a non-
sterile leaf. In the first case there was no evidence of digestion, and
none of aggregation and no discoloration of the glands. In the second
there was an evident discoloration of the glands, and aggregation was
observed. Tested, the fly now was swarming with bacteria among
which were gelatin-liquifying motile forms. In the last case digestion
of the insect proceeded with abundant evidence of aggregation and
discoloration. It was tried to obtain the putative protease by diffusion
into gelatin-sugar in the cold. On warming at ordinary temperatures
there was no liquifaction. They concluded that the digestion of insects
on Pinguicula is the result of bacterial activity, and while the authors
do not deny the presence of a protease secreted by the plant, they hold
its action to be negligible.

Thus the question has been reopened, and demands further critical
examination.

Pinguicula has long been supposed to have the ability to curdle
milk. Linnaeus (Flora Lapponica, p. 10) tells us that the Lapps used
it for the curdling of milk and that the peasants of the Italian Alps use it
similarly (Pfeffer, through Oppenheimer) . Francis Darwin also
records the fact that the same use was made of it by the farmers of
Wales "for the past 30 years" as previous to 1875. This probably
means a very much longer time (F. Darwin, in a footnote in Darwin,
1888).

The fact that some plants can cause coagulation in milk (notably
Galium veruni) was known to the ancients, according to Czapek. It
is not clear what precisely the function of a rennet on this plant would
be, but it seems that it is not a substance per se, but that the pro-
teolytic enzymes have the property of coagulation, as will be seen
beyond. In relation to this question, the following quotations were
sent me by Dr. Oke Gustafsson, translated and transmitted by
Dr. Jens Clausen, to both of whom I owe thanks: —

"This 'tatort' {Pinguicula vulgaris) has long been used in some of the more northern
provinces of Sweden, as for example Jamtland and Dalarne. It has been mixed with fresh
milk by smearing either the milk-sieve or the container with the glutinous leaves. For
a long time it has been a common view that the milk was changed to ropy- or long-milk
by its tough and viscid slime similar to cheese-lep (The milk has been given this name be-
cause it is so thick and tough [viscid] that it can be pulled into long strands). Through
experiments it has now been found that long-milk cannot always be produced with Pin-
guicula (the 'tatort'), if ever, but that on the contrary such milk can originate without
this medium." (Lindman).

Properties and uses in Norway and Denmark. — "When the leaves
are laid in milk it will curdle, although without separating from the
whey, and this milk, in Norway called 'Ta^ttemaelk' (ropy milk) will
make other milk curdle. From this the Norwegian name 'Tsettegrses'
(curdlegrass) and the Faroe name 'Undslaeva Greas' have their

Chapter IX — 113 — Pinguicula

origin. Especially the milk of reindeer is supposed to curdle. However,
it had been impossible for me to obtain information that at the present
time this plant is used in Norway for production of ropy milk, be-
cause usually the left-overs of milk curdled in this manner are used to
thicken fresh milk with.

"Previously this plant has been accused of producing liver sickness
(rot) in sheep but we now know that this is an effect of the liver
fluke, Distotnum hepaticum, which lives in wet pastures (Note by
J. Clausen, in ep.). The bees seek this plant, but stock do not eat it.
It is told that it will stain yellow. It is an indicator of moist, so-called
sour soil. In places it is used mixed with linseed oil as a home-remedy
against wounds." (Hornemann).

Dernby considered this whole question fully, citing the popular
belief in Scandinavia that both Pinguicula and Drosera procure when
in contact with milk a "long", that is, a very viscous coagulum.
Although the work of Troili-Petersson and Olsen-Sopp (Centralb.
f. Bact. II T^T,: 191 2) shows that these plants have nothing to do with
"langmjolk", yet the expressed sap of Pinguicula leaves does have
a definite effect on sweet milk, that is, on its casein. It produces a
viscous fluid of alkahne reaction, but the casein is not coagulated, but
broken down into simpler bodies. Dernby states the foflowing con-
clusions from experimental evidence: — (7) Dialysed expressed sap of
Pinguicula cannot make milk "thick"; (2) On the other hand it splits
casein of milk, but only partly, in a weakly alkaline field, just as
it does Witte-peptone under the same conditions; (3) The enzyme is very
similar to trypsin, working at an opt. pH of ca. 8; {4) No enzyme of
pepsin-erepsin character could be found.

Therapeutic effects. — P. Geddes pointed out that all alpine peasan-
try apply the leaves to the sores of cattle, and its healing effect, if such
there is, might be referred to the antiseptic properties. More recently
there have been more exact studies made of this property (McLean,
191 9) indicating the truth of Geddes' report.

Summarizing, we may conclude that Pinguicula is a carnivorous
plant inasmuch as it catches small insects and digests them, at least
in part, by means of its own ferments. The possible part played by bac-
teria is not excluded. Its leaves are very sensitive to too great "por-
tions" of food as GoEBEL truly said. Only minute insects can be
captured in nature, this being a matter of common observation. Large
insects or bits of fibrin, unless very small, cause decay beneath with
permanent injury to the tissues. A closer understanding of the chem-
ical nature of the digestive ferments has been attained by Dernby.

As to the power of the leaf to move, first observed by Darwin,
there can be no doubt of the fact, and that the stimulus, supplied
by the application of various kinds of substances, organic and in-
organic, is transmitted in some fashion, but only slowly. The short-
est time in which Darwin observed movement was 2 hours 17 minutes,
the stimulus being transmitted over a very short distance, a matter
probably of not more than 2 to 6 mm. Movements can be induced
by substances which do not cause increased secretion, such as fine
grains of sand, as I have also observed. Increased secretion follows
the application of sugar and proteins among others. But that following

Francis E. Lloyd — 114 — Carnivorous Plants I

, — ■ i

sugar does not contain ferments, indicating the abeyance of activity I

on the part of the sessile glands in this case. i

Literature Cited:

Batalin, a., Mechanik der Bewegungen der insektenfressenden Pflanzen. Flora 60:33-39; I

54-58; 65-73; 105-111; 129-144; 145-154, 1877 (Pingiiicula, pp. 150-154). _ {

CoLLA, Silvia, Sui fermenti secret! da Pingidcula alpina L. Annuario della Chanousia

3:144, 1937 (through Mirimanoff).
CzAPEK, F., Biochemie der Pflanzen. 3 vols. 825 pp. Jena, 1925.
Darwin, C, Irritability of Pingidcula. Gard. Chron. II, 2:15 and 19, 4 July, 1874.
Darwin, C, Insectivorous Plants. 2d Ed., 1875.
Dernby, K. G., Die proteolitischen Enzyme der Pingidcula vulgaris. Bioch. Zeitschr. 80:152-

158, 1917.
Fenner (see under Nepenthes).

Geddes, p., Chapters in modern Botany. New York 1893, 201 pp. 1

Goebei., K., Pflanzenbiologische Schilderungen, TI. 1891. I

GoRUP, (see under Nepenthes) \

Green, J. R., (see under Nepenthes). _ I

Gressner, H., Botanische Untersuchungen, i. Beobachtungen iiber Pingidcula vulgaris. |

Jahresber. d. evangel. Fiirstl. Bentheim'schen Gymn. Arnold, z. Burgsteinfurt. Iser- 1

lohn 1877. '

Hornemann, J. W., Fors0g til en dansk oekonomisk Plantelaere. Kj0benhavn 1821, pp. <

27-28. I

Hovelacque, M., Sur les propagules de Pingidcula vulgaris. C. R. 106:310, Feb. 1888.
Klein, J., Pingidcula alpina, als insektenfressende Pflanze und in anatomischer Beziehung.

Beitr. z. Biol. d. Pflanzen. 3:163-185, 1880.
Lindman, C. a. M., Bilder ur Nordens flora, p. 100. Stockholm, 1922. I

LoEW, O. & R. Aso, Benzoesaure in Pingidcula vulgaris. Bull. Agri. Coll. Tokyo Imp. 1

Univ. 7:411-412, 1907.
McLean, R. C, The anaerobic treatment of wounds in life and its maintenance. New 1

York, 1919. _ _ _ ;

Mirimanoff, A., Aggregation protoplasmique et contraction vacuolaire chez Pingidcula

vulgaris L. Bull. Soc. Bot. de Geneve II, 29:1-15, 1938.
MoRREN, E., Observations sur les precedes insecticides des Pinguicula. Bull. Acad. roy. ,

d. Sci. etc. Belg., 2 ser., 39:870, 1875.
NicoLOSi-RoNCATi, F., Contributo alia conoscenza citofisiologica delle glandule vegetali.

Bull. Soc. Bot. Ital. 1912: 186-193.
Olivet, R. & A. Mirimanoff, Pinguicula vulgaris L. est-elle une plante carnivore? BuU. j

Soc. Bot. Geneve II, 30:230-235, 1940.
Oppenheimer, C, Die Fermente und ihre Wirkungen. vol. 2, pp. 1106-1111. Leipzig, 1925.
Rees, (see under Nepenthes).
TiscHUTKiN, N., Die RoUe der Bacterien bei der Veriinderung der Eiweisstoffe auf den

Blattern von Pinguicula. Ber. d. d. bot. Ges. 7:346-355, 18S9.
Will (see under Nepenthes).
VON Willer, Vital-Microscopische Beobachtungen an Insektenfressenden Pflanzen. Trudy

Inst. Fiz. Narkomprosa (Trav. Inst. Rech. Phys. Moscou) 2:517-519, 1936. Not seen.

Describes a method of observing a single gland of Pinguicula exclusively, all others

being left intact, to be stained vitally and otherwise experimented on.

Chapter X
DROSERA

Number of species. — Geographical distribution. — Habitat. — Form and habit of the
plant. — Unfolding movements of the leaf. — The leaf (Form. Anatomy. Appendages.
Tentacles. Sessile glands, origin and structure, function. Locus of absorption. Other
glands). — Reproduction. — Carnivory, early observations. — Mucilage, origin. — Move-
ments of the tentacles (Early observations. Nitschke and Darwin). — Direction of bend-
ing. — Duration of response. — Leaf blade not receptive to stimulus. — Path of stimulus.

— Intensity of stimulus. — Mechanism of movement. — Behre's studies. — Aggregation.

— Digestion. — Enzymes. — The significance of carnivory for the plant.

The genus Drosera contains more than 90 species found in almost
all parts of the world. It reaches its greatest development in Australia
and is well represented in S. Africa. The most widely known, at least
historically, is the common sundew, ros solis, D. rotundifolia, the plant
which chiefly formed the subject of Darwin's extensive studies. This
and its allies, D. anglica, D. intermedia and filiformis, also well known
in the North Temperate zone, are modest representatives of the genus
as compared with such forms as D. gigantea of Australia or D. regia of
S. Africa.

Habitat. — It is very generally understood that Drosera grows
where the soil is poor in nutrient substances. Such a statement ap-
plies fully enough to the best known species of the northern hem-
isphere, D. rotundifolia, intermedia, filiformis, etc., but seems not to be
true of some species such as D. Whittakeri of S. Australia, where I
saw it growing on wooded slopes with a general vegetation. Even this,
however, though probably a richer soil than that of a sphagnum
swamp, is not a rich soil. One commonly linds D. rotundifolia in any
swamp where Sphagnum grows, and it grows plentifully in the chinks
of partially decayed floating or stranded logs, a favorite place. In the
Sequoia National Park, California, it is found in the wet open mead-
ows surrounded by Sequoia gigantea, growing on a dense floor of moss
(not Sphagnum). A more accurate picture is afforded by Weber (1902)
in his monograph describing the great swamps of Augustumal, in the
delta of the Memel River. There is in this swamp, as of course in
swamps elsewhere, a zonation of the vegetation. As one proceeds from
the margin to the middle, one finds that the ash content of the soil
and soil water becomes more and more reduced. It is only in the more
central parts that D. anglica and D. rotundifolia are to be found, and
these are the parts which are most lacking in salts. The vegetation
here consists of Spagnum with Cladonia uncinalis, Scirpus caespitosus,
Eriophorum vaginatum, Scheuchzeria palustris, Rhynchospora alba,
Vaccinium oxy coccus, and Andromeda polifolia — and therefore of few
species. This habitat was found to have a soil with the following com-
position in absolute and relative terms. In quoting these data, Schmld

Augustumal Diluvial clay

SWAMPS

SOIL

Potassium

0.044 (1)

1.06 (24)

Phosphoric acid

0.075 (i)

0.18 ( 2.4)

Calcium

0.217 (i)

2.86 (13.1)

Magnesia

0.138 (i)

0.88 ( 5.9)

Francis E. Lloyd — 116 ^ Carnivorous Plants

points out the absence of data on the nitrogen content, and cites, in
order to fill the gap, the fact, stated by Wollny (1897), that the soil
(raw humus) of a pine forest as compared with that of the Drosera
habitat, contains nitrogen in the proportion of 27:1. Even more strik-
ing than the fact that the habitat is of such poor quality in respect to
salt content is the further observation that the first immigrants onto
the newly cut turf surfaces after the removal of peats, is Drosera, and
this remains for a long time the only inhabitant of these raw peat
surfaces. We may recall in this connection that Correns (1896)
showed that tap water at a high temperature (54.4° C.) does not
cause movements of the tentacles, but that water devoid of CaCOs
and CO2 called forth reactions at that temperature. In this way he
detected a toxic effect of Ca and inferred that this substance in the
soil (at least too much of it) might be toxic.

Form and habit of the plant. — The commonest type of Drosera
consists of a slender stem crowned by a rosette of leaves with flowering
scapes growing in the leaf axils. It arises from a seedling (D. rotundi-
folia) which has a fugacious taproot, which, however, serves for the
formation of the earliest rosette of leaves (Nitschke, i860). Ac-
cording to Heinricher (1902) the taproot fails to elongate, but swells
into a rounded mass covered with root hairs. The cotyledons are
simple, spatulate, followed by leaves of the mature type, though small
and with fewer appendages (tentacles) than the latter. As the plant
grows the stem dies off behind. In winter the rosette is reduced to a
tight compact winter bud which may have no extending stem or roots.

Growing as it {D. rot^indifolia e.g.) does in mats of Sphagnum, the
differential growth rates of these plants brings it about that Sphagnum
by its more rapid growth during the cool months, overtops the Drosera
and in the warmer months the latter in its turn overtops the Sphagnum.
One sees, therefore, in a Drosera plant, which has grown in this way,
successive dead rosettes clinging to the dead stem, ending above in a
living rosette, as figured by Nitschke. Such are our familiar species
of the northern hemisphere. The leaves of the rosette when fully ex-
panded may be relatively small, as in D. rotundifolia, intermedia, etc.,
or very large and ligulate, as in a remarkable species, D. regina, de-
scribed by Miss E. L. Stephens from S. Africa. In this species the
leaves are 2 cm. broad by 35 cm. long. Or again the leaves may be
large and fern-like in aspect, with strong terete petioles with a once to
thrice parted leaf blade as in D. binata, D. dichotoma (S. Africa, Aus-
tralia). These make showy greenhouse plants, and have often been
cultivated and used for study, to be reported upon in some detail
beyond.

Or again the main stem may be elongated upward, only slightly in
D. capensis, an often cultivated form from S. Africa, with hgulate leaf
blades supported on rather long petioles {13 — 5,7). In the most stately
species D. gigantea the stem may be a meter long and many plants together
form a dense half shrubby tangle crowned with the numerous flowers
in panicles. The stems climb or clamber, partly by twisting and partly
by means of certain long-petioled leaves in which the leaf blade be-
comes a disc of attachment, its dense secretion forming an adhesive
(GoEBEL 1923). The stems are wiry, the leaves peltate and deeply
cupped. It is a pronounced sclerophyll, according to Czaja.

Chapter X — 117 — Drosera

Some species, after the seedling stage is passed, form tubers which
perennate, and send up strong stems ending in a rosette or whatever
type of above-ground parts it has. From the stems grow axillary,
positively geotropic shoots (droppers), at the ends of which new tubers
arise {See beyond for details). These species have no roots, while in
general the roots are always meagre in numbers and extent, a fact
which is well known (Schmid). The root hairs are numerous in some
species and their walls are suberized and persistent. In other species
the root hairs are sparse. Marloth has reported both conditions in
S. African species. In some species the roots are apparently replaced
by rhizoids. Diels has thus described for D. erythrorhiza the root-like
productions one to three in number from the base of each scale leaf.
These have no root cap, but are provided with "root" hairs (Diels
1906). GoEBEL comments on the nature of these structures, called
by the equivocal name of leaf-roots ("Blattwurzeln"), pointing out
that in the apex, while no root cap is present, there is an apical mer-
istem just behind the epidermis, the outer walls of which are thick-
ened, and which are evidently a protective mail for a boring apex,
which may be regarded perhaps as a reducing or reduced root end.
Their origin according to Diels, however, is exogenous, and he called
them leaf rhizoids, but leaves details of their origin not fully under-
stood. I have verified Diels' observation.

I have examined the "leaf roots" of D. erythrorhiza from West
Australia and am able to confirm Goebel's observation of very thick
outer cell walls of the apical cells (75 — 19). There is a meristem, but
this does not lie immediately behind the epidermis, but just back of
three cell layers within. The apex itself is composed of enlarged
epidermal cells underlain by other cells of similar appearance derived
from two subepidermal layers, and heavily loaded with large starch
grains. The apical cells constitute a boring organ which does not
slough off as does the root cap. If there is any renewal of substance,
this would be in new secretion of cell wall. The plant grows, however,
in very loose soil where friction against the growing tip is minimum in
amount. At all events I have examined a large number of "leaf-
roots" and have not found any evidence of renewal of epidermal cells.

In the axis of each rhizoid-bearing scale small tubers, having evi-
dently the function of reproduction, can be produced (Goebel 1933).

Unfolding movements of the leaf. — In many of the species of Dro-
sera {D. rotundifolia, D. pygmaea), the petiole is bent so that the
upper face of the blade becomes applied to the petiole {16 — 17). This
is brought about by the hyponasty of a more or less narrow zone of the
petiole at the base of the blade. In other species, however,
those in which the leaf blade is slender and filiform, there oc-
curs true circination, as in D. filiformis, D. regia (with slender
ligulate leaves with short petioles), D. binata, D. dichotoma, with
the volute facing the stem, and due again to hyponasty. Just
the opposite occurs in Drosophyllum and in Byblis linifolia. These
two directly opposite behaviors appear, according to Fenner (1904),
to be related to the need for protection of the tentacles since they are
on the upper surface in Drosera and on the lower in Drosophyllum,
but, it is to be noted, along the margins in both, with the result that

Francis E. Lloyd — 118 — Carnivorous Plants

in the volute a large number of the tentacles are exposed and cannot
receive protection from the overlying turn of the volute. I have ob-
served this and can confirm Goebel on the point. On the other hand,
GoEBEL proposes a causal explanation as follows. The production of
a great extension of surface by the growth of tentacles can act to in-
hibit the growth rate of that surface, and thus permit the more rapid
growth of the other face of the leaf, the lower in Drosera, the upper
in Drosophyllum (Goebel, 1924). In Byhlis gigantea the leaf shows no
such movements. The leaf grows in a basal zone, and the filiform blade
extends always straight on. In this the glands are more numerous on
the lower surface. Here the distribution of the very numerous glands
either has no inhibiting effect, or has an equal effect on all sides of the
leaf.

Of particular interest to us here are the leaves, which are the
mechanism for catching and digesting prey. These present a variety of
forms from a simple orbiculate bifacial leaf of small size {D. rotundi-
folia I cm. diam.) through linear {D. filiformis) to broad liguliform
tapering at both ends {D. regia). Or the blade may be once to twice
forked {D. binata, D. dichotoma) the petioles firm and cylindrical
("rush-like" as Darwin put it). Further, the leaf may be peltate,
either obliquely {D. pygmaea) or centrally {D. gigantea, D. peltata,
D. subhirtella), sometimes with two basal lobes {D. auriculata) making
the leaf base angular, a condition reaching its maximum expression in
D. lunata (E. Asia). In the seedlings of the peltate leafed species the pri-
mary seedling leaves are usually non-peltate, those of D. peltata re-
sembling the following leaves of D. rotundifolia (Diels, Goebel).

The leaf is conspicuous because of its glands raised on elongated
stalks, each bearing a drop of mucilage which is extremely viscid and
serves to entrap small insects. Erasmus Darwin thought that "Dro-
sera mucilage prevents small insects from infesting the leaves" (The
Botanic Garden, vol. 2, Canto i, p. 229).

Anatomy of the leaf blade. — The epidermis is composed of straight-
walled cells in D. rotundifolia and D. capensis, but in D. Whittakeri
the lower epidermis is wavy-walled, the upper straight-walled. In
these species the cells have many chloroplasts, absent from the lower
epidermis of D. rotundifolia (Solereder).

The internal parenchyma has no palisade, as pointed out by
NiTSCHKE, the whole being made up of rounded cells in rather few
courses, more in some species {D. Whittakeri) than in others {D. ro-
tundifolia). In the latter species there are, in the case examined by me,
3 to 5 courses of cells. The smallest are in contact with the upper
epidermis. Below there are much larger cells, the third course in con-
tact with the lower epidermis unless a fourth course occurs, when the
cells are somewhat smaller, but still larger than the upper course
cells (/J — 15). All the cells, usually including the epidermis (Solere-
der), contain chloroplasts. Stomata occur on both surfaces. The in-
tercellular spaces are large. This general structure is, as Schmid
(191 2) has said, rather primitive, a quality which is shared, in varying
degree, with insectivorous plants in general, indicating that this qual-
ity stands in a probable relation to carnivory. In these plants the
elaboration of starch and its metabolism and withdrawal are all slow
processes.

Chapter X — 119 — Drosera

It was observed by Schmid that during the absorption of materials
from the bodies of prey, the starch content of the tissues at the base
of the tentacles is lost. According to Spoehr (1923) the amino acids
are concerned with the metabolism of starch. From this Geessler
(1928) was prompted to investigate the influence of salts on the me-
tabolism of starch in the leaf of Drosera capensis. He found that in
this species, when the leaf is fed with insects or with various salts,
there is a disappearance of starch from the leaf. The leaves of D.
capensis are in summertime heavily loaded with starch. The starch
content is not lowered even when the plant is kept in the dark. Even
after 45 days in the dark in contact with distilled water, the leaf (at
temp. 36-38° C.) showed no reduction in starch. The sugar con-
tent is minute. In winter the leaves are starch-free, but there is as
little sugar as in summer. These facts, together with the high respira-
tory intensity, indicated to Giessler that the physiology of Drosera
resembles rather that of the animal than of the plant, in that there
is a protein respiration. He suggests that the starch is used in the
secretion of mucilage and in supplying the energy for the bending of
the tentacles and leaf blade in response to stimulation. In support of
his thesis he points out the abundant occurrence of labile albumin
(LoEw) in many carnivorous plants and mentions in support of this
the work of Erna Janson on aggregation, to which reference is made
elsewhere. It has often been asked if the carnivorous plants are not
animal-like in view of their habits, and this is at present answered as
above.

The absence of a palisade tissue in Drosera, already mentioned, is
not confined to this genus, but is generally though not universally true
of carnivorous plants. This lack stands, according to Schmid, in re-
lation to insectivory, the latter affording compensation. But Kos-
TYTSCHEW questioned this, and did experiments which he regarded as
proving that both Drosera and Pinguicula are quite as active as the
control plants which he used. As his figures are the only ones avail-
able, I give them. The amount of CO2 assimilated per i dm^ of leaf
surface: Drosera rotimdifolia, 4 cc, Tussilago farfara (control) 3.8 cc,
Pinguicula vulgaris, 38.4 cc, Aegopodium podagraria 18. i cc.

"Thus KosTYTSCHEw's experiments answered the question whether a carnivorous plant
can obtain its carbon nutrition through photosynthesis in the affirmative. The scant ex-
perimental data show, and the text implies, that Drosera and Pinguicula leaves, which
have not had access to animal nutrition for some time, carry on photosynthesis at a normal
rate. The observed rates are in good agreement with those estabhshed by Willstatter
and Stoll for a wide variety of green plants. Kostytschew's comparisons with Aiiricu-
laria and Lenina also bear this out.

"From his data on photos>Tithesis of Drosera it appears, however, that the rate of
carbon dioxide assimilation would have increased materially after feeding the plants with
insects. The experimental details have not been recorded in sufficient detail to permit of
a definitive decision. But the discovery of a measurable effect of the ingestion of animal
material on the rate of photosynthesis would open up a new approach to a study of the
problem of photosynthesis itself. The importance of such a possibiUty made it an easy
matter to obtain the co-operation of Dr. W. Arnold in carrying out some preliminary ex-
periments.

"Drosera and Pinguicula plants, previously not animal fed, were used for the experi-
ments. Single leaves were placed in distilled water in the center-cups of Warburg vessels.
A mixture of sodium carbonate and bicarbonate was introduced into the main chamber
in order to insure a constant carbon dioxide pressure in the gas phase. Photosynthesis

Francis E. Lloyd —120— Carnivorous Plants

was measured manometrically at 27° C. The rate was constant over a period of six hours,
at the end of which one leaf was fed with a fly and some egg albumen, while another was
kept as a control. Repeated measurements over a period of some 20 hours following the
feeding showed that the control leaf maintained a practically constant rate of both pho-
tosynthesis and respiration. The rate of oxygen production of the experimental leaf
appeared somewhat depressed, but its respiratory rate was considerably higher than that
of the control. By correcting the photos3Ti thesis measurements for respiration in the
usual way it was found that the corrected values do not differ significantly from the orig-
inal ones. The increased respiration obviously resulted from the availabiUty of substrates
for oxidation on the outside of the leaf, and may be caused by the plant itself or by con-
taminating micro-organisms. These experiments lend no support whatever to the idea of
an influence of feeding upon the rate of photosjm thesis of carnivorous plants." (C. B. van
NiEL, in ep.).

The appendages of the leaf. — There are several kinds of appendages
but they are not all common to all species of Drosera. Some are im-
portant physiologically in relation to the carnivorous habit, others not.
To the former belong the tentacles and sessile glands, common to all
species; to the latter are the glandular and eglandular trichomes seen
in D. rotundifolia and other N. hemisphere species and the glandular
trichomes found in such species as D. gigantea, and distributed over
the whole plant body (75 — 16). We may add, at this point, that
the fringes of trichome-like structures were regarded collectively as a
ligule by Nitschke. It is a fringed membrane formed at the sides and
across the leaf base in D. rotundifolia and some other species {16 — 18),
but is absent from many others {D. Whittakeri, D. peltata, D. gigantea,
etc.) It has been regarded as stipular and is so called in the taxonomic
literature (Diels) though Small (1939) takes another view, that the
apparent membrane is merely a linear cluster of trichomes. That
similar trichomes are found abundantly on the rest of the petiole sup-
ports his contention. On the other hand it is difficult not to see in
the huge ligulate "stipule" possessed by some species {D. paleacea,
D. pygmaea) {16 — 18) in Australia, in which they serve to protect
the bud during periods of drought (Diels), an integration of a fringe
as it occurs e.g. in D. rotundifolia.

Tentacles. — Of these, the stalked glands or tentacles are the most
conspicuous and have most frequently been described. They have often
called forth exclamatory remarks of wonder at their complex structure.
They have been described, but not always correctly, by Gronland,
Trecul, Nitschke, Warming, Darwin, Huie, Fenner, Homes and
probably others. The tentacles occur on the margin and upper surface
of the leaf blade and in some species on the tapering upper region of
the petiole, excepting those species which are strictly peltate.

The "tentacle" consists of a tapering stalk topped by an oval gland.
The stalk arises from the leaf surface, as a mass of tissue including all
the elements of the leaf structure, epidermis, parenchyma and vascular
tissue. The terms "trichome" and "hair" are therefore not suitable,
though they have been used.

The term "tentacle" is not a strict one; it has been equated with
"emergence" and serves if we think of the tentacle as an extension
of the leaf adapted to certain functions which makes'them so trichome-
like that they are no longer distinguishable from trichomes (Diels).
Nitschke and others regarded the tentacles as extensions of the leaf,
Warming as trichomes and Penzig as intergradients between
phyllome and trichome.

Chapter X — 121 — Drosera

In the upper reach the tentacle consists of the epidermis and one
course of parenchyma cells surrounding a very slender vascular strand
which extends from the leaf system up into the gland (75 — 6). This
was seen by Meyen in 1837, who supposed that it entered the gland.
This structure led Trecul (1855) to compare the tentacle with the
dicotyledonous stem, and to regard the adventitious buds described
first by Naudin as metamorphosed tentacles. Gronland called
them lobes, and Schacht, projections of the leaf. On the surface
as part of the epidermal system there are a few small sessile glands,
these being found also on the general leaf surface. They formed con-
venient marks by which H. D. Hooker was able to record changes
in the length of the tentacles during movement. The widened base
of the tentacle has, naturally, an increasing number of parenchyma
cells as the general leaf surface is approached. Similarly the vascular
system, consisting of spiral tracheids, may here consist of two or more
vessels, but above there is usually found only a single strand except
where two may overlap. Fenner did not see any phloem, and I can
only support him in this (ij — 7). The single vessel sets into a dense
mass of thick and short tracheids occupying the middle of the gland
(75 — 6) which, oval in form save when on a strictly marginal tentacle,
sits atop the narrow neck of the stalk. Those tentacles arising from
the leaf margin are bilaterally symmetrical, the stalk being extended
under the glandular structure proper in the form of a spoon holding
the gland in its bowl (ij — 9, n; 16 — 1-3). Darw^in records finding
intermediate forms, which I have also seen. The tentacles spring-
ing from the surface are increasingly radially symmetrical as the
margin of the leaf is left, are oval, and present the following struc-
ture.

The oval head of the tentacle consists of four layers of cells (75 —
6, 8). The innermost of these is a roughly oval mass of tracheids which
is connected by means of the vascular strand of spiral vessels in the
stalk with the system in the leaf. Surrounding this xylem mass,
three outermore layers cover it as a thimble, the flaring mouth of it
articulating with the somewhat expanded tip of the stalk. The layer
of cells in contact with the xylem mass is distinctly bell-shaped, and
was called by Fenner the parenchyma bell. The flaring wall of the
bell is composed of a single layer of elongated, curved cells, the ex-
posed ends of which come to the surface of the gland, and whose cuti-
cle is continuous with that of the gland above and the stalk below.
The inner ends articulate, at a point about half-way up the bell, with
shorter cells, forming the top of the bell. Both the transverse and
longitudinal walls of all cells are cuticularized so that when a gland has
been treated with sulfuric acid, these walls remain as a network (75 —
12) or, as it were, a cage formed of a continuous band of cuticularized
cell wall. In transverse section this band is T-shaped, the cross bar of
the T being narrow and placed towards the outside with respect to the
gland as a whole. Huie believed that only the outer part of the wall
(approximately one-half) is cuticularized, and abuts at the middle of
the wall on a pit connecting adjacent bell cells, the inner moiety of the
wall being Hgnified. Fenner did not see this, and I have been unable
to verify Huie's description. This parenchyma bell appears to func-

Francis E. Lloyd — 122 — Carnivorous Plants

tion as an endodermis, though Fenner questions Goebel's view that
water may pass only in one direction (outwardly). The outer ends of
these cells form a continuous single row of rounded outHnes like a string
of beads, seen in an entire gland, which limits the gland proper from the
uppermost transverse course of stalk cells. These cells were seen by
Warming, whose drawing Darwin reproduces. But Darwin (1875
2d. ed., p. 5) himself failed to see them, nor, said he, did Nitschke,
though one of his drawings seems to indicate that he did. Neither
did Gronland (1855) or Trecul (1855) see them.

Fitting over the parenchyma bell are the two layers of glandular
cells. The outer course is made up of columnar epidermal cells,
polygonal en face, their outer ends covered by a cuticle and their
radial, and sometimes outer walls strengthened by cellulose buttresses
and beams (75 — 13), as shown by Fenner. They are most pro-
nounced throughout the lateral reaches of the gland and diminish in
stature toward its apex, from which they are quite absent (Huie),
though Homes thinks they occur here, but are smaller, in much smaller
numbers and far apart (1928). Careful examination persuades me to
agree with Huie. They are obvious in the apical cells of the glands
of D. pygmaea. Naturally enough, the protoplasm of the cell fits into
the bays between the buttresses, and by the use of weak H2SO4 for
maceration, the protoplasts may be isolated and are then seen edged
with crenellations, interpreted by Haberlandt as sensitive papillae.
If this is a correct view, we must think that the glands are more sensi-
tive along their sides than on the apex, for which we have no evidence
one way or another.

The cuticle covers over the whole of the gland and is continuous
with that of the stalk. As Huie has said, it is quite continuous and is
not penetrated by pores (Gardiner) nor is it absent from the apical
cells (Goebel). Goebel's statement to this effect appears to have
been due to the observation of the earlier penetration of solutes through
these cells, but I have satisfied myself that methylene blue enters
equally rapidly over the entire surface of the gland. Prolonged treat-
ment with sulphuric acid leaves a very delicate continuous membrane
covering it. Yet as Huie says, the cuticle is readily penetrated by
silver nitrate, just as by methylene blue. Another observation of
Huie's I can confirm, namely, that in life the lateral walls of the
apical cells are often separated from each other by fissures tapering
inwardly between them, as if the walls had separated along the middle
fine. It is possible that this is what Franca saw in Drosophyllum,
interpreted by him as canals leading to the inner course of glandular
cells. The nucleus of these cells lies near the base and the cytoplasm
has a large vacuole in the outer moiety of the cell (in the resting con-
dition — see beyond under aggregation).

The second layer of glandular cells lies between the epidermis and
the parenchyma bell, and is composed of more depressed and irregular
cells, overlooked by Nitschke, but seen by Darwin, and correctly
described by Warming. The cells are irregular in shape fitting the
irregular bases of the epidermal cells without intercellular spaces.
The functions of these two glandular layers differ according to Homes
as we shall see.

Chapter X —123— Drosera

The emplacement of the glandular tissues is different in the marginal
tentacles. Here the end of the tentacle stalk is formed into a spoon,
in the bowl of which lies the gland. There is, as it were, a torsion of
the upper part of the tentacle so as to bring the gland on the upper
ventral surface. The complete homology of the two types is seen on
examination of a transverse section of the marginal gland {13 — 9, n;

zd — I, 2).

To be included as a specialized portion of the gland, or better a
portion of the tentacle acting in a specialized way in cooperation with
the gland, is, according to Fenner (1904), the uppermost course of
epidermal cells of the stalk, those, namely, which are in direct contact
with the tissues of the gland at its base. These cells are short and
being epidermal, they form a circle of 8-10 cells called by Fenner the
"Halskranz", or as we may call them, the neck cells. Sometimes
there are two rows of neck cells (Konopka), and this I note may be
the case in D. Whittakeri. The neck cells surround the parenchyma
cells of the same transverse course but these latter are not included in
the "Halskranz", as defined by Fenner, who describes the anatomi-
cal relations as follows. The neck cells are in contact above with the
lower ends of the emergent parenchyma bell cells, and with the outer
zone of the xylem mass of the gland. Inwardly they lie in contact
with the short parenchyma cells of the stalk, these in turn lying against
the inner zone of xylem tracheids and with the end of the stalk vascu-
lar bundle. Below, the neck cells impinge on the stalk epidermis.
They are, as one may say, in a strategic position to carry on a special
function, if Fenner is right in his interpretation. That they have a
function he beheves is evidenced by the presence of numerous pits in
their walls which He against the parenchyma bell cells and those of the
xylem, and furthermore, by the fact that their cuticle is porous. He
gives the following interpretation. The neck cells receive water from
the bell cells which bring water from the upper part of the xylem mass,
and from the lower xylem cells, presumably also from the parenchyma
transversely within the neck cells, and pass it outwardly through the
pores of the cuticle supplying fluid to dilute the viscid mucilage se-
creted by the glandular cells above. The glistening drop of mucilage
supported on the tentacle head is, says Fenner, pear-shaped, the
broad part of the drop being around the neck cells because the fluid
exudes chiefly from them. The reasoning here appears disingenuous.
Nor is his statement that the cuticle is porous acceptable since dyes
(methylene blue) never enter the outer surface of the neck cells, but
pass into the stalk only by diffusion through the gland, as I have
verified repeatedly. While crediting Fenner with imagination, it is
still permitted to doubt the correctness of his interpretation and even
the supposed facts on which it is based. Konopka indeed has taken
issue with Fenner, and his view is stated beyond.

The development of the tentacle has been worked out by Homes,
and it becomes evident that the outermost layer of the gland is purely
epidermic in origin, as would appear on the face of it. The second
layer, which might be interpreted as of epidermic origin, is shown to
be of parenchymatous origin. The third layer, the parenchyma bell,
is partly epidermic and partly parenchymatous (Fenner). Those

Francis E. Lloyd —124— Carnivorous Plants

cells which come to the surface at the base of the gland are epidermic.
They are narrower and longer than the others, which are of parenchy-
matous origin. The inclosed mass of reticulated, and annular and
spiral vessels are obviously an extension of the leaf vascular tissue
(j^_4_6). The developmental behavior of the gland in Drosera cor-
responds point for point with that of Drosophyllum (Fenner).

Functions of gland parts. — Such a complicated gland as above de-
scribed can scarcely be a simple matter physiologically. The reception
and transmission of stimuli, the secretion of mucilage, of one or more
ferments, probably of an odoriferous principle, water, and in the op-
posed direction, the absorption of the products of digestion are car-
ried on. Is it possible to assign any degree of specialization to the
various elements of structure? Homes (19296), having studied with
meticulous care the behavior of the cells in the matter of aggregation,
assigned to the outer layer, the epidermis, the function of "responding
directly to the necessities of secretion by the variation of its vacuome".
Its cells elaborate the substance secreted. That of the second layer is
the regulation of osmotic pressure. The third layer, the parenchyma
bell, takes no part in secretion (Homes, 1929^), p. 49). It may be as-
sumed, of course, that the cells of the bell allow the rapid transfer of
water from the inclosed xylem, but whether the movement is a one-
way one only, as Goebel suggested, or not, is difificult to say.

Reference has been made above to Haberlandt's view that the
protoplasmic processes lying between the buttresses of the epidermal
cells are sensitive organs, analogous to those seen by him_ in tendrils
and other plant parts. Konopka preferred another suggestion in 1930,
that the increased surface due to crenellation may be important also
in secretion and absorption, as a secondary advantage. Goebel has
regarded them in this way.

With respect to other parts of the gland Konopka has made some
further suggestions. The xylem bundle mass is, he says, composed of
spiral vessels of narrower bore in the central part, with wider lumened
tracheids surrounding them, and the more central vessels widen
in contact with the apical portion of the gland. The central vessels
are indeed often narrower than the outer, but other details it is diffi-
cult to accede. Konopka would attribute different functions to the
two regions, but beyond this regards the whole as a water storage
organ, which rather obviously it seems to be. He has, however, ex-
amined the behavior of the nuclei, and finds that during digestion and
absorption there occur changes in them which he interprets as con-
nected with taking up and transmitting nutrients from the outer tis-
sues of the gland to the stalk cells. He asserts that the nuclei of the
endodermis, of the xylem and of the stalk cells, show a gradient of such
changes, the nuclei of the more superficial tissues showing greater
changes in a quantitative sense than those of the deeper and more
removed tissues. To the endodermis (parenchyma bell) he attributes
the special function of a protective filter. It must be questioned
whether Konopka has advanced sufficient evidence to support this
hypothesis. Aside from the nuclear changes claimed to occur by
Konopka, there is no other change such as characterizes the secretion
cells, namely aggregation (Homes), during periods of activity. This

Chapter X — 125 — Drosera

seems to indicate that whatever the function of the endodermis, it is a
different one from that of the secretion layer, and this I believe is as
far as we can go in interpretation beyond admitting that substances
are transmitted, but not differentially.

KoNOPKA also questions Fenner's view about the neck cells. He
does violence to Fenner's definition of the neck cells by including the
uppermost parenchyma cells which lie somewhat (but very little) above
the level of the ring or circle ("Kranz") of neck cells. The neck cells,
as he uses the term, have membranes which resist the action of
concentrated sulfuric acid, and are similar in this respect to the endo-
dermis cells. Discarding Fenner's idea that they are especially con-
cerned with the transmission of water to the surface, he thinks that,
on the basis of his observation of the nuclear changes, which are sim-
ilar to those seen in the endodermal, tracheid and stalk cells, they
transmit absorbed materials downwardly to the stalk. This seems to
be a simple and natural view of the matter. But I have been unable
to see cuticularized walls in these cells, and Fenner says nothing of
this (75 — 12). Nor have others (Huie, Homes, myself) seen nuclei in
the xylem of the mature gland.

We may summarize what has been said in the few preceding para-
graphs by emphasizing the very complex functioning of the tentacle
gland, that it is, as a mechanism, relatively complex as compared with
many other known plant glands, but that we are far from recognizing
specific correlations between structure and function. It would seem
that the complexity of function is much greater than recognizable
structural differentiation.

Sessile glands. — ■ In addition to the stalked glands or tentacles
there are very numerous, small sessile glands, or, as Darwin called
them, "papillae". They were described for the European species by
Nitschke and others, and in detail by Fenner, who traced their
development. They are to be found on both leaf surfaces, on the
stalks of the tentacles, and elsewhere (petioles, scapes). The glands
project dome-shaped from the leaf surface, are little larger than the
stomata in area, and consist of a capital of two cells, which may be
rounded and compact, or more or less elongated into obliquely placed
cylinders. These stand on a short stalk of compressed cells in two
courses, each course of two cells. The basal cells have cuticularized
inner walls. These in turn stand on two epidermal cells (75 — -18).
Fenner describes also a variant of the fundamental form. It occcurs
on the petioles, and consists of a more or less elongated stalk with a
capital of about four cells.

The origin of the sessile glands is purely epidermal (Fenner).
The mother cells are two short epidermal ones which by tangential
division give rise to a pair of capital cells, the base of which is again
cut off to make stalk cells. The remaining true capital cells are two
in number and may remain rounded, or elongate more or less into two
divergent short cylindrical cells, seen on the base of the tentacles and
on the petiole (Nitschke). In D. Whittakeri these glands are much
larger and more complicated in structure and consist of twelve cells,
eight outer surrounding a core of four inner, the whole being supported
on a very short biseriate stalk of longitudinally compressed cells.

Francis E. Lloyd — 126 — Carnivorous Plants

Other glandular trichomes occur in D. gigantea (seen by Darwin) and
probably in other species. These are stalked, bear an oval gland, and
look superficially like the tentacles, but do not have their elaborate
differentiation. They are to be found scattered on the petioles and
stems; on the latter they are quite numerous. I failed to observe any
secretion. Though the gland is covered with cuticle, they absorb dye
readily. Their structure is indicated in i§ — i6. In origin they are
epidermal, but in the base there is a small involvement of parenchyma
as it is rather broad, the stalk tapering upward into a uniseriate por-
tion just beneath the gland. What function these can serve, if any, is
not known. Small flies have been observed sticking to them.

Function of sessile glands. — Darwin observed that aggregation
takes place in the sessile glands during the digestion of prey, and
thought therefore that they are concerned in the absorption of sub-
stances derived therefrom, "but this cannot be the case with the pa-
pillae on the backs of the leaves or on the petiole." It is not clear if
he meant this merely because of unfavorable position. But Fenner
held that the sessile glands of the concave leaf surface are alone capable
of absorption, pointing out that those of the dorsal surface are small,
and usually lose the capital cells. The active glands display cytoplas-
mic changes (Darwin's aggregation evidently) during the absorption
of nutriment. Because nuclear changes also intervene, Rosenberg
aligned himself with these authors. To all this Konopka opposes a
contrary opinion. Nuclear changes such as Rosenberg observed are
also to be seen in other glands, certainly not concerned in the absorp-
tion of substances; and the "middle layer" (endodermis) also is to be
found in nectaries, hydathodes, etc. He believes the sessile glands to
be hydathodes. They never, he continues, show such far-reaching
changes in nuclear behavior as do the tentacle cells, and there never
occur the " Digestionsballen " which he found in tentacle gland cells.
Nor have the glands any connection with the vascular tissues; they
develop much earlier than the tentacles, and occur on both leaf faces.
These points argue that the sessile glands are not absorptive. There
is, Konopka believes, much greater probabihty that they serve the
purpose of water secretion. In support of this view he cites as facts
(a) the "not small" vascular system of the roots; (b) the rich supply
of root hairs; (c) the wetness of the substrate; (d) the active passage
of water through the plant and (e) the high relative humidity of the
habitat, tending to reduce transpiration. And Schmid, he says, had
found that there is only a slow transfer of water to the tentacle glands
following the experimental removal of the mucilage drop, while on the
capture of prey there is an extraordinary increase of fluid supplied
from the leaf during the digestion of prey (as Darwin and others have
observed), all speaking for a process of guttation. Admitting the
above as facts (though Schmid 's results seem to question some of
them) Konopka arrives at an interesting interpretation of the whole
situation: the sessile glands draw off water from the leaf, supplying it
for the process of digestion and thus at the same time exert suction on
the tentacles, thus increasing absorption by them. These glands, he
says, may be roughly compared with the animal kidney which with-
draws water from the body thus making room for more to be absorbed.

Chapter X — 127 — Drosera

In support of the idea he recalls the case of the trap of Utricularia,
which is known to excrete water from the glands on its outer surface
(glands not much dissimilar from the ones in question) and to absorb
nutrition and water from the interior by means of the bifid and quad-
rifid trichome glands. Since these two sets of glands in Utricularia are
the only non-cuticularized areas of the inner and outer surfaces of the
trap and since the cuticle elsewhere is impermeable {e.g. to dyes) we
are forced to recognize its pecuHar glandular action as involving the
two sets of glands, as Czaja, Merl and Nold have beHeved. This
view would harmonize our ideas about the two apparently widely
different structures, leaf and trap.

The free flow of watery secretion observed during the earlier stages
of digestion or just previous thereto, even if Konopk.a.'s view is correct,
does not preclude the possibility that the sessile glands may not con-
tribute to the efficiency of the leaf by exercising the function of ab-
sorption as well. We may, therefore, direct our attention briefly to
the specific question of locus or loci of absorption of the leaf.

The locus of absorption. — Previous to the studies of Oudman, there
had always been a vagueness about the point of entrance of substances
absorbed by the leaf. Three possibilities there are: (i) that they enter
through the tentacles; or (2) through the papiUae; and (j) through the
epidermis, which according to Nitschke, has no cuticle. The last may
be at once excluded as Nitschke's statement is not true. Aside from
direct proof with sulfuric acid, the dift'usion of e.g. caffeine (Kok) into
the leaf takes place through the papillae, and not through nearby
epidermal cells.

With regard to the tentacles the fact of aggregation in the stalk
cells following on the application of various substances (insects, caffeine,
etc.) would seem to indicate at once that absorption can and does take
place through the glands. Darwin indeed regarded aggregation as
proof of absorption. Pfeffer, however, pointed out that this might
be the result of the stimulating eft'ect of minimal quantities of ma-
terial with no quantitative relations indicating absorption. Some such
substance has been thought to be necessary to procure aggregation,
that is, a specific aggregation-stimulating substance formed in the
gland (Akerman, 191 7; Coelingh, 1929). Ali Kok determined the
rate of transport of caffeine from the glands into the tentacle stalks.
Changes in the structure of the cytoplasm and nucleus (studies by
HuiE, Rosenberg, Konopka and Ziegenspeck, and Kruck on Utricu-
laria), were referred by them to the activity of these structures (cyto-
plasm and nucleus) in response to the absorption of various foods.
Taking up of food by the tentacles has been generally assumed, as
e.g. by GoEBEL, Fenner, Ruschmann. Oudman points out, however,
that there is little positive information and that even if the tentacles
do absorb, their role may be small and of secondary significance.

That the papillae, small sessile glands of various sizes, smallest on
the tentacle stalks, largest on the leaf blade, where they occur on both
surfaces, are concerned in absorption has been expressed by Darv;in,
and by Rosenberg, both of whom saw the ready passage of sub-
stances through them into the tissues. Rosenberg used methylene
blue (as I have repeatedly done). Fenner and Coelingh, as also

Francis E. Lloyd —128— Carnivorous Plants

Darwin and Rosenberg, saw that aggregation and granulation occur
in response to the entrance of various substances, but this is true of
the tentacles, also, and proves as much and as little in both cases.
To be sure it was thought that the papillae produce no secretion ex-
ternally escaping, and this has perhaps influenced the judgment. As
OuDMAN remarks, here also as in the case of tentacles quantitative
results had not been forthcoming. He therefore endeavored to supply

these.

Having first assured himself that the N- content of the leaves (of
Drosera capensis) under the circumstances under which he worked, is
nearly constant, Oudman then arranged a simple experiment (i) so
that the more marginal tentacles were surrounded by agar (2%), with
asparagin (1.5%), and (2) so that the mixture was poured on the back
of the leaf taking precautions against capillary flow. He obtained
these results :

Treatment of the Leaf

N IN % OF INCREASE IN
FRESH WEIGHT 24 HOURS

Control 2.07 —
Asparagin on the marginal tentacles 3.54 1.47
Asparagin on the back of the leaf 3.31 1.24

A second experiment, greater precautions against capillary flow: —

Control 2.01 —
Asparagin on the bordering tentacles 3.53 1.52
Asparagin on back of the leaf 3.38 1.37

From these figures it was evident that asparagin is taken up both
by the tentacles and by the back of the leaf. By comparing the total
area of the tentacle glands with that of the back of the leaf he found
that the amount of asparagin absorbed by the tentacles was six times
that absorbed by the back of the leaf. Two explanations presented
themselves, namely, either that the tentacle heads (glands) are better
adapted to this purpose than the leaf epidermis (which would be ruled
out by the fact that the epidermis is cuticularized, as above said);
or that the absorption by the leaf-back takes place only at certain
points, that is, through the papillae, through which it has been ob-
served that entrance can take place (Darwin, Rosenberg, Kok).
Oudman adopted the latter view, and inferred that in nature both the
tentacles and the papillae are made use of for the absorption of food,
but rather the papillae of the upper side of the leaf than those of the
lower. Oudman also examined into the question of the influence of
various factors (temperature, concentration of the applied materials,
the course of absorption in relation to time, the nature of the ap-
pHed material, the influence of the glands and narcosis).

As would be expected, the higher the temperature within physio-
logical limits, the more rapid the absorption. But whether this is
due to the greater rapidity of transportation, or to the greater uptake
by the glands, does not appear. The same with increasing concentra-
tions of applied substance (asparagin). In the course of absorption,
the rate was greater after the first period (3-6 hrs.), than at first, and
falls off again after 9 hours. This, it may be suggested, may be due to

Chapter X —129— Drosera

the dilution of the applied material by the secretion of the glands
during the beginning period, and to equiHbrium during the later period.

All substances are not absorbed at equal rates. Darwin noted
that they did not procure aggregation at the same rate. Oudman
found that caffeine is much more rapidly absorbed than asparagin,
although the latter has the smaller molecule. This may be due to the
path taken. Caffeine enters the vacuole and is there precipitated, and
fresh caffeine must traverse the zone of precipitation. Asparagin
probably passes along the path provided by the protoplasm. By
following the localization of fluorescence it was shown that fluorescein
does this. If the tentacles are removed, leaving the stalks open at the
outer end (due to the operation), less material is absorbed, but the
difference is not related to the exposed surface, it being much greater
for tentacles with the glands removed. The glands therefore offer
some hindrance, perhaps because they are quite complex organs, excreting
at the same time as absorbing. The presence of an endodermis (the
parenchyma bell, Fenner) may have some regulatory effect, but this
is not known to be the case. It is worthy of note that narcosis (with
ether) inhibits the penetration of asparagin more than caffeine, the
former traversing the protoplasm, the latter the vacuole. Caffeine is
known to penetrate into the vacuole with great rapidity (Bokorny,
Akerman, Erna Janson) and in any event it has to pass only a thin
layer of cytoplasm while asparagin is forced to pass lengthwise the
cells within the cytoplasm.

In a later paper by Arisz and Oudman (1937), making use of an
improved method of applying the reagents to the tentacles, Oudman's
figures describing the rate of absorption of caffeine and of asparagin
were confirmed. Caffeine is absorbed in the fashion of a physical
diffusion, while asparagin shows a maximum penetration in the second
period, and low rates in the first and third periods. Nevertheless more
asparagin penetrated into the leaf blade as shown by tests after the
removal of the tentacles before analysis. It seems obvious that the
conclusion that the paths followed by these substances are different is
justified, namely that caffeine travels by way of the vacuoles and
asparagin through the cytoplasm, yet in spite of the narrowness of the
path through the cytoplasm, the latter moves more readily. This again
seems to be due to the taking up of the caffeine by precipitation, a
subsequent wave of diffusion having to overstep the zone of pre-
cipitation.

An attempt was made by Arisz and Oudman to determine the in-
fluence of aggregation upon the transport of asparagin. Aggregation
was first induced by suitable reagents (sahcin 0.25% and KH2PO4
0.1% solutions) with a "remarkable result" that now more asparagin
was taken up during the first period (contrary to the above mentioned
rates). Since asparagin itself causes aggregation, during the first
period aggregation takes place, and during the second period, ag-
gregation now having taken place, penetration goes on more rapidly
because of this earlier induced aggregation. This behavior, that is,
aggregation, has on the other hand no effect on the rate of transport
of caffeine.

Reproduction by seeds and by buds {"regeneration''). — While Dros-

Francis E. Lloyd — 130 — Carnivorous Plants

era reproduces itself through seeds, it is, on the other hand, extraor-
dinarily prohfic by means of non-sexual multipHcation making use of
brood bodies {D. pygmaea, Goebel) and tubers, of strong axillary buds
and especially and above all of budding from the leaves. So frequently
and vigorously is the last method used that it would seem to rival
that by seed (Behre).

The seedlings are very small, the cotyledons either escaping from
the seed coat (Nitschke, Lubbock, Goebel) or remaining perman-
ently embedded therein {D. peltata and D. auriculata, Vickery 1933).
The earlier leaves in all cases are rounded (spatulate), indicating this
to be the primitive form for this genus (Leavitt, 1903. 1909). The
leaf blades are provided with a few glands, both marginal (Nitschke)
and on the disc, 5 on each (D. rotundifolia, Leavitt). The radicle
is short, provided with root hairs and fugacious. As the shoot develops,
adventitious roots put out from the stem, and, as this dies away with
the extension of growth, new adventitious roots are produced above.
The root system cannot be said to be abundant (Schmid).

In some species, e.g. D. rotundifolia (Nitschke), the axillary buds
below the rosette form at once secondary rosettes, similar to the chief
rosette, and as the stem decays they are separated, to propagate the
plant. In one group (Ergaleium) tubers are formed. These have been
described in their static condition by Diels (1906) and Morrison
(1905), and very fully, from the point of view of development,
by Vickery, from whose paper (1933) the present account is taken.
She worked with the two species D. peltata and D. auriculata which I
saw growing about Sydney, N. S. W. When exhumed, the stem below
the epigaeal rosette extends downward a matter of a few centimeters,
is clothed with scale leaves, and emerges from a small hard rounded
tuber clothed with loose membranous envelopes, which when peeled
off leave a smooth white tuber. This at the upwardly directed apex
bears an "eye", a depressed scaly bud which can develop into a new
plant {16 ■ — II, 12). The genesis of this structure is seen in the seed-
ling as an axillary shoot bearing normally only scales and growing
downwards {16 — 13). This is a "dropper". Reaching a certain
depth the end bends upward, and develops into a corm. While this
structure normally elongates upward to form a rosette at the surface
of the ground, if more or less exposed to light it may produce at once a
partial or complete rosette of normal leaves. Such leaves may arise
even from the extending dropper instead of scale leaves. An old tuber,
as it becomes exhausted, is usually replaced by another produced
laterally on the end of the dropper axis close to it. In Australia es-
pecially this form of reproduction is of common occurrence.

The underground tubers, as Goebel has pointed out, are doubtless
important as storage reserves of food and water which can tide the
plant over during a season when the rosette of leaves disappears. Some
of the Australian species have strong coloring matter in their tissues,
as is evident from the staining of herbarium sheets on drying. It con-
tains two substances, a red one CuHsOs, and a yellow C11H8O4, the
latter in only small amounts. Rennie (1893) had shown "that the
Os-compound formed a triacetate and was probably a trihydroxy-
methylnaphthaquinone, whereas the 04-compound gave a diacetate

Chapter X —131— Drosera

and appeared to be a dihydroxymethylnaphthaquinone." This was
confirmed by Macbeth, Price and Winzor, who called these sub-
stances hydroxydroserone and droserone respectively, determining the
constitution of the hydroxydroserone.

Reproduction by means of gemmae. — The case of D. pygmaea de-
scribed by GoEBEL (1908) is one of a small group of species in which a
very highly specialized method of non-sexual reproduction takes place,
viz., by means of gemmae. D. pygmaea is a very small plant, about
1.5 cm. in diameter, and consists of a tight rosette of minute acentri-
cally peltate leaves with fleshy petioles which appear to be the im-
portant chlorophyllous parts. On the approach of the resting season
there are formed small brood bodies, resembling superficially those of
Marchantia, clustered in the center of the rosette. The gemma itself
is a small, ovate, hard mass of tissue, flattish on the dorsal surface, with
a deep depression at the base of the ventral surface, in which develops
a minute bud which gives rise to a plant {16 — 14-^17). At the base it
is attached to a cylindrical hyaline stalk of some length. At the point
of attachment to the brood body it is constricted, and is here fragile,
so that the brood body is easily detached. The stalk is marcescent,
drying up iw situ. The brood bodies measure about 0.5 by 0.7 mm.
and contain an abundance of food in the form of fat and starch.

I received material of D. pygmaea collected by Dr. Pat Brough
near Sydney, N. S. W., in response to my request, on two occasions,
viz., in Nov., 1939 and in April, 1940. In the former no signs of gem-
mae were to be found; in the latter they were present in various stages
of development. In none of the specimens could brood bodies be seen
openly exposed, as represented in Goebel's drawing (1908). The plants
were perhaps still too young. The structure of the gemmae was as Goebel
described them. He suggested their homology with leaves, but it is to
be noticed that there is no suggestion of stipules. They arise in a
ring about a dished growing point, and stand in several ranks around
it {16 — 17). Around them young leaves have already begun develop-
ment, the older of these expanding. The gemmae seem therefore to
represent the culmination of a growth period, and they would be
set free during the winter season in the natural habitat. Professor
Buller suggests to me that the rosette, with its gemmae at the center,
may be regarded as a "splash cup", Hke those of the bird's nest fungi.

Of much more general occurrence is another method, namely, by
budding from the leaf. This is by no means of recent observation.
First seen by Naudin in 1840, it has been described by numerous
others, at least thirteen in number. The historical aspect of this mat-
ter has been well summarized by Behre (1929).

Naudin in D. intermedia (1840) and Kirschleger (1855) in D.
longifolia had observed the fact of budding from the leaf surfaces, and
that the origin was "probably endogenous" (Naudin). Nitschke's
account was sufficiently extended and exact so that Behre found little
to correct, so far as general morphology was involved. The earliest
anatomical study was made by Beijerinck (1886), estabUshing the exo-
genous origin of the leaf buds. Leavitt (1899, 1903, 1909) pointed
out that the earlier leaves of the leaf buds of even such extreme forms
as D. binata, have rounded leaves characteristic of D. rotundifolia,

Francis E. Lloyd —132— Carnivorous Plants

regarding this a repetition of phylogeny. Winkler (1903) observed the
lack of polarity in the occurrence of leaf buds in D. capensis, as in
Torenia and Begonia, and further for the first time showed clearly that
the buds arise not from retained embryonal tissue, but by redifferentia-
tion of the leaf tissues.

Exact studies of the mode of origin of the adventitious buds of the
leaf surface have been made by Behre (1929) and by Vickery (1933),
the latter author independently confirming the former in all essentials.
Such leaf buds arise on the blade always from the bases of tentacles,
usually on the adaxial surface, but occasionally laterally or even adax-
ially (Behre). They are often visible in a few days if during that time
the leaf has been separated from the plant and kept under moist con-
ditions. The cells involved have in all cases arrived at maturity, and
there is no sign of the persistence of embryonic tissue. The cells there-
fore undergo a true rejuvenation passing from an adult, vacuolated
condition into one of high protoplasmic content with accompanying
changes in the nucleus. They then undergo cell division previous to
growth, the earliest divisions, in general, being anticlinal, followed by
perichnal {16 — 7, 8). Increase in size now overtakes the newly active
cells, and a simple outgrowth emerges exogenously, this gradually in-
volving the whole of the base of the tentacle (Vickery) {16 — 9, 10).
The vegetation point having been defined at the scene of the earliest
divisions, this is now raised by the growth activities of the paren-
chyma of the upper moiety of the mother leaf in the immediate vicinity,
carrying up the tentacle so that this now appears to arise from the bud,
rather than the bud from it. Whether the new vascular tissue, that of
the bud, becomes articulated with the older, that of the leaf, is not
clear. Doubtless this occurs if the leaf does not decay, as observed by
Vickery. If, however, the leaf does decay, this may be questioned.
Robinson (1909) asserts that no connection occurs. The vegetation
point having been established, leaves appear on the bud and a new
plantlet becomes established, roots being formed secondarily. The
earlier leaves frequently show abnormalities, as I have observed,
such as the lateral fusion of contiguous leaf primordia, producing more
or less laterally doubled leaves. Nepionic leaves occur. Leavitt
(1903) was able to produce such even from the terminal bud by cutting
off the stem below it and removing the leaves as they expanded. D.
intermedia, which bears only radially symmetrical tentacles normally,
under such condition of ''malnutrition " bears on nepionic leaves
spoon-shaped lateral tentacles like those of D. rotundifolia. The fre-
quency and ease with which all this occurs, as already mentioned,
makes it probable that this method of reproduction rivals, in its re-
sults, reproduction by seed. I have at my hand now a small flower
pot which a few months ago carried three small plants sent to me from
Ontario by my friend Professor R. B. Thomson. These at the present
writing have died down to winter buds, and I count at least a dozen
minute plantlets which I observed to have arisen from old and at
length decaying leaves.

Behre has further described the origin of plantlets from the leaf
stalk. As in the case of the blade, such always occur on the upper
surface, with the exception of D. capefisis and D. binata, in which they

Chapter X — 133 — Drosera

may occur on the under side. Due to the different anatomical struc-
ture, the origin is more various, for the epidermis may not in all cases
take active part in the earlier cell divisions. These occur usually in
the vicinity of stomata or near the bases of trichomes or of sessile
glands, but can arise also on the stalk of the inflorescence or even from
the latter itself, as axillary buds however (Robinson 1909). Since the
flower stalk is radial in structure, the buds arise on all sides, and on
account of the closed cylinder of sclerenchyma, they never articulate
with the vascular system of the flower stalk. Adventitious buds may
arise from roots also, in which case they are, as would be expected,
endogenous in origin.

In the case of D. spathulata Behre found regeneration by bud for-
mation to take place indirectly from callus, previously formed on the
cut end of a leaf stalk.

Miss MouLAERT (1937) obtained adventitious buds from leaves,
isolated petioles, hypocotyls, stems, scapes, receptacles and sepals.
Following the formation of epiphyllous buds, she observed the develop-
ment of cushions of tissue ("bourrelets") extending from the base of
the plantlet toward the petiole. These are of three kinds, those which
remain as mere thickenings in the parenchyma above the veins, and
which she called "undifferentiated"; those which act as a liaison be-
tween the plantlet and a root which has already differentiated ad-
ventitiously nearby and in which a vascular connection between shoot
and root becomes established; and third, a kind which is formed near
the plantlet which does connect with it, an example of "affolement
cellulaire. "

Another observation made by Moulaert is the occurrence of ab-
sorbing hairs, structures quite like root hairs, which arise from the
upper surface of the leaf blade or from the basal part of the stem of
the plantlet. They are very abundant and their walls are brown as
in the case of root hairs.

Conditions determining the incidence of leaf buds. — It has gen-
erally been observed that the production of adventitious buds takes
place only under conditions of high humidity, and apparently the
higher the better. In order to obtain them the practice is to remove
leaves and place them on moist moss, or float them on water, in cov-
ered vessels (Graves 1897; Grout 1898; Ames 1899; Robinson 1909;
Leavitt 1903; Salisbury 191 5; Vickery 1933, and others). But the
matter seems not to be quite so straight-forward as this. Dixon (1901)
found that such buds occur on plants in abundance when they have
been allowed to dry out gradually on their bed of Sphagnum under a
bell-jar, during a period of two months. Confirmatory of this is
Behre's observation that leaves which had been removed and sus-
pended in a moist chamber, but not so moist as to prevent some wilt-
ing, will produce many buds. A too great plenitude of moisture
therefore appears to mask a delicate balance of affairs between the
leaf and its environment.

As to temperatures, Ames (1899) thought that low temperatures
were favorable. Vickery found a wide favorable range. My own
experience favors the idea that D. rotundifolia at any rate is active in
this way at prevailing cool temperatures.

Francis E. Lloyd — 134 — Carnivorous Plants

Wounding, necessary in such plants as Begonia (Goebel 1903), in
itself is of no influence (Behre). It has been thought that the removal
of the chief shoot (particularly the growing shoot) is a stimulus, that
is to say, the disturbance of correlations (Behre), which is attained
simply by the removal of the leaf. The weight of this point seems not
to be great since budding occurs in abundance on leaves still attached,
in the case of D. peltata, though rather more slowly than when the
leaves have been removed (Vickery). When some of the glands have
been injured or removed, the leaf will still produce buds, but only from
the bases of uninjured glands (Vickery), indicating that the gland may
contribute something in the form of a growth substance (see Coe-
LiNGH 1929).

Nevertheless Behre did find certain correlations. The removal of
the growing point always increased the leaf budding, though em-
bedding it in gypsum plaster did not. If the removal of the growing
point incited the development of an axillary bud, this itself would
inhibit bud formation, though if at the same time the vascular tissue
had been suitably cut, before the axillary bud was put into activation,
buds were formed. Behre further did this experiment: after removal
of the growing point the leaves were cut longitudinally in some cases
and transversely, but not sufhcient for amputation, in others. Only
on the outer parts of transversely cut leaves did buds arise, while on
plants similarly treated but with the growing point not removed, the
result was negative. Yet Behre recorded the occurrence of an ad-
ventitious bud on a leaf on a plantlet, itself produced adventitiously
from a scape, with the growing point active (1929, Fig. i). These results
with D. rotundifolia could not be obtained with D. capensis. But the
facts as they stand support the view that there is a delicate inter-
relation between the growing point and the inclination to regeneration
(Goebel) as observed in numerous other plants. Thus we are led to
consider what the internal conditions in the plants may be which de-
termine or control such phenomena. Here the food materials may
play a role or hormones may act as regulators, but this question is too
far away from our present purpose, though it may as well be pointed
out that Behre did experiments in which he reduced leaves to a con-
dition of pronounced hunger in darkness with the deprivation of CO2
and yet obtained regeneration, from which he concluded that "there
is no doubt that regeneration is put into activity by some other stimu-
lus than a surplus of nutrient materials," thus indicating the presence
of specific substances, hormones perhaps, which could procure the
results.

Polarity. — The fact of polarity is one of so general observation
that Behre naturally raised the question in regard to Drosera, finding
but little evidence that it obtains, except to a slight extent in the case
of D. capensis and D. filiformis. Adventitious buds are not related in
position to the stronger vascular strands, but are found scattered in-
differently, arising usually from the abaxial surfaces of tentacle bases,
though they may be found on the side or on the adaxial aspect. If
small pieces of the leaf blade are made, more buds arise than would
otherwise, and even on the leaf margin where they do not occur except
when a narrow band (i mm. broad) is made by a cut parallel to the

Chapter X — 135 — Drosera

margin. They never arise on the lower leaf surface. That buds arise
on other parts where there are no tentacles indicated that if it were
possible to obtain leaf pieces large enough and free of tentacles, they
would arise also from the upper leaf surface proper. The age of the
leaf makes little or no difference. It is remarkable in this connection
that even young leaves when removed from the plant will continue
their development under suitable conditions of light and moisture. If
the entire leaf, blade and stalk are removed, buds occur on the blade.
If the blade is then removed, buds occur on the outer end of the
petiole where there are tentacles, though not always on a tentacle base.
If the tentacle bearing part is now removed, a bud may arise at any
point, no polarity being shown. If now the conditions are so ar-
ranged that the petiole is kept moist and the blade relatively dry, the
petiole will regenerate instead of the blade. In D. capensis, however,
there is a distinct tendency for buds to appear near the leaf apex, this
species having long leaves with narrow blade which unrolls during
growth. This is true in both old and young leaves, and is regarded by
Behre as evidence of polarity. This polarity may be easily masked,
however, by placing a leaf with its petiole in moist sand and the blade
in the air, when buds now appear toward the basal end. Winkler
(1913) had observed a similar behavior in D. filiformis, which has long
cylindrical leaves. It is curious that the long slender leaves of such
species as D. binata do not exhibit the same tendency. The readier
production of buds near the leaf apex in D. capensis, but in the case of
young, not older scapes, is conditioned by the young state of the
tissues. The readiness of roots to produce buds is well known and
made use of for propagating exotic species, but here also they may
arise quite indifferently in position, and no polarity can be de-
tected.

Carnivory. — The attention of botanists was first attracted to
Drosera as an insectivorous plant by the observation that the tentacles
are capable of movement. This was made in 1779 (Hooker 1875),
when a physician of Bremen, Dr. A. W. Roth, noted as follows: ''that
many leaves were folded together from the apex toward the base, and
that all the hairs were bent like a bow, but that there was no apparent
change in the leaf stalk." When he opened the leaves he found cap-
tured insects, and was driven to compare Drosera with Dionaea, think-
ing that it had the same power of motion as the latter. He records an
experiment which he did. "With a pair of tweezers I placed an ant
upon the middle of the leaf of Drosera rotundifolia but so as not to
disturb the plant. The ant endeavored to escape, but was held fast
by the clammy juice at the points of the hairs, which was drawn out
by its feet into fine threads. In some minutes the short hairs on the
disc of the leaf began to bend, then the long hairs, and laid themselves
on the insect. After a while the leaf began to bend, and in some hours
the end of the leaf was so bent inwards as to touch the base. The ant
died in fifteen minutes, which was before all the hairs had bent them-
selves" (fde Hooker, 1875). At about this time (1780) similar ob-
servations were made independently by Dr. Whately, "an eminent
London surgeon" (E. Darwin: Botanic Garden, pt. 2, p. 24) as re-
ported by his friend Mr. Gardom, a Derbyshire botanist. "On in-

Francis E. Lloyd — 136 — Carnivorous Plants

specting some of the contracted leaves we observed a small insect or
fly very closely imprisoned therein, which occasioned some astonish-
ment as to how it happened to get into so confined a situation. After-
wards, on Mr. Whately's centrically pressing with a pin other leaves
which were yet in their natural and expanded form, we observed a
remarkable sudden and elastic spring of the leaves, so as to be inverted
upwards and, as it were, encircling the pin, which evidently showed the
method by which the fly came into its embarrassing position."
(Withering 1796). It is unfortunate that Dr. Whately did not
record his observations himself since the rate of movement seems, by a
trick of memory, to have been exaggerated by the writer, Mr. Gardom.
As late as 1855 the facts were denied by Trecul, but in i860 Nitschke
made a thoroughgoing study, substantiating the earlier observations,
to be followed by Darwin, who had been heralded both by Hooker
and by Asa Gray, to whom Darwin had previously communicated
his results. Of 267 pages of Darwin's book on Insectivorous Plants
230 are devoted to an extraordinarily minute examination of the activ-
ities of Drosera, attesting to his immense patience and determination
to uncover every secret possible.

Following Darwin various trends of investigation can be followed.
His observation of the phenomenon of aggregation was the beginning
of numerous studies of the cytological changes in glandular and other
cells, summarized by Homes. Other trends have been in the field of
anatomy, already discusssed, of digestion and nutrition and of the
nature of the movements, all to be duly considered.

Mucilage. — While the papillae have not been observed to throw
off secretion, unless it be water (Konopka), the glands of the tentacles
are very conspicuous because each bears a drop of mucilage of high
viscidity, clear and ghstening, secreted by and supported on it (ij — 9).
The glands are charged with red pigment, so that the shining drops of
mucilage lend to the leaf a brilliant red hue. Since these persist as
well during the sunshine as otherwhile, we have the name "sundew"
common among Europeans. This mucilage, because of its briUiance
and reflected color, may be interpreted as a visible lure; it is at all
events an effective means of capturing prey of small dimensions, if it
ventures to alight on the glands. A delicate fungus-like odor which
has been detected by various observers (Geddes) may be an additional
factor of allure. The insect caught is soon (Nitschke: 15 min.) wet
all over and smothered by the secretion, which upon stimulation is said
to flow more freely. Darwin investigated the secretion activity on
the application of various kinds of substances and found that not only
does the secretion increase in the gland directly stimulated, but in
nearby glands as well, as the result of transmitted stimulus. When
the stimulating material is nitrogenous the secretion becomes acid, sup-
plying an important condition for digestion. The amount of secretion
which becomes applied to the captured prey is increased not only by a
more ample supply of secretion, but by the movement of the tentacles
which bring more glands than originally stimulated into contact with
the prey. The secretion, Darwin showed, is possessed of antiseptic
properties, and thus inhibits the action of bacteria. In his experiments
he found that bits of meat and of albumen placed on the Drosera leaf

Chapter X — 137 — Drosera

underwent changes, shown to be due to digestion, and were found to
be free of bacteria, while similar pieces of material placed on wet moss
"swarmed with infusoria."

Chemically the mucilage appears to be a sort of hydrocellulose, but
the seat of its secretion is not known. Like other cases of mucilage it
may be a product of an alteration of the cell wall, or it may be an
exudation from the protoplast. In any event it is permeated by other
substances in which its power of digestion rests — enzyme, acid, some
antiseptic substances, and latterly Weber has suspected the presence
of ascorbic acid.

Small (1939) has advanced the notion that the mucilage is se-
creted only by the lateral cells of the gland, and not by the apical cells.
His evidence is seen in internal reflecting surfaces, stated to be present
at the apex and absent from the lateral cells, between them and the
mucilage. For my part I fail to find such reflecting surfaces. On the
other hand, if a piece of leaf with glands which have been thoroughly
wiped off with filter paper is placed in parafhn oil and carefully ex-
amined to find glands on which no trace of mucilage is visible, these
in the course of one to several hours will show numerous droplets of
mucilage oozing away from the surface as well at the apex as on the
sides of the gland (75— 14). Weber (1938) by means of sodium
oleate has demonstrated to his own satisfaction rods or streams of
mucilage radiating from the gland surface at every point. I have not
been able to confirm this. If the glands are watched under a binocular
dissecting microscope, in the course of a short time it will be noticed
that the surface of an opalescent mucilage drop is wrinkled longi-
tudinally, and by this time the surface of the drop has lost its glassy
look. It is evident that there is a surface concentration of some sub-
stance or substances. As one watches steadily, one sees an occasional
explosion on the surface as if some minute particle or droplet had on
arriving there from inside immediately spread over it. As the wrink-
ling progresses the drop becomes pear-shaped, the broad end above the
apex. With cessation of evaporation, the drop will assume its oval
form. The mucilage is a jelly-hke mass. If two glands with drops are
approached so that they touch and then are moved apart, the drops
will largely separate, adhering by only a slender thread. If a drop is
touched with a corner of filter paper at its basal margin and, on ad-
hering, the mucilage is pulled away upwards toward the gland apex, it
will tear away and extend asymmetrically from the gland apex. When
a drop is pulled out, it at first refuses to leave the gland. Only when
there is sufficient adhesion and pull, the whole mass, after a certain
amount of stretching, will pull away suddenly. These and similar
evidences indicate that the mucilage has a sort of structure. When
dry, it shows double refraction, but not when wet (Weber). It is not
so stiff a jelly as that of Drosophyllum, which pulls away readily in a
mass, but is otherwise similar.

One other apparently trifling observation which I have made may
be mentioned here. I have noticed that, over the apical half of a
gland there are in the immediate vicinity of the gland surface minute
plaques of clear colorless substance not soluble in sulfuric acid, rounded
or sometimes angular in shape {15 — 10). Sulfuric acid dissolves the

Francis E. Lloyd — 138 — Carnivorous Plants

mucilage, and the cuticle remains intact. They might be delicate
flakes of cuticle exfohated from the remaining cuticle, but of this there
is no certainty.

Movements of tentacles and leaf blade. — We must go back to 1782
to find the first record of studies of the modes of behavior of the ten-
tacles and leaf blade. These were carried on by the above mentioned
Dr. Roth, botanist as well as physician. He was stimulated to study
Drosera by reading Ellis' letter to Linnaeus in 1770 announcing the
discovery of Dionaea muscipula; and in his essay he makes cogent
comparisons between these two, the only then known carnivorous plants.

According to Roth, if an ant be placed on a leaf, the glands re-
spond by bending, first the centrally placed, then, but much more
slowly, the glands most distant. Finally the leaf blade bends either
transversely, its apex approaching its base, or if the stimulus, say a
small fly, has been placed laterally, the side may bend over. The
rates of movement depend on external conditions, and are most rapid
in warm sultry weather. He remarks that D. longifolia reacts more
readily than D. rotundifolia, and that rain reduces sensitivity.

The next contribution of major importance, by Nitschke, did not
appear till i860, eighty years later. Meanwhile, however, several bota-
nists had observed and discussed the matter. Somewhat previous to
1835 ^- P- °E Candolle had observed the response of the tentacles.
Treviranus (1838) quoted Roth (1782) but said that he failed to get
the results described by him. Hayne (date about this time, see
Nitschke i860) saw the response of the tentacles and that, at length,
the leaf blade bent and became spoon-shaped. In 1837 Meyen re-
viewed previous observations and while he could confirm the fact that
the tentacles as also the leaf blade were bent, he maintained the idea
that it was due not to irritability, but to the activity of a struggUng in-
sect pulhng over the tentacles toward itself. Milde (1852), however,
put this right by experiment. He placed small flies on the leaf, and ob-
served in 5 min. the outer tentacles bending inwards. Next day
the whole leaf was bent, and in 5 days again unrolled. A useful skep-
tic appeared in 1855 in the person of Trecul, who thought that the
insects were caught by young leaves which then retained their youth
position. Came then Nitschke (i860), who was the first to attack the
problem in a sustained way and with a critical attitude. His first
argument was directed against Trecul, and he established the general
correctness of Roth's observations. He believed that when a stimulus
has been applied at some point by the apphcation of an insect, the
surrounding tentacles bend their heads directly toward this point,
whether the position of the stimulating object is central or lateral.
The marginal tentacles move, he says, always in the "most direct"
path toward the point of stimulation. On this point the reader is re-
ferred to the work of Behre beyond. Nitschke regarded the behavior
as an expression of true irritability, and that Meyen's view that the
action of the tentacles is purely passive is wrong for a number of
reasons, especially cogent being the fact that young leaves do not se-
crete mucilage, and that neither they nor aged leaves are sensitive.
First when the leaves are widely open and rich in secretion is this the
case; even dead insects procure movements, if indeed somewhat less

Chapter X — 139 — Drosera

vigorous ones. He found, however, no response to simple mechanical
stimulation, but this was found later to be wrong. Equally so his view
that a stimulating body attached to the back of a leaf induced re-
sponses whereby the tentacles turned backward to embrace the body
quite as well as forward. He found that the leaf may repeat the per-
formance after recovery on renewal of secretion, and further that the
effect of a given stimulus depends on the distance it has to travel. The
movements can take place under water and in response to soHd bodies
and acids in weak solution. The rate of response is aflfected by tem-
perature but not by Hght. It is then chiefly to Nitschke and to Dar-
win that we owe many original observations which furnish a picture of
the direction and rapidity of the movements of the leaf and tentacles.
The general facts first and most readily observed are the following.
If a suitable stimulus is received by any group of leaf tentacles, say
near the middle of the leaf, or on or near the "disc" in the case of
D. rotundifolia, in the course of a few minutes a bending of the near-
by tentacles is to be observed until, the stimulus evidently travelling
radially, it reaches even the extreme marginal tentacles which then
bend over. If the stimulus is sufficient even the leaf blade responds
in like manner. Goebel figures a leaf of D. intermedia which had com-
pletely folded over to embrace the body of a large fly which had been
caught. D. capensis was found to be particularly good at this. I
placed a single Drosophila flylet on a leaf and in the course of time the
marginal tentacles, as well of course as those nearby, had responded.
Finally the whole apex of the leaf bent over (ij — 8). With regard to
the leaf blade not all the species of Drosera behave in this way. Goe-
bel observed that D. hinata does not, nor does D. dichotoma, and
probably others. From such observations it is evident that the stim-
ulus received by a tentacle travels to its base and radially from there
to neighboring tentacles, which then respond. A casual glance at a
leaf displaying these responses, one in which the tentacles are bent
over towards the middle of the disc (speaking of D. rotundifolia) sug-
gests that the normal movement of the tentacles is along radial lines.
The dorsiventral flatness of the tentacles would seem to condition them
to move thus. But Nitschke saw that the matter is not so simple.
He said that the tentacles receiving the stimulus bend over in the di-
rection of the point at which the stimulus was received, irrespective of
its position, so that, if a fly is caught at some eccentric point, the
tentacles affected bend over toward this point and not toward the
center of the disc. Apparently the direction of movement of the stim-
ulus determined the appropriate direction of movement of the tentacle.
There is an apparent exception to be noted. Darwin found that
when a marginal tentacle is stimulated, it bends over, but no response
is called forth in the neighboring marginal tentacles. Only when the
marginal tentacle originally stimulated brings its glands with its stimu-
lating material into contact with the glands of the disc, is a stimulus
provided by the latter which now calls forth a response of the mar-
ginal tentacles hitherto not affected.

The duration of the response depends on the nature of the stimu-
lus. Here I quote from Darwin (p. 19) "The central glands of a leaf
were irritated with a small camel's hair brush, and in 70 minutes

Francis E. Lloyd — 140 — Carnivorous Plants

several of the outer tentacles were inflected; in 5 hours all the sub-
marginal tentacles were fully inflected; next morning after an interval of

22 hours they were fully expanded I then put a dead fly in the

center of (a) leaf, and next morning it was closely clasped; five days
after the leaf reexpanded and the tentacles, with their glands sur-
rounded by secretion, were ready to act again."

A given stimulus acting somewhere on one side of the leaf will affect
the marginal tentacles on that side sooner than those of the other side
further away; or indeed, only one side of the leaf may be called into
action. In the case of a cup-shaped peltate leaf {D. gigantea) I have
observed that the total result of such movements is to bring the prey
into the depths of the cup, where, in the course of time, only the chitin-
ous remains of the captured insects are to be found. This result is
perhaps contributed to by the surface tension of the drop of secretion
which more or less fills the cup.

It was thought by Nitschke that even the back of the leaf could
accept stimuli and transmit them to the tentacles, but Darwin was
unable to cause any response by stimulating the leaf blade proper, on
the front or the back. In order to locate the sensitive or sense per-
ceptive points, Darwin removed the gland from a tentacle, whereupon
the latter made a brief response by slightly bending but soon regained
its erstwhile posture. When stimulus was applied to the cut tentacle,
no response followed. But if now the disc tentacles were stimulated,
the amputated tentacle responded, as if the head were not missing.
The stalk of a tentacle, no more than the leaf or petiole, can receive a
stimulus. In any event, the marginal tentacles are not so sensitive as
the rest, nor are they affected by rain drops. Small (1939) denies
this. That the disc tentacles are more sensitive may appear to be the
case because the stalks of these are very short, and the tentacles are
closer together so that a given stimulus does not have to travel so far
to elicit response. And although the stimulus travels radially from a
point of stimulation, Darwin found that it travels more readily longi-
tudinally than transversely across the leaf blade. The stimulus may
travel quite across the blade so that when it is applied to the tentacles
on one margin, those of the opposite may respond; but in spite of
repetition of the stimulus, the opposite tentacles will open again, from
which Darwin argued that the "motor discharge must be more power-
ful at first then afterward." It was asked by Darwin whether the
motor impulse travels through the vascular tissue, but this turned out
not to be the case, certainly "not exclusively," for the tentacles of a
group surrounding the point of stimulus will respond all at a uniform
rate notwithstanding the fact that the vascular connections are very
unequal in length; indeed the course of the vascular tissues in the leaf
as a whole does not permit the view in question when the uniformity
of response of the tentacles is considered.

The intensity of the stimulus necessary to procure response was a
matter of much concern to Darwin. He endeavored to get some
measure of intensity by weighing small pieces of hair, etc., which would
prove efficient. The following quotation embodies an expression of his

reflections on this " it is an extraordinary fact that a little bit of

soft thread 1/50 of an inch in length and weighing 1/8197 of a grain,

Chapter X — 141 — Drosera

or of a human hair 8/1000 of an inch in length and weighing only
1/78740 of a grain (.000822 milligram) or particles of precipitated
chalk, after resting for a short time on a gland, should induce some
change in its cells, exciting them to transmit a motor impulse through-
out the whole length of the pedicel, consisting of about 20 cells, to
near its base, causing this part to bend, and the tentacle to sweep
through an angle of above 80 degrees".

It was generally conceded by both Nitschke and Darwin that
dead bodies do not provoke so much response as hving and therefore
moving bodies. This was explained by Pfeffer by pointing out that
mere constant contact does not produce response, but that there must
be both direct contact with the gland and friction on its surface. The
mucilaginous drop can prevent direct contact as in the case of rain or
quicksilver (which Pfeffer tried) or even particles suspended in it un-
less by their weight they fall against the sensitive surface. That the
minute particles of hair used by Darwin should produce the results
observed may be understood better when, as Pfeffer showed, vibra-
tion of the table or floor causes movements of such particles on the
surface of the gland sufficient to stimulate it.

In addition to non-living substances, Darwin tested the reactions
of the tentacles to a large variety of organic materials with the purpose
of determining what digestive juice or juices are secreted by the leaves
of Drosera. His contribution to the problem of digestion will more
suitably be considered under the appropriate caption beyond. Here
it will be mentioned that he seemed to regard the movements of the
tentacles and the length of time they remain inflected as evidence of
the nutritional value to the plant of the material exposed to them.
But he himself records a various behavior of the tentacles in this re-
gard. He says in conclusion "The substances which are digested by
Drosera act on the leaves very differently. Some cause much more
energetic and rapid inflection of the tentacles and keep them inflected
for a much longer time, than do others. We are thus led to believe

that the former are more nutritious than the latter " This

generalization can hardly hold. Robinson found that pure creatin was
digested but caused no bending of the tentacles. As Schmid points
out, Darwin's work, rightly or wrongly, led emphasis to be too strongly
placed on the Drosera mechanism being an adaptation for the obtain-
ing of protein nutrition. While it is true that, to quote Darwin again,

" inorganic substances, or such substances as are not attacked by

the secretion, act much less quickly and efficiently than organic sub-
stances yielding soluble matter which is absorbed" it is also true that
some nitrogenous bodies equally do not, and therefore it is impossible
to formulate a rule. Darwin himself records the failure of urea to
procure movements. What explanation serves when HCl, boric acid,
malic acid and camphor stimulate to movement when Ca, Mg and K
salts generally do not? And ammonium phosphate was found more
energetic than other ammonium salts though containing less nitrogen.
But because potassium phosphate is taken up Darwin argued a need
for phosphorus. Schmid, considering this phase of the insectivory
problem, himself tested the action of pure salts and concluded that the
movements of tentacles alone cannot lead to any real index of the
value of insectivory from the nutritional-ecological point of view.

Francis E. Lloyd

— 142

Carnivorous Plants

As little indeed may one thus argue as about the nutritional value
of food taken by man from the action of the salivary glands, adds
ScHMED. It seems proper to conclude that the reactions of the ten-
tacles are general rather than specific. The length of time they remain
inflected, however, seems, in the absence of injury (several times noted
by Darwin) to be generally correlated with their opportunity for ab-
sorption.

Mechanism of tentacle movement. — Nitschke pointed out that al-
though the tentacles can bend, there are no special motile organs, such
as occur e.g. in Mimosa. What then is the nature of the bending
movements of the tentacle? Though Darwin obtained no hght on this
question, it was answered by Batalin (1877). He made spaced marks
on the sides of the tentacle, and found that after a movement was com-
pleted, the distances had increased. When the recovery is complete,
these distances are maintained, showing that the bending is a growth
phenomenon. This was shown true also of the leaf blade. H. D.
Hooker (191 6) investigated the matter more thoroughly. In making

Fig. 3. — Drosera rotundifolia. — A, Side views of a tentacle in process_ of bending,
beginning with the bottom figure; B, same in process of unbending, beginning with the
top figure; C, Side views of the same tentacle before and at close of the reaction (after
Hooker).

his measurements of the tentacles during bending he made use of
natural marks supplied by the minute sessile glands to be found on the
surface of the tentacle stalk. By means of these measurements and of
camera lucida drawings, he got a detailed record of changes in dimen-
sions during bending and recovery. A set of his drawings are here re-
produced (Text fig. 3). Hooker found, as did Batalin, that the
movement, whether bending or unbending, is a growth phenomenon.
During bending acceleration of growth begins near the base along the
back (the convex surface) of the tentacle, and moves upward during the
bending phase, so that the tentacle end moves through an^ angle
of 215 to 270 degrees, beginning the movement within 1.5 minutes,
completing it in a few hours, or sometimes in as short a time as 17
min. 30 sec. (Darwin). The unbending movement results from in-
creased growth on the now concave side, and takes place at once if the
stimulus was a brief one, or is delayed as when the tentacles have
closed over prey. Here also growth begins near the base and moves
upward toward the gland. During neither phase is the growth neces-

Chapter X — 143 — Drosera

sarily limited to one side, but the difference of rate is obvious and
produces the same result. Since the movement of the tentacle is a
matter of growth, and since there is a limit of total growth, the num-
ber of times bending may be repeated is limited. Darwin found the
number is three, and this was confirmed by Hooker. Though the two
movements constitute practically a continuous reaction, at least when
a single brief stimulus is originally applied, the unbending reaction fol-
lows a stimulus inherent in the internal conditions (such as tissue ten-
sions) set up during bending, and is tropic (autotropic, autonomous,
Behre) in nature. Since the entire armament of tentacles may not be
used in any one grasping of prey, the leaf as a whole may react more
than three times, even though a single tentacle cannot. The short
radially structured tentacles of the disc do not react by bending to
stimuli applied directly to the glands, but only to stimuli received
through the glands of other tentacles. Hooker regards the response as
tropic while the original response of the lateral tentacles is evidently
nastic, though the unbending response is tropic. Both Darwin and
NiTSCHKE recorded their behef that marginal tentacles when stimu-
lated indirectly bend toward the point of stimulation. Hooker takes
exception to this, saying that he was unable to get evidence of it,
and thinks that they normally bend toward the middle of the disc, that
is, nastically.

Exceptions to this he thought "to be purely accidental." None-
theless, Hooker was sufficiently impressed with his observations
to state that "most of the marginal tentacles which reacted to the
conducted impulse" from the discal tentacles "in bending toward the
center of the leaf bent hkewise in the direction of the source of excite-
ment. " The bending of the discal tentacles is, however, always toward
the point of original stimulation, and cannot be stimulated directly.
The response is tropic, but "in all probability" the movements "are
likewise the result of differential growth on opposite sides" of the
tentacle base. The method used was not applicable to the determina-
tion of this fact.

Behre (1929) admitted that Hooker's conclusions were nearly
right, but was evidently impressed by the discrepancies admitted
by him. He, therefore, attacked the problem at this point, and ana-
lysed the movement of the tentacles more rigorously, controlling his ob-
servation by means of a horizontal measuring microscope with a scale
in the field. He recorded accurately the movements of tentacles rela-
tive to each other and to the position of the source of stimulation, and
made them available to the reader by means of maps showing the
paths of movements.

In the case of D. rotundifolia he found that, according to their be-
havior the tentacles can be divided into three groups, namely, {a) mar-
ginal, the outermost standing exactly on, or very near to the leaf mar-
gin; {h) an outer zone of discal tentacles of one to three rows, called by
him the "surface outer tentacles"; and (c) the discal tentacles within
(&), or "central tentacles". With some sHght differences due to the
posture of the tentacles, the same holds for other species investigated
{D. binata, intermedia, capensis, spathulata). His observations yielded
the following results, and here it may be injected that he used in most

Francis E. Lloyd — 144 — Carnivorous Plants

cases small and uniformly sized objects for stimulation, viz., the eggs
of the wood-louse.

The responses of the strictly marginal ("outer marginal") tentacles
are somewhat slower than those standing just within the margin ("in-
ner marginal")- Their reaction to a direct stimulus (that is, one ap-
plied to the glands of these tentacles) is, however, always strictly
nastic; their function is to bring the prey into contact with the discal
glands. The sensitivity and quickness of reaction are surprising. The
reaction may begin in lo seconds, and was seen to make a complete
excursion of i8o degrees in 20 seconds, the movement being visible to
the naked eye. This was a maximum, however. It must be clear that
the direction of movement is in a single plane normal to the leaf mar-
gin. Prompt and rapid as their response to direct stimulus is, they re-
spond to indirect stimulus, derived from stimulated discal tentacles,
only slowly and weakly. At best, a reaction may be detected in 10
minutes, but the total excursion is short. Only when the leaf is heavily
fed, especially with living prey, do the marginal tentacles indirectly stim-
ulated actually reach the prey. If the stimulus is derived from a small
insect, the excursions of the marginal tentacles are incomplete, are soon
reversed and can be of no use to the plant, though, since complete
bendings can occur only three times at best, the meagreness of re-
sponse may be regarded as an economy of effort. Full expenditure of
effort is made only when the prey falls on the marginal tentacles, when
by bending fully they bring it into contact with the inner tentacles
thus exposing it to much greater digestive surface. The movements
are at first nastic. Since in D. rokmdifolia the orbicular form of the
leaf results in nastic and tropistic reactions acting in the same direc-
tion, the observer is and has been naturally deceived. Only when the
reactions are observed in such leaves as those of D. intermedia and D.
binata is it seen clearly that, while the reaction of the marginal ten-
tacles to direct stimulation is at first nastic, in the course of the ex-
cursion the direction of movement may be modified by tropistic re-
actions, especially clear in D. binata, and in this is the account of
Hooker amplified.

In the case of the central or discal tentacles, there is no response to
direct stimulus, that is, stimulating material placed on a single tentacle
produces no movement in that tentacle. But the stimulus is quickly
transmitted to nearby tentacles and these then bend toward the point
of stimulation, that is tropistically. The rate of reaction is here much
more dependent on temperature — from an hour to 20 or so, according
to circumstances.

Between the central disc tentacles and the marginal lies a narrow
zone of outer surface tentacles, in size grading between them, being in
such species as D. binata as long as the marginal tentacles, or longer.
Their reactions are more complicated than those of the tentacles of the
other two zones, since they combine properties of both. They react
nastically to direct stimulus and as rapidly as the marginal tentacles,
and this character distinguishes them at once from the central disc
tentacles. Toward indirect stimulus their reactions are both nastic and
tropistic, and the resulting excursions are rapid and more extended than
those of the marginal tentacles to indirect stimulus, and result in bring-

Chapter X —145— Drosera

ing the glands into contact with the prey. The tropistic movement is
slower. The case of D. hinata well illustrates the behavior, because of
the cylindrical form of the leaf. A small fragment of meat was placed
on an outer surface tentacle. This responded at first quickly, and in
the course of five hours brought the prey into contact with the discal
tentacles. In two hours the nearby outer surface tentacles began their
excursions which were at first (for four and a half hours) nastic. The
next morning it was evident that tropistic movements had set in,
since by then all the glands were in contact with the prey. When,
however, in this species the stimulus is applied to the discal tentacles,
the reactions of the outer surface tentacles are entirely, or very nearly
entirely, tropistic. The case of D. capensis was of peculiar interest,
since in this species stimulus of an outer tentacle procured tropistic
reactions of its neighbors so that their glands would have travelled
the shortest way to the place where the prey was deposited on the
discal tentacles (the completion of the movement was not observed by
Behre) and not as in D. hinata, at first nastically (carrying the glands
away from a direct path) and later tropistically, correcting the error.

Behre, having pointed out such minor differences in behavior as
between different species, remarks that, since the nastic and tropic re-
sponses are influenced differently by different temperatures, as when
the nastic responses are arrested by a high temperature while the tro-
pistic are stimulated, such differences may account in part at least for
various behaviors. By and large, however, the various species act in

the same way.

Aggregation. — Darwin observed that, following stimulation, the
contents of the gland cells first and later of those of the pedicel, dis-
play changes in appearance due to a rearrangement of the protoplasm
and vacuole which he termed "aggregation." The total effect is suf-
ficient to be seen by the naked eye, if pigment is present, in the change
of color of the gland. In this way it is possible to follow the direction,
if not the extent of the movement of a stimulus. While Darwin's de-
scription of these changes was incorrect, they stimulated a great
amount of work directed toward their elucidation. Those who have
seen at Down House the tools Darwin worked with may well wonder
at the extent and acuteness of observation which characterize his work
in this particular. Taken with the general state of the knowledge of
the cell in his day, the observations of Darwin are the more sur-
prising.

Darwin gives his observations as follows: "If .... . tentacles that
have never been excited or become inflected be examined, the cells
forming the pedicels are seen to be filled with a homogeneous purple

fluid. The walls are Hned by a layer of protoplasm ". " If a

tentacle is examined some hours after a gland has been excited by re-
peated touches, or by an organic or inorganic particle placed on it, or
by the absorption of certain fluids, it presents a wholly changed ap-
pearance. The cells, instead of being filled with a homogeneous purple
fluid, now contain variously shaped masses of purple matter suspended

in a colorless fluid By whatever cause the process may have

been excited, it commences with the glands, and then travels down the
tentacles The Httle masses of aggregated matter are of the

Francis E. Lloyd — 146 — Carnivorous Plants

most diversified shapes, often spherical or oval, sometimes much
elongated, or quite irregular with thread- or necklace-like or club-
formed projections they consist of thick, apparently viscid mat-
ter ... . " " .... these Httle masses incessantly change their form

resembling the movements of Amoebae, or white blood corpuscles."
" We may therefore conclude that they consist of protoplasm."

Francis Darwin in 1876 concurred with his father, but later in
1888 reversed his position, pointing out that Darwin was in error in
thinking that the aggregated masses consisted merely of protoplasm,
but that they are concentrations or precipitations of the cell sap, and
that their amoeboid movements are the result of streaming protoplasm
which moulds the passive masses into a variety of forms" (Darwin,
2d. ed. 1875, note by Francis Darwin, p. 34) in agreement with
Pfeffer's views as pointed out in his Osmotische tlntersuchungen. Fran-
cis Darwin's volte face resulted from the pubHcation of views by
ScHiMPER, by Gardiner, and by de Vries. These we presently ex-
amine. ScHiMPER made his studies while in the U. S. A. where he was
evidently impressed with his opportunities. He examined Sarracenia
purpurea, Drosera intermedia and Utricularia cornuta.

Examining the epidermal and subepidermal cells of the tissues of
the lower part of the pitcher of Sarracenia, when such cells had been
exposed to nutrient substances, he observed that they showed, in con-
trast to those not fed, the following behavior. The single vacuole
containing tannin was found to be now broken up into two or more,
becoming, because of the concentration of their tannin solution, more
highly refringent. These vacuoles were found not to be suspended in
the cell sap, but themselves represented the whole of the sap, and were
found now to be suspended in a swollen protoplasm. " That under
the influence of certain substances the protoplasm attains a greater
capacity for swelhng seems probably to be of direct significance for
nutrition."

Recalling Darwin's statement that the aggregations are suspended
in the cell sap, Schimper examined Drosera intermedia tentacles. Here
he found, as in Sarracenia, that the protoplasm is swollen, the tannin
bearing vacuoles contracted. " By plasmolysis (with NaCl) it is seen
with the greatest clearness that here also that which appears to be the
cell sap is really only the much swollen protoplasm. After extraction
with alcohol, the protoplasm remains as a beautiful framework of
meshes."

Gardiner in 1886, apparently without having seen Schimper's
paper, described his own observations thus. " The chief phenomena

induced in the stalk cells" " most marked when stimulated by

food" "are that the protoplasmic utricle swells up and encroaches

on its own vacuole, that granules appear in the protoplasm and that
the movement of rotation increases in vigor." The cell becomes less
turgid. " The protoplasm in swelling abstracts water from its own
vacuole and in so doing leaves the sap in a more concentrated con-
dition." Going on to describe the protoplasmic activity of move-
ment he says that the reduced vacuole becomes fragmented, the re-
sulting small vacuoles become droplets, pear-shaped bodies and long
string-like processes (77 — i), just as described by Darwin.

Chapter X — 147 — Drosera

De Vries' studies were most illuminating. He examined cells of the
tentacle stalk. He says that the whole process of aggregation falls into
two periods. In the earlier period there is a pronounced increase in the
rate of cyclosis of the protoplasm, accompanied by growing complexity
of the currents (" differentiation"). Many accounts ignore this,
though Gardiner mentioned it. During the second period there is a
breaking up of the vacuole into a varying number of smaller ones, the
more obvious phase usually seen. These periods are not sharply de-
fined, the first passing over gradually into the second.

The rapidly circulating protoplasm, with its breaking up into new
streams, furnishes a mechanism for subdividing the originally single
vacuole which in the meantime loses some of its sap. This escapes
through the wall of the vacuole (the " tonoplast") into the space be-
tween this and the protoplasm. This escaped sap retains its osmotic
pressure, since tentacles in aggregation are as rigid as otherwise. Left
behind, however, are the pigment, albuminoids and tannin which can
be precipitated within the resulting vacuoles. The vacuol