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 Plants

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 vacuoles, however,
are now less rigid and more readily broken up by the cutting streams
of protoplasm. There result eventually many lesser vacuoles as drop-
lets of various shapes, especially slender tubular ones, constantly stirred
up by the moving protoplasm, and thus constantly changing positions.
De Vries attempts an interpretation of these changes by suggesting
that the heightened circulation of protoplasm may serve to facilitate
the movements of nutritive materials absorbed by the glands; that
the contraction of the vacuoles is connected with the partition of sub-
stances (acids, enzymes).

Here it should be pointed out that the process of aggregation,
thought to be observed in stimulated tentacles, is quite independent of
the process of inflexion of the tentacles, as Darwin pointed out. I
myself have observed that the bending of the tentacles occurs in a
region where no aggregation had taken place. As pointed out by Jost
(Benecke-Jost, 1924) aggregation takes place in a primarily stimu-
lated tentacle downwards from cell to cell, but also later in those
tentacles secondarily stimulated, but does not in these precede the
bending movement, and moreover proceeds not from the base upwards
but from the gland downward. Aggregation can, therefore, have no
relation to the transmission of stimulus, likely enough as it first seemed.

I have mentioned above that, on the escape of sap from the vac-
uoles, there is left behind a variety of substances, pigment, tannin,
albumin, which can be precipitated by various chemical agents (al-
kaloids, weak bases) and then appear as minute droplets or granula-
tions which coalesce into larger ones, and which in the case of tannin
and albumin can become brittle masses. Such precipitation was con-
fused by Darwin with true aggregation. Gardiner called this " pas-
sive aggregation," Goebel " granulation." Glauer (1S87) {fide F.
Darwin in Darwin, 1875) and Bokorny (1889) extended the dis-
tinction, the latter recognizing other kinds of aggregation to the num-
ber of four, viz.: the contraction of the entire protoplasmic utricle,
contraction and division of the vacuoles {'' true aggregation"), the
precipitation of albumin in the vacuoles and fourth, in the protoplasm.

The albumin in question has been called " active albumin" by O.

Francis E. Lloj'd — 148 — Carnivorous Plants

LoEW and Bokorny, so designated by them because of its imputed
peculiar properties which place it in a category of substances which
may be regarded as bridging the gap between the non-living and the
living parts of the cell. These peculiarities were recited by Erna
Janson, who, working in Loew's laboratory, examined aggregation
from the point of view thus indicated. Her paper (1920) cannot be
said to indicate a full apprehension of the observations of those workers
(Gardiner, de Vries, Akerman, especially the last two named) who
had described in much detail the curious and complicated happenings
which take place during aggregation in the Darwinian sense. Before
resuming Janson's work, it would profit us to look first at that of
Akerman, who while differing from de Vries in the interpretation of
certain details, nevertheless agrees with him about the general trend of
affairs. Akerman used pepsin as the stimulating substance which,
when applied to the gland, quickly causes movement responses and ag-
gregation. The course of events he described as follows:

At first the peripheral protoplasm is thin, and displays rotation
(cyclosis). This movement, as threads and ultimately plates of proto-
plasm arise, becomes more comphcated, and changes into a true circu-
lation, becoming more and more active. Meanwhile the peripheral
layer of protoplasm thickens, even to twice the original thickness.
Folds of protoplasm, impinging on the vacuole, become strands which
become thicker and more and more extensive till they cut into the
vacuole and ultimately break it up into numerous small ones. The
change in volume of the protoplasm is accompanied by a reduction in
volume of the vacuoles, the substances in solution therein (the red pig-
ment, tannin, etc.) becoming more concentrated till the specific gravity
of the two are reversed, that of the sap increasing, as shown by cen-
trifuging. At length the vacuoles display remarkable activity. They
elongate, become vermiform and move, creeping about each other in a
most dramatic fashion. In my own experience it has been found diffi-
cult to make drawings of them as the movements, apparently slow, are
fast enough to defy adequate representation of the proceedings. Aker-
man affords a fair idea of the condition regarded for the moment as
static in his Fig. 3, but even this does not do the matter justice. But
the photographic reproduction of three exposures of a single cell two
and fifteen minutes apart and two two minutes apart, are truly
illuminating (77 — i). We may here recall de Vries' own figures, not
to be ignored. Such activity can be sustained for hours, even days.
At last a reversal of changes sets in and proceeds till at length the
primary condition is attained. During the forward progression
Schimper thought to have observed the formation of new vacuoles,
new evidently as they did not contain red pigment; de Vrees also.
These Akerman did not see. He also supports Gardiner, Schimper,
Goebel in his view that there is no separation and independence of the
vacuole walls {" tonoplast" of de Vries) from the peripheral proto-
plasm, and no accumulation of water between them. Indeed, he goes
so far as to state it as his considered opinion that the important and
characteristic feature of aggregation is the swelling of the protoplasm,
the vacuolar phenomena being resultant and secondary. In support
of this view Akerman again tested a long series of substances and

Chapter X — 149 — Drosera

found that some (albumin, pepsin, peptone, phosphoric acid, ethyl
alcohol) cause swelling of the protoplasm and the accompanying ap-
pearances of aggregation, while others (basic substances such as am-
monia, carbonates of ammonia, sodium or potassium, alkaloids), do not
cause protoplasmic swelling, but only a precipitation in the vacuole
with unchanged volume. Further, that these latter substances can in-
hibit the action of the former.

The difference between true aggregation and " granulation" in
which merely precipitation occurs within the vacuole, is emphasized by
the results of plasmolytic studies. Akerman found that during ag-
gregation there is an increase in turgor pressure of about 5 atm. in the
cells involved. De Vries, it is true, found no changes while G.^rdener
beheved that they " lose their turgidity" (he made no experiments to
show this). As Coelingh points out, the difference in the use of the
terms turgidity and turgor pressure (dependent on the concentration of
the sap) is to be noted. In the case of cells in which precipitation oc-
curred, there is no change.

From experiments to determine if there is an influence of the gland
on aggregation, Akerm.a.n found that in pieces of tentacle stalk from
which (a) the gland with one-third and (b) with two-thirds of the stalk
removed, no aggregation at all could be procured in the remaining
portion of the tentacle. If only the gland and one-third of the stalk
had been removed, a weak or no aggregation occurred, according to the
test substance used. Akerman tried eight agents.

That injury is not involved is shown by the fact that aggregation
intervenes in small pieces of the upper region of the tentacle when pep-
sin or meat extract are applied.

These results in Akerman's opinion pointed to the presence of a
substance, resident or formed in the gland and in the more apical stalk
cells, which can on suitable stimulation procure aggregation. It looks
therefore, as if two substances are required to cause aggregation, one
inherent and one which must be supplied from the outside as a stimu-
lant. As Darwin recorded, however, aggregation occurs when the
tentacles are stimulated by mechanical means only. If a second sub-
stance is required, this must mean that the second substance is also
formed in the gland when so stimulated {see beyond).

Miss Coelingh, working at Utrecht, where the study of growth
substances was then being actively prosecuted, repeated some of Aker-
man's work, supporting his conclusion that during the progress of ag-
gregation there is an increase in the osmotic value of the sap, but that
this also takes place when the tentacles are successfully stimulated by
mechanical means. Having also verified Akerman's observation on the
effect of the gland on aggregation in the stalk (finding shght but
not significant differences) Coelingh proceeded to test the value of
his theory of an aggregation-promoting substance. To this end she
made extracts of glands in distilled water. Such extracts may be
preserved dry, and can then withstand heating to 100° C. In
order to determine the virtue of gland-extract, Coelingh proceeded
to obtain "empty" pieces of tentacle stalk, that is, such as would
not show aggregation when a "stimulating" agent (beef extract) was
apphed. Such were obtained of four kinds: (/) small strips of leaf

Francis E. Lloyd — 150 — Carnivorous Plants

blade with mere stumps of tentacles, (2) such pieces, or short pieces
of tentacle stalk placed for some time in a slightly hypotonic solu-
tion of cane sugar, remaining a few days to allow the putative sub-
stance to escape by diffusion; (j) young, still unfolded leaf tentacles,
the glands being removed and {4) a young but fully unfolded leaf
plunged into ethyl alcohol (96%) for 4-6 seconds. The alcohol be-
cause of the relative penetrabihty of the cuticles could not penetrate
into the leaf beyond the glands so that living tentacles with dead
glands were provided. Various difficulties involved need not be re-
cited here.

On testing the responses of thus prepared material to a " stimulat-
ing" substance in the presence and absence of gland-extract, it was
found that, with few exceptions, aggregation did not occur in the
presence of either alone, but only when both were presented together.
As control the author used extracts of Drosera leafstalks, with negative
results.

CoELiNGH also suspected that the aggregation-promoting substance
may occur also in the capital cells of other glandular structures of
Drosera since by the use of a stimulant (such as pepsin) she found it
possible to get aggregation in stalk cells of sessile and other glandular
trichomes and in leaf blade cells, even on the lower surface opposite
the bases of tentacles. On the other hand she could never obtain ag-
gregation in other parts of the plant where tentacles do not occur:
petals, sepals, ovary-wall, adventitious roots, young stipules.

It was found also that there are substances (saliva, diastase, tryp-
sin) which act as the theoretical aggregation-promoting substance,
and in a search for a possible clue as to its nature, some such sub-
stances were tested on '^ empty" tissues. Saliva added to pepsin acts
positively, even after heating to exclude enzymes, but saliva also acts
alone, as I have found. The composition of saHva being known, the
various components were tried and only the phosphates were active.
For example a 0.1% solution of Na2HP04 was active and sahva di-
luted to this concentration of that salt was also. Aggregation follows
on the use of pepsin plus growth-substance of Indian corn on "empty"
pieces of tentacle, while pepsin or meat-extract alone have no effect.
Among organic N-substances, aspartic acid, asparagin and leucine have
the greatest action; creatin, alanine and urea are doubtful or negative;
guanine and ethylurethane no effect whatever. Substances which
lower surface tension (saponin, amylalcohol) have no effect. The
presence of oxygen, as Darwin had found, is necessary. Since the
swelling of protoplasm is considered characteristic of aggregation, per-
haps the chief one (Schimper), it was thought that the pB. of the ag-
gregating agents would betray an influence, but on experimentation
only negative results were obtained.

CoELiNGH in a discussion of all the facts observed fully agrees with
Akerman, that two substances are needed to procure aggregation:
i) one with property A, which does not cause aggregation, but which
conditions cells to aggregate; and 2) one with B, which provides a
stimulus (pepsin, etc.) to aggregation. But it may not be assumed that
all substances have only one of these properties A and B but not both,
for some may have both in various degrees of efficiency. This is indi-

Chapter X —161— Drosera

cated in the behavior of various substances. It seems sure, however,
that some substances have only the property B, such as meat extract,
pepsin, peptone Witte, none of which can procure aggregation in empty
cells, that is cells devoid of A. But substances having A only seem
not to occur. On the other hand some substances can cause aggre-
gation in both normal and ''empty" cells, e.g., saliva, gland extract,
etc., and these, therefore, seem to have both properties A and B.
If this were true, we would be helped in understanding why aggrega-
tion can be procured by merely mechanical stimulation by assuming
that a substance containing AB is secreted only on stimulation, then
to diffuse into the stalk cells. To this it can be objected that the
cells of tentacle stalks with the glands removed can respond to pepsin
alone, indicating that B is already present in the stalk cells.

But more, for it appears that tentacles are to be found in a state
of aggregation when they have never been stimulated in any way so
far as this can be ruled out by conditions of culture, as Homes
subsequently showed. The hypothetical substance we are looking
for must therefore inhere and have both properties A and B. Since, if
this is to be assumed to be the case, aggregation may not be present
till stimulation occurs, it must be argued that stimulation merely ac-
tivates B.

Thus Miss CoELiNGH brings her argument into a purely theoretical
atmosphere which she, herself, finds hard to breathe, and says that
speculation without further experimentation affords no sure guidance.
But it must be evident that the problem of aggregation is most in-
triguing.

To the previously cited results of Schimper, Gardiner, de Vries,
Akerman, in which they clearly distinguished between true aggregation
and Goebel's granulation, Erna Janson took a diametrically opposite
position, stating that all aggregation is due to the precipitation of mate-
rials in solution in the cell sap. Neither the tonoplast (de Vries) nor the
swelHng of the protoplasm have any part in the process. It does not
appear in her paper, however, that she has taken sufficient cognisance of
that condition called by de Vries " true " aggregation. Her figures, which
are very crude, give no hint that she brought this into her field of con-
sideration. She showed, what Bokorny also had shown at length,
that certain reagents (alkaloids, other weak bases) procure precipita-
tions in the vacuoles. Beyond this she failed to show that "true
aggregation" could thus be achieved. Nor did Bokorny's figure of
aggregation in Drosera show more. Considered as contributions to
the nature of the vacuolar contents, these papers have value. As
furnishing enlightenment on the nature of that kind of aggregation
which follows on mechanical stimulation and on feeding with pepsin,
peptone, etc. (though in these was suspected the presence of ammonia),
the work of Janson is very limited. For it is quite certain, and I
speak now from my own observations, that during true aggregation
there is no slightest evidence of precipitation from beginning to end.

Following the studies of Darwin and Gardiner on aggregation a
new phase of the subject was entered upon, in which the cytological
changes taking place within the cells of the gland especially were ex-
amined. Lily Huie examined the behavior of the nucleus during se-

Francis E. Lloyd —152— Carnivorous Plants

cretion, digestion and absorption (1897-9). Much later Kruck (1931)
did the same for the gland cells within the traps of Utricularia (to be
mentioned elsewhere).

QuiNTANiLHA (1926) (on DrosopJiyllum), Dufrenoy and Homes ex-
amined the cells as a whole with reference to the comportment of the
vacuome. Homes has carried on most meticulous studies. He set
himself the task of determining first of all what happens during the
development of the cell during the ontogeny of the leaf and gland.
The next step was to determine if it is possible to describe definitely
the conditions of the glandular cells during repose, so far as this con-
dition might be reaHzed. To this end he depended not on plants col-
lected in the field, but on those raised under control, and which there-
fore, though secreting, were known not to have been stimulated by
insects, etc., but only by light, humidity, temperature. He then
studied glands during " digestion", that is glands which were actively
secreting enzymes and absorbing.

The development of the vacuome during the ontogeny of the vari-
ous tissues, including the glands, follows a general course beginning
with minute droplets of material in solution (primordia or metachro-
mata). These enlarge, remaining spherical or more or less elongating,
and then give rise to rods or fine filaments which are straight or curved,
simple or branched, giving rise to thicker, short, large and massive
rods, turning gradually into vacuoles of very irregular shape. From
these by confluence arises a single massive vacuole with concentrated
contents from which may arise a vacuome with a fine, becoming a
coarser, even large and massive network from which in turn a defini-
tive single vacuole with " diluted contents" arises. Homes then re-
marks that these observed stages are not strictly common to all tissues,
e.g., parenchyma cells do not pass through a network stage, which
stage in any event is very transitory, and marks tissues or cells which
are very active, or are at a moment of particular cell activity. There
are, he continues, the following two essential phases: growth by the
augmentation of vacuolar substance, followed by growth by simple hy-
dration. All tissues commence with a vacuome in the form of meta-
chromata, and finish (certain tissues, including the glands, apart) with a
single vacuole with dilute contents. The more, however, a region
preserves the capability of specialization, the slower the evolution of
the vacuole. The glands, therefore, remain in a relatively juvenile
condition, and consequently more capable of immediate activity.

In his second paper Homes followed the behavior of glandular cells
in order to find out what happens in them during the secretion of mu-
cilage only, so as to be able to distinguish later between this activity
and that which takes place during digestion and absorption.

He found first of all that under natural conditions the gland cells
presented no uniform structural condition. But since in nature, that
is in plants studied in situ, it is impossible to know the precise history
of a given leaf, since it may have been digesting and recovered. Homes
raised plants from winter buds under control, regulating within ap-
preciable hmits the amount of illumination and humidity. Much to
his surprise, the glands of such plants were as Httle uniform in cytolog-
ical features as those in situ. That is, instead of finding a uniform

Chapter X —153— Drosera

condition in the gland cells supposedly in a state of repose, he found
all possible conditions. It should be here noted that various authors
had previously described the resting condition in various ways. Darwin,
ScHiMPER, DE Vrees, Bokorny, Goebel, Akerman thought that in
this condition the gland cells have a single large vacuole. Gardiner
described the gland cell as consisting of a fine meshwork of protoplasm
holding red sap in the interstices. Dufrenoy held a similar view, be-
lieving that there are many small vacuoles. Quentanilha found a
meshwork of threadlike vacuoles. Allowing for discrepancies of under-
standing it is obvious that Homes' findings bring these various views
into some harmony, since no one condition prevails in the resting gland
cells.

Having compared living and fixed material it was established to
Homes' satisfaction that his fixative (osmic acid, 2%, i part; mercuric
chloride 2.5%, 4 parts) preserved the cell structure accurately. Know-
ing also that the glands in the middle of the leaf are more active in
secretion than those standing along the limb (since the inhibiting
action of the environment on the former is less effective), he made a
statistical study of the gland cells as to the form of the vacuome. But
a statistical study called for standard conditions. These were supplied
by growing plants under three sets of conditions, inciting minimum,
medium and maximum secretory activity. From these plants Homes
obtained evidence of the behavior of the gland cells under these various
conditions, and found that, though all states of activity occurred in all
cases, there was a preponderance of one state over the others in any
one experimental set of conditions. Relying on the relative abundance
of the different structural states of activity, four '' witness" glands
were chosen to serve as criteria, called types A, B, C, and D. Briefly
stated, and neglecting details, the outer glandular cells of these types
have vacuomes of the following character:

T5T)e A. the vacuome consists of a single large vacuole.

Type B. — — — — numerous small droplets.

T3T)e C. — — — — fewer, larger droplets.

Type D. — — — — a thicker reticulum.

It was then found that glands at minimum, medium and maximum
secretory activity displayed the above cytological characters in the
following percentages:

Glands of

A

B

C

D

Variants

Minimum activity
Medium "
Maximum "

97%
0
0

0

30%
15%

0

65%

25%

0
0

60%

3%
5%
0

showing clearly that when glands are at minimum activity, the
glandular cells are in the state in which they have a single large vacu-
ole, while when in maximum activity, the vacuome is a thick reticulum.
Intermediate conditions characterize glands of approximately medium
activity. In the second layer there is a similar course of events, but
these are not so pronounced, and do not follow the changes of the outer
layer promptly, scarcely ever doing more than fragmenting the
vacuole into two or three parts. The cells of the parenchyme bell (the

Francis E. Lloyd — 164 — Carnivorous Plants

third layer) never show these changes. There would, therefore, seem
to be a difference of function of these layers, the innermost taking no
part in secretion. Under severe conditions caUing for active secretion,
the outer layer may display a certain degree of plasmolysis, the cells
recovering by drawing water from the second layer. Homes regards
this plasmolysis as a normal event, and attributes it to the concentra-
tion of the mucilage by rapid evaporation so that it becomes hyper-
tonic to the cell-sap. Though normal, it occurs infrequently in plants
under usual conditions. Such plasmolysis can cause directly the frag-
mentation of the vacuole, as shown by plasmolysis studies in general.

Summarizing, it seems certain that the rate of secretion obeys
changes in the external conditions, quite apart from any responses to
irritability to chemical substances which are presented at and during
digestion. The changes described as aggregation under which these
can be subsumed are therefore not alone the result of stimulation in
the usual sense (mechanical, chemical).

The purpose of Homes' final paper was to determine if during di-
gestion there is a specific activity of the glandular cells different from
that during secretion of mucilage. The same methods were used as
previously and the various behaviors of gland cells fed with raw egg
albumin compared with those from leaves without nourishment. We
recall that during secretion, according to its intensity, any one of four
types of condition above mentioned may be found, these types being
a) cells with one large vacuole, h) cells with many small round drop-
lets, c) cells with irregular droplets, and d) cells with a thick reticulum.

"If now the glands are fed, at whatever state they may be at the moment, a first
rapid change will occur in the direction from A to D, or from B to D, but if they are al-
ready in the state D, no change will be apparent. Later, after the reticulimi stage has
been reached, a gradual concentration of the vacuole takes place till a single concentrated
vacuole is present in each glandular cell. Then this single vacuole becomes more and
more hydrated and finally a large diluted vacuole is present and persists till the end of the
digestion. Practically any type of vacuole present during digestion can thus be found in
a "resting" tentacle, and it is consequently impossible, by examining a single tentacle, to
tell if it is, was or is going to be in the process of digestion.

"But the vacuolar changes revealed by the statistical study can also be related to the
intensity of the secretory process or even more generally to the exchange of water between
the gland and the external medium. The first change towards the reticulum expresses
the increase of secretion which takes place immediately after feeding; the later hydration
of the single vacuole indicates the decrease of secretion and the beginning of absorption.

"Aggregation is thus not the characteristic result of 'excitation' in a carnivorous plant,
but simply the expression of any rapid change in the water content, as can happen during
secretion or absorption of a liquid." (Homes in ep.).

Thus stands the problem at the present. We are impressed with
what appears to be a relatively simple cytological behavior of the
cytoplasm (not speaking for the moment of the nucleus) and by the phys-
iological complexities indicated by the behavior of the glands and
tentacles during the process of digestion during which there must be
secretion and escape of enzyme (even though they may be al-
ready present in the mucilage), acid or acids, a substance inhibiting
bacterial action (Goebel), perhaps an odorous principle, and a more
active secretion of mucilage which ceases at the end of digestion and
absorption. Meanwhile, absorption takes place. Simple as the cyto-
logical behavior may be, it is, I feel, not yet thoroughly understood.

Chapter X — 155 — Drosera

From studies during the past three summers I can confirm Homes in
his claim that resting glands, that is glands found on young vigorous
resting leaves, can be found in a resting condition, meaning with cells
having a single grand vacuole, or with its cells in a condition which is
distinctly otherwise, that is, with an appearance which may be in-
terpreted as a mass of smaller irregular vacuoles, or as a network, prob-
ably fairly represented by Homes' fig. 60-63, 1932- Guided by what
we have seen in the tentacle cells, in which it seems reasonably certain
that the vermiform condition can give rise by confluence to a network,
temporary though it may be, we may agree that the same occurs in
the gland cells, as Homes believes, and he may be quite right.

I have seen evidence, however, that, during the period after feed-
ing (with raw and purified egg albumin, pepsin, peptone Witte,
saliva) while aggregation follows in the tentacle cells, meaning specifi-
cally the breaking up of the grand vacuole into smaller ones, these be-
coming numerous, slender, actively agitated vermiform bodies of high
refringence (as so well depicted by Akerman), to suffer at length con-
fluence and total reversion, this series of changes is not followed in the
cells of both glandular courses, though some approach to it may be
observed in the lateral epidermal cells seen en face, and in the apical
cells of a gland which were distinctly not in the grand vacuolar state,
having instead a number of smaller irregular vacuoles, in appearance
at any rate.

Lateral cells viewed en face present a crenated outline in conformity
with the buttresses which cut up the periphery of the cell into bays,
seen by Franca in Drosophyllum. I observed that in these cells, in
saliva, a droplet containing pigment would be formed in each bay,
there being as many droplets as bays. Dufrenoy's drawings indicate
that he saw the same condition, which he called aggregation, but ap-
parently did not connect the early form of the vacuole with the cre-
nated walls. These droplets formed in the bays might remain as such
or might run together to form a single drop containing all the pigment
of the cell, depending apparently on the size of the droplets and the
vigor of the process. In the apical cells, also in saliva, the whole mass
of vacuoles, whatever may have been their exact state, became con-
fluent and there was formed a single large drop, corresponding to
Homes' condensed vacuole. Here then we have a case which appears
to conform with Homes' observations. On the other hand, when the
gland cells are, to begin with, in the dilute vacuolar state, each having a
single grand vacuole (Homes' "vacuole diluee") when treated with
egg albumin, raw or purified, the first sign of response is to be seen in
ten minutes in the cells of the internal course in the lateral region of
the gland. The change advances toward its apex. In 90 minutes simi-
lar droplets appear in the epidermis lateral cells low down near the
base of the gland, again advancing toward the apex. These droplets
do not apparently appear in the grand vacuole but in the cytoplasm,
but this is a point very difficult to make out. In the course of time
the confluence of droplets yields a single large drop. This seems to be
what Darwin saw: " In 15 min. I distinctly saw extremely minute
spheres of protoplasm aggregating themselves in the purple fluid; these
rapidly increased in size, both within the cells of glands and of the

Francis E. Lloyd — 156 — Carnivorous Plants

upper ends of the pedicels." In my preparations aggregation, in the
generally understood sense, occurred in the upper end of the tentacle
stalk. Similar appearances were seen in glands treated with pepsin,
and peptone Witte (each i% soln.), in that the first thing noted is the
appearance of drops and there is no breaking up of the grand vacuole
into parts, as is to be constantly observed in the tentacles. In KH2-
PO4 (tried by Coelingh as a component of sahva) in i % soln. the
gland cells behaved so far differently that its action can be only doubt-
fully compared with that of saliva since vacuoles without pigment are
formed which push aside the cell contents, producing a distinctly path-
ological effect. Aggregation occurred in the tentacles, but this also oc-
curs in water. In order to check on my observation, I took a piece of
leaf which had lain in oil for 48 hours, and which showed very clearly
that there was no aggregation at all in any gland cells. After removing
the oil, which does not adhere to the glands because of the mucilage, I
treated it with saliva. In 15 min. droplets had appeared in the lateral
cells of the inner course of gland cells (C-II), the epidermis (C-I) re-
maining quite clear. At the end of 50 min. all the cells of C-II had each
a large drop, appearing in the apical cells last. All the lateral cells of
C-I had drops in them, a few small droplets in some apical cells. In
an hour's time, droplets had appeared in many of the apical cells of
C-I and they were evidently enlarging.

These observations seem to indicate that during the ordinary course
of events when the glands are secreting mucilage the condition of the
gland cells may be found in various states such as Homes had de-
scribed. If the glands are fed they may follow one of two courses.
If the gland cells are filled with smaller vacuoles (are aggregated), these
become confluent to form a single large drop, the condensed vacuole of
Homes. If the gland cells are in the resting condition, that is, are in
the grand vacuolar stage with no sign of aggregation, the course of
events on feeding consists in the formation of small droplets in the
cells of the inner glandular course in the lateral regions of the gland.
This advances till the apical cells of this course have all formed drop-
lets. These grow until by their size they form an optically dense layer.
In the meantime, droplets have appeared in the epidermis, in the
lateral region. This also progresses toward the apex, until all the epi-
dermal cells are involved. The drops all contain pigment. In the
course of change, the gland becomes denser and darker in color as Dar-
win observed. During six hours aggregation had occurred in the tenta-
cles and by next morning it had progressed quite to their bases and
into the leaf tissue about their bases.

Whatever the final agreement as to the course of aggregation, which
appears to be different in detail, if the same in results, in gland and
tentacle, the whole activity is most extraordinary, and demands much
further study before any final answer can be given as to the relation of
aggregation to secretion and absorption.

Studies of cytoplasm and nucleus. — The studies of the living cell
leading to our knowledge of aggregation led to a desire to know more
of those details of behavior of the cytoplasm and especially of the
nucleus which cannot be discovered by the methods used for obser-
vation of the li\'ing material. Accordingly the method of fixation fol-

Chapter X —157— Drosera

lowed by sectioning and staining came into use first by Gardiner,
followed later by Lily Huie, Rosenberg and Konopka. Homes'
work we have already mentioned as especially bearing on aggregation.
Gardiner believed that the mucilage of the glands is secreted as a
" formed matter" within vacuoles which grow after stimulation, the
protoplasm being reduced in amount to be later restored by new
growth from the vicinity of the nucleus. He observed the presence of
the crystalloid " rhabdoids". It is not unfair to say that his work was
rather meagre in amount, and he did not use staining reactions. Huie
on the other hand did a sustained piece of work. She traced rhythmic
changes in cytoplasm, nucleus and nucleolus, attempting to interpret
these as chemical and morphological reactions connected with and
following stimulation by food materials of various kinds, which, ac-
cording to their nature, were followed by quantitatively various re-
actions, though quahtatively similar. We may leave out of account
the reactions to non-organic substances which produced in any event
only very fleeting cytological changes. Following feeding there is a
reduction in the volume of the basophile cytoplasm until it becomes
scanty in amount and eosinophile in character confirming Gardiner's
similar observation. Following this but preceding the restoration of the
cytoplasm there is an increase in the volume of the basophile chromo-
somes accompanied by a reduced amount of nucleolar chromatin. The
beginning of restoration of the cytoplasm is to be seen in an accumula-
tion of neutrophile dense cytoplasm surrounding the nucleus correspond-
ing chemically (as indicated by color reaction) and morphologically
(size of granules) to the intranuclear plasm. At completion of
cytoplasmic restoration the nuclear chromatin is reduced in amount,
and the nucleolar chromatin increases. What is left of the chromoso-
mal bodies (chromatin) is finally aggregated into definite V-shaped
bodies of a constant number (eight) characteristic of the plant, which
" proves" that this is a mark of nuclear activity and not merely a
feature of mitosis. Huie's second paper (1899) adds nothing to the
above account in general, but she concludes that the nucleus is the
seat of metaboHc activity, and that the usefulness of a food can be
judged by the condition of the '' nuclear organs."

Rosenberg (1899) supported and extended Huie's studies. He
saw similar changes, but in lesser degree, in the endodermis, tracheids
and stalk cells. He diverged, however, from her interpretation of the
masses of chromatin as chromosomal, for he found them in no constant
number, nor did he find splitting as in prophase, but on the contrary
much difference in form and size. He observed, however, certain
bodies lying on or near the nuclear membrane, called generally "pseu-
donucleoli", which he termed prochromosomes. These occur in a con-
stant number, and are the chromosomes.

Konopka and Ziegenspeck (1929) studied D. rotimdifolia, D.
binata and D. anglica. After the glands have been fed 24 hours with
various proteins, including pollen, droplets appear in the cytoplasm
near the nucleus (intermediate food products) at first always in the
inner gland cells layer, later in the outer, and also in endodermis
(parenchyme bell) tracheids, stalk cells and leaf blade cells. They in-
crease in number and size, and at length are overtaken by a sort of

Francis E. Lloyd — 158 — Carnivorous Plants

disruption, show hollows and cracks, and appear to be in some vague
but perhaps intimate connection with the nucleus, indicating that they
are products taken up by it. At the " high point" of feeding the
nucleus has enlarged and the membrane becomes less definite and
finally disappears so that the chromosomes and nuclear materials ap-
pear to he free in the cytoplasm, or at any rate in a nuclear lymph.
KoNOPKA observed in the resting stage the bodies which Rosenberg
called prochromosomes, but could not confirm his belief that they oc-
cur in a constant number, regarding them rather to be of nucleolar
nature. After feeding the nucleoli become reduced in size, and lie in a
large vacuole (" Hof") from which, in many cases, he could observe
canals leading to the cytoplasm, bringing the nucleus into more inti-
mate contact with the cytoplasm. Here it is to be regretted that cost
prevented the reproduction of his photographs, since the figures are
unsatisfactory to a degree, and show no convincing evidence of this.
The chromatin on the other hand exhibits increase and occurs in larger
masses of various form. After 24 hours they become very evident by
their clearness and size, and besides granules of various sizes there ap-
pear rod-shaped structures (" Rhabdoids"?) lying near the periphery
(of the nucleus) sometimes paired and connected by fibers with the
interior of the nucleus, but in no constant number. The nucleolus has
now been dissolved, and appears as a pale, rather than, according to
Rosenberg, a distinct body. The chromatin rods now unite to form
threads and rings. Among them the granules, which Rosenberg
thought to be prochromosomes, are secondary nucleoh, indicating
enhanced nucleolar activity. They always he in vacuoles (" Hof")
and have no connection with the nuclear structure, show disinte-
gration and disappearance at the high point of the reaction. During
long periods of digestion these events appear not to progress steadily,
but rather to pulsate — there is a rhythm in behavior. These nucleolar
structures are regarded by Konopka as supplying materials for forming
mucilage and for digestion, and he is inchned to regard the central and
peripheral nucleoh as having different functions. The whole aspect of
the changes in the chromatin and nucleolus indicates that these changes
are connected with ferment production. With the escape of the fer-
ment the chromatin shrinks in ring forms, so that one might regard the
nucleus now as being in a spireme stage, leading to mitosis, or merely
as a reformation of chromosomes in a somatic condition. They are at
all events true chromosomes.

It has been attempted to summarize the above work with not
too great brevity, so that the reader may appreciate the difficul-
ties of interpretation. It is not too much to say that, while it has
been shown experimentally that changes in the cell do indeed occur
during digestion, and while we have become aware to some extent
what, in detail, these changes are, it must still be recognized that we
are yet lacking general agreement as to the precise nature of many of
these details, and much less are in a position to attribute precise func-
tions to the various structures seen.

Digestion; enzymes. — Darwin (1875) proclaimed the digestive
power of the secretion of Drosera tentacles. He fed the leaves proteins,
connective tissue, cartilage, gelatin, to find that these were attacked.

Chapter X —169— Drosera

The presence of acid being a condition for peptic digestion, he observed
that the inner tentacles of the disc of the leaf were more acid than the
outer. This may have been because of the greater number of tentacles
per unit of area. Darwin thought that the acidity of the secretion of
the leaf is increased on the absorption of nitrogenous substances de-
rived from the captured insect.

Opposed to the general trend of opinion was that of Morren
(1875), Batalin (1877), TiscHUTKiN (1889), and of Dubois (1899), all
of whom were persuaded that the digestion of insects by Drosera is
always the result of bacterial action, so that the results of others, to
be detailed below, were not without opposition.

Rees and Will (1875) made a glycerin extract of the leaves and
found it weakly acid and to contain an enzyme which in the presence
of weak HCl exercised a peptonizing action. Lawson Tate (1875)
collected the secretion from the tentacles, sweeping the leaves with a
feather (he used Drosera dichotoma), mixed it with water and pre-
cipitated it with cholesterin. The precipitate was found to coagulate
milk, and this he referred to the action of a ferment which he named
droserin. In 191 1 Miss J. White reinvestigated the matter; leaves
were removed, washed with previously boiled water with added chloro-
form and chopped with a sterile knife. The bits were then placed in
lukewarm boiled water with chloroform as antiseptic, shaken vigor-
ously for 2 hours. To the filtrate was added an equal part of satu-
rated ammonium sulfate, from which a filtrate was obtained which
contained a principle which could attack fibrin but only in an acid
medium. The product gave the biuret reactions. Abderhalden
(1906) had found that in its presence peptides are not spHt.

Dernby in 191 7 obtained a glycerin extract of the leaves from
which by means of dialysis he obtained an enzyme which he regarded
as a pepsin. This worked at an acidity of pR 5 as optimum. No
tryptase or ereptase was found.

Miss Robinson (1909) tested the digestive effect of Drosera on
" purer proteins" then available. She found that acid-albumin, alkali
albuminate and edestin were digested, but " somewhat less readily"
than dry egg-white, fibrin, tendo-mucoid and nucleoproteins. Col-
lagen and elastin proved entirely indigestible. Though creatin did not
cause a bending of the tentacles, it was readily dissolved, meanwhile
remaining in contact with the leaf for three days. It is important, in
view of Darwin's opposed idea, that the lack of movement of the
tentacles is not an indication of the non-nutritional value of the sub-
stances applied; nor did Darwin find that the positive response indi-
cates the contrary.

Beginning in 1930 Okahara pubHshed a series of papers deahng
with the matter. He first dealt with the question of the actual oc-
currence of a digestive ferment in Drosera. The leaves were extracted
with glycerine and water for several days (with toluene) and the press
juice then filtered off. The mother solution thus obtained showed
strong acidity. The enzyme was separated by means of acetone and
redissolved for experimentation.

He concluded that there exists in the leaves of Drosera a powerful
proteolytic enzyme which, acting on proteins, hydrolyses them to

Francis E. Lloyd — 160 — Carnivorous Plants

proteoses and peptones, with an optimum activity at pB. 1.5, an acidity
high for plant enzymes and suggesting a resemblance to animal pepsin.
In order to determine to what extent the Drosera pepsin is identical in
action with animal pepsin, Okahara observed the influence of poisons
(quinine hydrochloride and atoxyl) on them. He failed to find a strict
parallel, since the enzyme activity under certain conditions was re-
pressed in the one and accelerated in the other.

Okahara's second paper (1930J) is of a more general nature, deal-
ing with the effect of toxic substances on pepsin with a view to
illuminating his earher observations cited just above. Though the sub-
ject may be regarded as controversial, it remains true that there is a
substance capable of digestion of proteins in Drosera leaves.

In a third paper (1931) Okahara gave the results of inquiry into
the optimum acidities of various acids for the enzyme activity. He
had observed the occurrence of formic acid in Drosera, and was
prompted to investigate the comparative effect of various acids on the
action of a proteolytic enzyme on edestin and found that the optimum
acidities for various acids differ, and that the decrease from the op-
timum acidity parallels the decrease of the electrical dissociation con-
stant.

While Okahara's second and third papers do not immediately
concern Drosera, they have been mentioned in this connection since
their bearing will doubtless be made clear by further studies. His
fourth paper, however, bears directly on the controversial question,
do bacteria play a role in the digestion of carnivorous plants and in
particular of Drosera? Nepenthes was examined also in this connection.
The paper was published in 1933. The author isolated from the plants
studied a series of bacteria and moulds. Experiments with these
in media held at two acidities, />H 5-6 and /?H 3.3, afforded the following
results. Three of the moulds acted on Witte's peptone and glycocoll
at ^H 3.3. The other organisms attacked various nitrogenous com-
pounds supplied (Witte's peptone, glycylglycine, glycocoll and alanine)
falling into two groups which cooperate to reduce these substances to
ammonia. Okahara concludes, that, while the plant enzymes may
themselves take a leading part in the breaking down of proteins,
such organisms as were isolated from the plants mentioned ''may
also cooperate in the completion of the process." Okahara to this
extent supports the views of Labbe (1904) and of Stutzer (1926),
the former for Drosera and the latter for Utricularia.

Following Okahara, Linderstr0m-Lang and Holter (1934) again
raised the question whether digestion in Drosera rotundifolia is es-
sentially different from that in other plants and similar to that in
animals, or do they depend on resorption of the products of bacterial
action?

Accordingly, the secretions from the glands and from the leaf
tissues (from the blade, that is) were examined separately, in order
to answer specific questions, to wit: (i) whether proteinase is secreted
by the glands; (2) what position among the proteolytic enzymes
it takes; (3) in what quantities it occurs and how these quantities
behave in relation to the endoproteinase to be expected in the leaf
tissues.

Chapter X —161— Drosera

Their method of obtaining the enzymes was as follows. The se-
cretion of the glands was taken up by filter paper and because of its
viscosity was diluted. That of the leaf tissues was extracted with
glycerin. Both were tested on edestin.

It was found that the gland secretion, with its optimuni at pB.
^.^, was far more active than that of the leaf blade, with its opti-
mum at 4.6, or even of the secretion, extracted from removed glands
with maximum activity at ^H 3.8. It is admitted that the last may
be due to the overlapping of the action of the two enzymes in ques-
tion. The authors concluded, "We have to do with a quite different
distribution of two enzymes, of which the one occurring in the se-
cretion is a proteinase for the purpose of digestion." Further it was
pointed out that the distinct function of a proteinase, optimum activity
at pB. 3.2, does not harmonize with Dernby's results who found
the maximum activity on acid albumin at ^H 5, nor with Okahara's
with carmine fibrin, maximum activity at pE. 1.4. Dernby's results
may have come about because he used masses of total leaf, but those
of Okahara's are regarded as distinctly antagonistic. This may be
due to the possibiUty that the Japanese plant may differ physiolog-
ically from the European. Merck's pepsin acted on edestin at pB.
1.8; therefore, the proteinase of Drosera and pepsin are not iden-
tical. , . o ,

Recent and still unpublished work done by A. Akerman and
L. G. M. Baas Becking using D. capensis yielded definite evidence
that peptic fermentation takes place. The method used was the fol-
lowing: A single tentacle on the leaf edge was plunged into distilled
water held in a small paraffin cup. Under these conditions the water
retained its initial pB. of 5.8 for at least 24 hours. When, however,
a solution of NH4CI (cone. 25 mgr/L) was used the ^H fell from 5.8
to 2.0, from which it is evident that this salt served to stimulate the
production of acid. Equivalent solutions of CaCl2, NaCl and MgClz
gave no action, while KCl produced only a very slight change in
pB. When egg albumin {pB 7.0) was placed on a tentacle, the pB
changed to 3.0. Carmine-fibrin when treated with a leaf extract was
digested indicating the presence of a peptic enzyme, effective at an
optimum pB of 2-3.0 while it has been shown by J. de Zeeuw that
digestion in Nepenthes takes place at about pB 4.0. Since the leaf
extract was not bacteria free. Prof. Baas Becking (Sept. 1935) pointed
out that the proof for Drosera is not absolute, and final proof will
require experiments with bacteria-free plants. It is further noted
that Drosera proteinase takes a middle position between pepsin and
papain {in ep.).

Darwin observed that milk when placed on Drosera leaves was
soon coagulated. Green mentions this under the heading "Vegetable
Rennet," presumably because of the more obvious inference that milk
coagulation is brought about by a rennet, as perhaps is the case
when Galium verum is used for the preparation of curds for cheese
making, also mentioned by Green. As has been seen, a similar ac-
tion of Pinguicula in coagulating milk is not attributed to the pres-
ence of a rennet (Dernby 191 7), and if so, this may be equally true
of Drosera. Darwin does not speak of a rennet, but does remark

Francis E. Lloyd

— 162 — Carnivorous Plants

on the digestive effect of the secretion on casein, which harmonizes
with Dernby's view that a trypsin is present (in Pinguicula).

The general conclusion may be drawn from the foregoing summaries
of work done on digestion in Drosera that this plant does indeed
secrete a ferment which can act upon proteins and reduce them to
substances which can be and are absorbed for nutriment. If food
materials in the form of an abundance of insects, pollen, etc. (Darwin)
are present so that the antiseptic effect is incomplete, bacteria may
(particularly according to Okahara) assist in rendering such foods

available to the plant.

The abundance of fats in the bodies of insects would suggest
the presence of a lipase in Drosera, but such has not been found.
Whether lecithin and fatty acids might be absorbed by infiltration
(as lecithin is taken in by the human intestine according to Shocot-
YOFF, fide ScHMiD, 1912) is a matter of speculation, though Goebel
thought that he found fats to be absorbed by the glands of Pingui-
cula and Utricularia.

Significance of carnivory for the plant. ~ Although. Darwin left
no room for doubt that Drosera is able to catch, digest and absorb
the products of digestion, it remained a question if this abihty is
of advantage to the plant in furthering its growth and development.
It was natural that Francis Darwin (1878) should take up the cud-
gels in his father's behalf. He grew plants, obtained from the field,
in shallow dishes duly protected so as to prevent insects from reaching
them. These he divided into two lots, one of which he fed, the other
remaining unfed. The result showed the indubitable conclusion that
the plants which were fed were more vigorous, produced more and
stronger inflorescences and seed than the unfed. Similar results were
obtained by Kellermann and v. Raumer (1878). Busgen (1883)
then pointed out that plants grown from winter buds show a wide
range of development to begin with, so that an experiment with these
is really a handicap race. To avoid this he used plants grown from
seed, so that his plants started out from scratch. The results were
even more striking than those of the previous workers. In the table
herewith the results of the three authors are compared in terms of
percentage, the quantities for the unfed plants being 100: —

Fr. Darwin Kellermann & Busgen

1878 V. Raumer, 1878 1883

Number of inflorescences 164.9

Number of capsules i94-4

Total weight of seed 379-7

100 152: 100 300: 100

100 174: 100 533: 100

100 capsules 205: 100

And in terms of dry weight: Solution Spring

nutrient water

Winter buds, i Feb 173: 100

Winter buds, 3 Apr 213: 100

Entire plant, end of 2nd. year 296: 100 174: 100

Inflorescences 141. i : 100

Plants minus Inflorescences 12 1.5: 100

{flde Busgen)

While these figures speak for themselves, I venture to quote briefly
from these authors. Kellermann and von Raumer: "The general

Chapter X —163— Drosera

result is not to be doubted, that in all essential points the fed plants
forge ahead of the unfed"; Busgen: "We must therefore take it
as proven that animal stuflfs are transferred to the plant and that
they are of great significance to it for its development, namely, for
the development of its fruit, etc."; and finally Francis Darwin
said: "These results show clearly that insectivorous plants derive
great advantage from animal food."

Just previously to the publications of Kellermann and v.
Raumer's work Pfeffer (1877) had grown Drosera rotundifolia plants
from winter buds under cover to prevent the access of insects, and
observed that they grew vigorously, evidently leading to the conclu-
sion that the carnivory is not always a necessity. Regel (1879) went
further than this, claiming that the carnivorous habit is a distinct dis-
advantage because he observed that the leaves {Drosera filiformis)
were often injured by feeding, and that fed leaves die sooner than
unfed ones. It is, however, well known, as Goebel pointed out,
that overfeeding often causes decay of the leaf; and to deplore the
earlier passing of fed, not overfed, leaves is to ignore the possible
good which may have accrued to the plant in the meantime. And
Haberlandt was of the opinion, based on field observations in Java,
that Nepenthes pitchers appeared to have but a meagre booty, that
insectivory is a sort of semi-superfluous, luxus adaptation. In this
Massart {through Haberlandt), having had similar field experiences,
agreed. Nor did Goebel regard the role of insectivory in the struggle
for existence very seriously — it is useful, he said, but not obligatory,
and the plant does not meet much competition in its natural habitat.
Such more or less contrary views have in the long run been brought
to a focus in the idea now generally accepted that carnivory is a very
striking and useful adaptation, which, though not always obHgatory,
can under circumstances better the condition of the plant. Ad-
ditional questions, no less important, however, arose. It will be
noted that the above researches were overshadowed by the sole idea
of animal food, as supplying chiefly proteids, and this has crept into
the textbooks as the dominant thought. Stahl, in 1900, published
a long dissertation on the significance of mycorrhizal arrangements
in plants, in which he instituted comparisons between those plants
with the carnivorous plants, all of which grew in sterile soils. Sar-
racenia had been shown by MacDougal (1899) to be free of mycor-
rhiza, nor had it been found otherwise in Pinguicula (Schlicht, 1889,
through Stahl), Drosera, or Nepenthes (Janse, 1896, through Stahl)
and this is now known to be the case for all carnivores. As com-
pared with true parasites, mycorrhizal plants and autotrophic plants
with very extensive roots, those plants which avail themselves of capil-
lary water and in which many forms of animal life perish and are
entangled in the foHage {e.g. mosses), and carnivores have poor roots
and therefore httle means for obtaining the materials of the soil^ no-
tably lacking in salts, especially those of phosphorus and potassium,
in which they grow. And while it may be true that it may be shown
by experiment that carnivores may obtain all their requirements
through their roots, if plentifully supphed to the substrate, this does
not show that in a state of nature their arrangements for obtaining

Francis E, Lloyd —164— Carnivorous Plants

these materials are superfluous or useless, since they live in nature
and not under experimental conditions. Stahl indicated the low ash
content of the leaves of carnivores, and advanced this additional fact
as an argument for the significance of carnivory. There is generally
also a depression of transpiration due to situation in the habitat,
and where transpiration is low, some other means of obtaining salts
is called for. The leaves of Nepenthes, when exposed to situations
where transpiration can act freely, do not make pitchers, and have
a higher ash content than those low down and exposed to higher
humidity, where also, as in the case of seedlings (Goebel), pitchers
are immediately produced following the cotyledons. Stahl thus
argues: the carnivory has been dominated by the idea that it is
an adaptation to obtain proteins; but the soils in which carnivorous
plants grow are notoriously poor ones, and therefore the question of
how the carnivorous plants obtain substances aside from nitrogen is
in need of investigation. This in 1900. Pfeffer had indicated
this problem (1877), thinking particularly of phosphorus compounds,
and his and Stahl's suggestions were fruitful ones.*

In 191 2, Weyland and Schmid both entertained this idea, Wey-
LAND showing that there was Kttle of K and P to be found in the
meagre roots of Drosera, and Schmid finding these elements present
in the leaves of this plant after insect feeding, whereas before this
they were not to be found or only in meagre amounts (Ruschmann,
1914). Oosterhuis (1927) pushed investigation further along in this
direction. He asked the question whether the lack of any particular
mineral in the soil could be compensated for by insect feeding. He
summed up his experiments by saying that (i) mineral nutrients can
be taken up by the roots; (2) even if an abundant supply of nutrient
salts is present, the plants can not grow as well as if insect-fed; (3) in
view of the fact that in his experiments plants grown on a substrate
poor in salts but insect-fed prospered beyond plants grown on salt-
rich medium but not insect-fed, he argued that the significance of
insectivory Hes in the uptake of the cleavage-products of proteins
out of insects by the plant. The absorption of salts from the in-
sect is not excluded, but is not the important factor. Summarily
stated, in the lack of nitrogen in the soil, the plants can be supphed
this by insect prey, and then flourish better than when grown in
a substrate with Knop's solution supplying all elements. That plants fed
with insects have a higher nitrogen content than those grown ap seedlings
on turf carrying Knop's solution strengthened him in this view.

Behre (1929), stimulated by an expression of doubt by Diels
(1906) as to the value of insectivory, asserted that such value had
not been proved, and instituted experiments of his own to test the
matter. He found that plants grown in distilled water but plenti-
fully fed with flies or meat throve very much better than those grown
in distilled water, or even in a v/eak Knop solution (M cone), in
both cases not fed. The differences noted became much more evident
toward the end of the second summer. He concluded that insectivory
is indeed of great moment to the plant. An important value, it
seemed to Behre, hes in the taking up of inorganic salts, and that

* Peyronel (1932) argues that if mycorrhizal fungi are parasitic, they should be found
in carnivorous plants, but he found none in Drosera or Pinguicnla. Mycorrhiza occurs
chiefly when soils are poor in nitrates and ammoniacal salts, but rich in organic matter.

Chapter X —165— TtTosera.

the lack of such salts in the natural environment is compensated for

in this way.

Came Oudm.\n in 1936 with further proofs. The virtue of his
experiments lies in the fact that his experimental plants {Drosera
capensis) were grown from seed, and the seedlings carefully chosen
for their uniformity, and in the further fact that the plants were
grown on a very uniform substrate of powdered peat which had pre-
viously been very thoroughly washed. Several sets of plants with
{a) distilled water, (b) nutrient solution without N, and (c) Knop's
solution, were set up and either not fed at aU, or fed with asparagin
1.5%, peptone 1.5%, gelatin 2% (against dist. water alone), gelatin
plus Knop (against dist. water only in the substrate), Knop solution
alone, and finally with insects. He found that plants grown on salt-
poor substrate, but fed insects, were quite normal. Plants grown
on N-free substrate could make use of asparagin and peptone as well
as the N-compounds occurring in insects. Plants well supplied with
nutrient salts, incl. nitrogen compounds, can grow well in the absence of
leaf-feeding with insects. Drosera can obtain nitrogen if this is not pres-
ent in the substrate, through its leaves, and this in organic form.
It can also take up through its leaves not only N, but other salts
as well. Oudman's conclusions correspond quite fully with those of
OosTERHUis. There can, therefore, be no sort of doubt that the
ability to absorb substances (mineral salts as well as N) is of sig-
nificance to the plant. It should be added that gelatin and glutin,
a derivative of gelatin, cause degeneration of the tentacles, so that
in time they entirely disappear.

The presence of ascorbic acid in D. intermedia, suspected by Weber
(1938), was soon after demonstrated in the leaves of this plant by
Neubauer (1939) who claims to have found a content nearly as high as
that of a "well known paprika preparation", which itself contains
20-fold that of lemon juice. On this Weber (1940) again examined
the leaves of the same species after having been fed peptone powder,
and obtained evidence of a heightening of cell activity, accompanied by
an increase in vitamin-C content. This being a non-nitrogenous com-
pound, the significance of these results is quite problematical.

Literature Cited:

Akermam, a., Untersuchungen iiber die Aggregation in den Tentakeln von Drosera rotun-

difolia. Bot. Notiser 1917:145-192.
Ames, O., An easy method of propagating Drosera filiformis. Rhodora 1:172, 1899.
Arisz, W. H. & J. OuDMAN, On the influence of aggregation on the transport of asparagine

and caffeine in the tentacles of Drosera capensis. Proc. K. Akad. Amst. 40:3-11, 1937.
Batalin, a., Mechanik der Bewegungen der insektenfressenden Pflanzen. Flora 60:33-39;

54-58; 65-73; 105-111; 129-144; 145-154, 1877 {Drosera, 33-73; Diotiaea, 105-150;

Pinguicula, 150-154).
Beck, A. B., A. K. Macbeth and F. L. Winzor, The absorption spectra of hydroxynaph-

tlioquinones and of the coloring matter of Drosera Whittakeri. Austral. Jour. Exp. Biol.

Med. Sci. 12:203-212, 1934.
Behre, Karl, Physiologische und zytologische Untersuchungen liber Drosera. Diss. Ham-
burg, 1929. Planta 7:208-306, 1929.
Beijerinck, M. W., Beobachtungen und Betrachtungen uber Wurzelknospen und Neben-

wurzeln. Verzamelde Geschriften van Beijerinck 2:7-121 (1886).
Benecke, W. & L. JosT, Pflanzenphysiologie. 4. Aufl., Jena 1924.
Bennett, A. W., The absorptive glands of carnivorous plants. Mo. Microscopical Journ.

15:1-5, 1876 (n.v.).
BOKORNY, Th., Uber Aggregation. Jahrb. wiss. Bot. 20:427, 1889.

Francis E. Lloyd — 166 — Carnivorous Plants

BusGEN, M., Die Bedeutung des Insektenfanges fiir Drosera rotundifolia. Bot. Zeitung
41:569-577; 585-594, 1883.

DE Candolle, a. p., Physiologic Vegetale. Paris, 1832. (Translated by J. Roeper. Stutt-
gart and Tubingen, 1835) (vol. 2, p. 652 of this refers to Drosera).

CoELiNGH, W. M., Over stoffen, die invloed uitoefenen op de aggregatie bij Drosera.
Proefschrift Amsterdam, N. V. Hollandiadrukkerij, Baam, 1929, 74 pp., cf. also her
English resume in K. Ak. Wet. Amst. 32:973, 1929.

CORRENS, C, Zur Physiologie von Drosera rotundifolia. Bot. Ztg. 54:21-26, 1896.

CzAjA, A. Th., Insectivoren. Handworterbuch der Naturwissenschaften 5:655-666, 1934.

Darwin, Charles, Insectivorous Plants. 2nd. Ed. of 1875.

Darwin, Erasmus, The Botanic Garden. London 1791; New York 1798.

Darwin, F., The process of aggregation in the tentacles of Drosera rotundifolia. Q. Jour.
Mic. Sci. 16:309-319, 1876. ,

Darwin, F., Experiments on the nutrition of Drosera rotundifolia. J. Lmn. Soc. Bot. 17:
17-32, 1878.

Dernby, K. G., Notiz betreffend die proteolytischen Enzyme der Drosera rotundtfolta.
Bioch. Zeitschr. 78:197, 1917.

Deels, L., Droseraceae. Das Pflanzenreich, IV, 112, 1906.

DiELS, L., Blattrhizoiden in Drosera. Ber. D. Bot. Gesellsch. 24:189-191, 1906.

Dixon, H. H., Adventitious buds on Drosera rotundifolia. Notes Bot. Sch. Trinity Coll.
Dublin 144-145, 1901 {n.v.).

Drude, O., Die insectenfressenden Pflanzen. Schenk's Handbuch der Botanik 1:113-146,
1 88 1 (Literature complete up to 188 1). -r ,• a

Dubois, R., Absence de zymase digestive des albuminoides chez le Drosera longifoha. Ann.
Soc. Linn, de Lyon II, 45:79-80, 1898, 1899 {n.v.).

DUERENOY, J., Modifications cytologiques des cellules des polls de Drosera rotundifolta.
C. R. Soc. de Biol. 97:86-89, 1927.

Fenner, C. a., Beitrage zur Kenntnis der Anatomic, Entwickelungsgeschichte und Bio-
logic der Laubblatter und Drusen ciniger Insectivoren. Flora 93:335-434, i904-

Franca, {see under Drosophyllum). .

Frankland, Acids of the secretion of Drosera tentacles, in Darwin's Insectivorous Plants,

2d. ed., pp. 73-76, 1875- . . , . r .1. 1 J n • n

Gardiner, W., On the phenomena accompanying stimulation of the gland ceUs in Drosera

dichotoma. Proc. R. Soc. London, 39:229-234, 1886.
Geddes, {see under Pinguicnla). . . r. t^i

Giessler, a., Einfluss von Salzlosungen auf die Starkeverarbeitung bei Drosera. tlora

23:133-190, 1928.
Glauer, Jahresber. (1887) d. Schles. Gesellsch. f. vaterl. Cultur 3:167, 1886.
Goebel, K., Pflanzenbiologische Schilderungcn, II. Marburg 1891.
GoEBEL, K., Bull. Torrey Bot. Club 30:179-205, 1903.
Goebel, K., Brutknospen bei Drosera pygmaea und einigen Monokotylen. Flora 98:324-

335, 1908.
Goebel, K., Organographie der Pflanzen 3:1497, 1617, 1923, Erganzungsbd. 171, 212, 1924.
Gra\^s, J. A., Notes on Drosera. Plant World 1:28, 1897.

Green, J. R., {see under Nepenthes). , c • T\r

Gronland, J., Note sur les organes glanduleux du genre Drosera. Ann. d. bci. nat., IV

scr. Bot. 3:297-303, 1855.
Grout, A. J., Adventitious buds on Drosera rotundifolia. Am. JNat. 32:114, i»9».
Haberlandt, G., Eine botanische Tropenreise. 1893, S. 228.
Haberlandt, G., Physiologische Pflanzenanatomie. 6. Aufl., Leipzig 1924.
Hayne, F. G., Getreue Darstellung der Arzneigewachse. {through Nitschke, i860).
Heinricher, E., Zur Kenntnis von Drosera. Zeitschr. des Ferdinandeums f. Tirol 3, 46, 1902.
HoMis, M., Evolution du vacuome au cours de la differenciation des tissus chez Drosera

intermedia Hayne. Bull. CI. Sci. Acad. R. Bclg., 5 ser., 13:731-746, 1927-
Homes, M., Developpement des feuilles ct des tentacules chez Drosera intermedia Hayne.

Comportcment du vacuome. Bull. CI. Sci. Acad. R. Bclg., 5 ser., 14:70-88, 1928.
Homes, M., La question des plantes carnivores, principalement au point de vue cytologique.

Bull. Soc. R. Bot. Belg. 61:147-159, 1929a. ^ ^

Homes, M., Modifications cytologiques au cours du fonctionnement des organes secreteurs

chez Drosera, I. Modifications dans les feuilles non nourries. Mem. Acad. R. Belg.

CI. Sci., ser. 2, 10:1-54, 1929ft; II. Modifications dans les feuiUes nourries. Ibid.

Hooker, h! D., jr.. Physiological observations on Drosera rotundifolia. Bull. Torr. Bot.
Hooker, H. D., Jr., Mechanics of movement in Drosera rotundifolia. Bull. Torr. Bot. Club

Hooker, J. D.,' Address to the Dept. of Botany and Zoology, B. A. A. S. Belfast Meet-
ing, 1874. Report, pp. 102-116, 1875. . u f J-

HuiE, L. M., (a) Changes in the cell organs of Drosera rotundifolia produced by teeding
with egg-albumin. Q. Journ. Mic. Sci. 39:387-425, 1897.

Chapter X — 167 — Drosera

(b) Further studies of cytological changes produced in Drosera. Q. Journ. Mic. Sci.
42:203-222, 1899.

(c) Changes in the gland cells of Drosera produced by various food materials. .\nn.
Bot. 12:560-561, 1898.

Janse, 1896 {through Stahl igoo).

Janson, E., Studien iiber die Aggregationserscheinungen in den Tentakeln von Drosera.
(Diss. Munchen, extracted and repaged from Beih. bot. Centralbl. 37, 1920, zi PP-)-

JosT, L., see Benecke-Jost.

Kellermaxn, Ch. & E. v. Raumer, Vegetationsversuche an Drosera rotundifolia, mit und
ohne Fleischfiitterung. Bot. Ztg. 36:209-218; 225-229, 1878.

KiRSCHXEGER, M., Notc sur quelques anomalies vegetales. Bull. Soc. Bot. France 2:722-
724, 1855.

KoK, Alida C. a., tjber den Transport korperfremder Stoffe durch parenchymatisches Ge-
webe. Proefschrift, Groningen. Rec. des Trav. bot. neerl. 30:23-139, 1932/3.

KoNOPKA, K., Die Rolle des Kerns bei Verdauung, Sekretion und Reizbewegung der Dro-
sera rotundifolia. Schriften Konigsberger Gelehrt. Gesellsch. Natur%viss. Kl. 7(2) :i3-
112, 1930.

KoNOPKA, K. & ZiEGENSPECK, H., Der Kern des Z)ro.reratentakels und die Fermentbildung.
Protoplasma 7:62-71, 1929.

KosTYTSCHEw, S., Die FhotosyTithese der Insektivoren. Ber. D. Bot. Gesellsch. 41:277-280,
1923.

Kruck, M., {see under Utricularia).

Labbe, E., Du role des microorganismes dans . . . digestions observes chez Drosera rotundi-
folia. These ficole Pharm. Paris, 1904. (Bot. Centralb. 102:333, 1906) {n.v.).

Leavitt, R. G., Adventitious plants of Drosera. Rhodora 1:206-208, 1899. Reversionary
stages experimentally induced in Drosera intermedia. Rhodora 5:265-272, 1903. Seed-
lings and adventitious plants of Drosera. Torreya 9:200-203, 1909.

Leavitt, R. G., Translocation of characters in plants. Rhodora 7:13-20, 1905.

Linderstr0m-Lang & H. Holter, Ergebnisse der Enzymforschung 3'3°9~33'i> i934-

Lubbock, J., A Contribution to our Knowledge of Seedlings. London 1892.

Macbeth, A. Killen, Jr., J. R. Price, F. L. Winzor & A. B. Beck, The coloring mat-
ters of Drosera Whittakeri, I. The absorption spectra and colour reactions of hy-
droxynaphthaquinones. Journ. Chem. Soc. 1:325-333, 1935-

Macbeth, A. K. & Winzor, F. L., The coloring matters of Drosera Whittakeri, IL Journ.
Chem. Soc. 1935:334-336-

MacDougal, D. T., Symbiotic saprophytism. Ann. Bot. 13:1-47, 1899.

Marxoth, R., Flora of South Africa, II (i):26, 1925.

Massart, J., Un botaniste en Malaisie. Gand 1895, p. 253.

Meyen, F. J. F., iJber die Secretionsorgane der Plianzen. A memoir pubhshed in 1837.

MiLDE, J., IJber die Reizbarkeit der Blatter von Drosera rotundifolia L. {through Nitschke,
i860). Bot. Zeit. 10:540, 1852.

Mirimanoff, a., Remarques sur la secretion des tentacules de Drosera. Protoplasma 33, 1939.

Morren, E., La theorie des plantes carnivores et irritables. Bull, de I'Acad. Roy. Belg.,
II, 40:1040 seq., 1875 (1876) (seconde Edition revue et amelioree dans Bull. Fed. Soc.
Hort. 1875).

Morren, £., Note sur les precedes insecticides du Drosera rotundifolia. Bull. Acad. Roy.
Belg. II, 40:7, 1875 (seconde edition dans la Belg. Hort. 25, 1875).

Morren, £., Note sur le Drosera binata, sa structure et ses precedes insecticides. Bull.
Acad. Roy. Belg., II, 40, No. 11, 1875.

Morrison, A., Note on the formation of the bulb in Western AustraUan species of Drosera.
Trans. Proc. R. Bot. Soc. Edin. 22:419-424, 1905 {n.v.).

MouLAERT, B., La regeneration asexuelle chez Drosera. Bull. Soc. R. Belg. Bot. 19:154. i937-

Naudin, M., Note sur les bourgeons nes sur une feuille de Drosera intermedia. Ann. d.
Sci. Nat., II ser. bot., 14:14-16, 1840.

Neubauer, Maria, Vitamin C in der Pflanze. Protoplasma 33:345-370, i939-

Nitschke, Th., Wachstumsverhaltnisse des rundblatterigen Sonnentaues. Bot. Ztg. 18:57-
61; 65-69, i860.

Nitschke, Th., Uber die Reizbarkeit der Blatter von Drosera rotutidifolia. Bot. Zeitg.
18:229-234; 237-243; 245-250, i860.

Nitschke, Th., Morphologie des Blattes von Drosera rotundifolia. Bot. Zeitg. 19:145-151;
233-235; 241-246; 252-255, 1861.

Oels, W., Vergl. Anat. der Droseraceen. Diss. Breslau, 1879.

Okahara, K., Physiological studies on Drosera, I. On the proteolytic enzyme of Drosera
rotundifolia. Sci. Rep. Tohoku Imp. Univ., 4 ser. Biol, 5:573-59°, ^93°^-

Okahara, K., Studies, II. On the effect of quinine and atoxyl on pepsin. Ibid. 5:739-
755, 19306.

Okahara, K., Studies, III. The effect of various acids on the digestion of protein by pep-
sin. Ibid. 6:573-595, 1931- ,. . , . u J- • •

Okahara, K., On the role of microorganisms in the digestion of insect bodies in insectiv-
orous plants. Bot. Mag. Tok. 46:353-357, 1932 (In Japanese with resume in English).

Francis E. Lloyd — 168 — Carnivorous Plants

Okahara, K., Physiological Studies on Drosera, IV. On the function of microorganisms in

the digestion of insect bodies by insectivorous plants. Sci. Rep. Tohoku Imp. Univ.,

4 ser. Biol., 8:151-168, 1933.
OosTERHUis, J., Over de invloed van insectenvoeding op Drosera. Diss. Groningen, 1927.
OuDMAN, J., NahrstoflF-aufnahme und Transport durch die Blatter von Drosera capensis.

K. Akad. V. Wetensch. Amsterdam, Proc. 38:3-15, 1935.
OuDMAN, J., iJber Aufnahme und Transport N-haltiger Verbindungen durch die Blatter von

Drosera capensis. Proefschrift, Groningen. Amsterdam, Mulder & Zn. 1936. Ext. Rec.

Trav. bot. Neerl. 33:351-433. 1936-
Peyronel, B., Bol. Sez. Ital. Soc. Int. Microb. 4:483, 1932.
Pfeffer, W., Osmotische Untersuchungen. Leipzig 1877.
Pfeffer, W., tJber fleischfressende Pflanzen und iiber die Emahrung durch Aufnahme or-

ganischer Stoffe iiberhaupt. Landwirtsch. Jahrb. 6:969-998, 1877.
Pfeffer, W., Zur Kenntnis der Kontaktreize. Unters. Bot. Inst. Tubingen 1:483, 1884.
QuiNTANiLHA {see under Drosophyllum).
Rees, M. & H. Will, Einige Bemerkungen iiber fleischfressende Pflanzen. Bot. Zeitung

33:713, 1875 {see also under Nepenthes).
Regel, E., Futterungsversuche mit Drosera longifolia Sm. und Drosera rotundlfolia L. Bot.

Zeitung 37:645, 1879 {n.v.).
Rennie, E. D., The coloring matters of Drosera Whittakeri. Journ. Chem. Soc. 63:1083-

1089, 1893.
Robinson, W. J., Experiments on Drosera rotundifolia as to its protein digesting power.

Torreya 9:109-114, 1909.
Robinson, W. J., Reproduction by budding in Drosera. Torreya 9:89-96, 1909.
Rosenberg, O., Physiologische und zytologische Untersuchungen iiber Drosera rotundifolia.

Upsala 1899.
Roth, A. W., Von der Reizbarkeit des sogenannten Sonnentaues. Beitr. z. Bot. i, 1782.
RuscHMANN, G., Zur Oekologie von Pinguicula und Drosera. Diss. Jena, 1914.
Salisbury, E., On the occurrence of vegetative propagation in Drosera. Ann. Bot. 29:308-

3io> 1915-
ScHiMPER, A. F. W., Notizen iiber insectenfressende Pflanzen. Bot. Zeitung 40:225-234; 241-

248, 1882.
SCHLICHT, 1889 {through E. Stahl).

SCHMID, G., Beitrage zur Oekologie der insektivoren Pflanzen. Flora 104:335-383, 1912.
Small, J., Intimate camera studies of flowers and plants. Gard. Chron. 105:178, 1939.
SoLEREDER, H., Systematische Anatomie der Dicotyledonen. Stuttgart 1899 and 1908.
Spoehr, H. a. and J. M. McGee, Studies in Plant Respiration and Photosynthesis.

Carnegie Institution Publ. 325, 1923.
Stahl, E., Der Sinn der Mycorhizenbild ung. Jahrb. wiss. Bot. 34:539-668, 1900 {re car-
nivorous plants, pp. 656-661).
Stephens, E. L., A new sundew, Drosera regia Stephens, from Cape Province. Trans. R.

Soc. S. Afr. 13:309-312, 1926.
Stutzer {see under Utricularia).

Tate, Lawson, Insectivorous Plants. Nature 12:251, 1875.
Tischxjtkin, N., Die Rolle der Bacterien bei den Veranderungen der Eiweisstoffe auf den

Blattern von Pinguicula. Ber. d. D. Bot. Gesellsch. 7:346, 1889.
Treat, Mary, Observations on the Sundew. Am. Nat. 7:705-708, 1873.
Tr^cul, a.. Organisation des glandes pedicellees des feuilles du Drosera rotundifolia. Ann.

Sci. nat., IV ser. Bot., 3:303-311, 1855.
Treviranus, C. L., Physiologie der Gewachse. 2(2):759, Bonn 1838 {n.v.).
Troll, 1939 {see under Nepenthes).
Vickery, Joyce W., Vegetative reproduction in Drosera peltata and D. auriculata. Proc.

Linn. Soc. N. S. W. 58:245-269, 1933.
Vries, H. de, tJber die Aggregation im Protoplasma von Drosera rotundifolia. Bot. Zeitung

44:1, 17, 23, 57, 1886.
Warming, E., Sur la difference entre les trichomes .... Verhandl. der Natur. Gesellsch.

in Kopenhagen 1873 {n.v.).
Weber, Fr., Notizen iiber den Z)ro5era-Tentakel-SchIeim. Protoplasma 31:289-292, 1938.
Weber, Fr., Vitamin-C-Gehalt gefiitterter Z)rojrera-Blatter. Ber. d. d. b. Gesellsch. 18:370-

373> 1940.
Weyland, H., Zur Ernahrungsphysiologie mykotropher Pflanzen. Jahrb. f. wiss. Bot. 51:

1-80, 1912.
Whately, in E. Darwin, Botanic Garden, London 1791.

White, J., The proteolytic enzyme of Drosera. Proc. R. Soc. London B. 83:134-139, 1911.
Winkler, H., Uber regenerative Sprossbildung auf den Blattern von Torenia asiatica L.

Ber. D. Bot. Gesellsch. 21:96-107, 1903.
WiNZOR, F. L., The coloring matters of Drosera Whittakeri, III. The synthesis of hydroxy-

droserone. Journ. Chem. Soc. 1935, 1:336-338.
Withering, W., Arrangement of British Plants, ed. 3, London, 1796.
WoLLNY, Ew., Die Zersetzung der organischen Stoffe und die Humusbildung. Heidelberg,

1897.

Chapter XI
CARNIVOROUS FUNGI

Occurrence. — Habit. — Glands. — Secretion. — Digestion.

Among the multifarious activities of fungi, that of zoophagy has
been well known for a very long time. Of this one of the best known
examples is the behavior of Cordyceps, which invades the bodies of
caterpillars of various species and sizes. After displacing the substance
of the body of the larva attacked, preserving its form, however, in the
sclerotium thus formed, the fungus then sends up a linear stalk bearing
the fruiting bodies, the sclerotium being buried in the soil (since
it was there that the larva was destroyed), and the fruiting stalk
rising above in free air. The study of this kind of pathology in rela-
tion to lower forms (algae, small water animals) was being pursued
by the botanist, W. Zopf, in Austria, when there came to his atten-
tion just previous to 1888 a fungus which attacked eelworms {Anguil-
lulidae) .

In the various cultures which he was observing, there were numer-
ous living eelworms and many dead ones tangled with and variously
penetrated by the hyphae of the fungus. The question then arose
in his mind as to whether the fungus is purely saprophytic, pene-
trating only already dead worms, or does the fungus attack and kill
the Kving animal? In answering this question experimentally, Zopf
made the first discovery of a fungus which traps a living animal.

The fungus was Arthrohotrys oligospora, first described by Fres-
ENros (1850-63). It is found in all kinds of more or less decayed
matter — mats of old algae for example — and makes a thin veil of
mycelium of septate hyphae over the surface. From it there
extend slender septate conidiophores bearing pear-shaped two-celled
spores. The peculiar feature is the occurrence on the hyphae of many
slings or loops of various sizes, formed by the sharp curving of a
growing branch which turns upon itself and fuses by its end with its
base. From one loop a second and from this a third may arise, and
thus is formed a tangle of loops lying in all positions, as Woronin
had already observed. It was Zopf, however, who first saw that
living eel-worms were actually caught by these loops, either by the
tail or by the head. The fact that when once the worm has by chance
inserted one end or the other into a loop, it cannot free itself again,
was definitely observed. The eelworm he used was Telenckus scandens,
which infests wheat. One observed, on being caught, struggled vio-
lently for a half-hour, then became quieter and finally died in 2.5
hours. Why the eelworm cannot free itself when once trapped he
attempted to explain by analogy, using as a model a rubber loop
just big enough to allow a finger to enter. When one attempts to
withdraw the finger, the rubber band clamps on the surface and holds
the finger. The clamping effect is due to springiness of the loops, he

Francis E. Lloyd

170 —

Carnivorous Plants

thought. After the animal succumbs, branches from the loop penetrate
its body, and withdraw nutriment (Text fig. 4).

What happens in the case of another similar organism, Dactylella
bembicoides Drechsler, was explained by Couch (1937). In this
plant the loops are composed of a short branch of three cells turned
upon themselves. Fusion occurs between the end and basal cells,
and a neat ring is thus formed. By growing the fungus on agar
to allow of clear microscopic observation, he saw an astonishing thing,
that when an eelworm pokes his head or tail into a ring, the ring
immediately clamps on it by the sudden swelling of the three cells
(Text fig. 4, d, e, i). Couch records his opinion that the rings are
formed most abundantly in media poor in "available food supply,"

Fig. 4. — A, B, Zoophagus insidians (after Gicklhorn 1922); C, Dactylella tylopaga
attacking Amoeba (after Drechsler 1935^); D, E, F, Dactylella bembicoides which attacks
nematodes (after Couch 1937); G, H, Arthrobotrys oligospora, which attacks nematodes
(after Zopf 1888); I, Dactylella betubicoides (after Couch).

judging from experiments which he did. He attempted to get rings
to close on fine glass rods, with limited success, so that it seemed
unlikely that mechanical stimulation suffices. He did find, however,
that heat (water, at t,3 to 75° C.) will cause the rings to close,
but that the temperature of the animal's body enters in as a factor
in nature is quite unlikely. Couch therefore fell back on the perhaps
correct explanation that the fungus responds to a chemical stimulus
from the worm's body. A 1% solution of lactic acid caused a slight
swelling. While this was uncertain. Couch observed that in every
case "when a nematode thrusts its head or tail into one of the rings
it closes practically instantaneously by the simultaneous swelHng of
the three cells of the ring." Later, new hyphal branches penetrate
the body of the prey. Among these predacious species are included
Trichothecium, Arthrobotrys, Dactylaria, Monacrosporium and Dactylella,
all figured by Drechsler (1934a). Still others may be expected to
turn up.

But not all the ring forming fungi act in the same way. We owe

Chapter XI — 171 — Carnivorous

much knowledge about these to Drechsler, who points out (1933c)
that some species have loops the component cells of which do not
swell to constrict the loop, and that these catch their prey by means
of a strong adhesive found on the inner surface of the loop. It is
not unlikely that this is the case in the plant studied by Zopf who
did not observe constriction of the loops. One species has the loop
borne on a very slender stalk which may be broken off during the
struggles of the worm, but this does not obviate death and destruc-
tion, as the cells of the loop can still form their penetrating hyphae
(Drechsler, 1933^).

Some of these species, e.g. Dactylaria Candida (Nees) Sacc, have
in addition to the loops a second organ for the catching of the prey,
called "globular" bodies. These are round knobs on short hyphae.
On the knob is secreted a patch of strong adhesive, by which the
eelworm is caught. In the course of a short time, a penetrating haus-
torium grows through the adhesive pad and enters the animal's body.
In these and the other cases above mentioned, after the prey is per-
meated with haustorial hyphae, and after these have withdrawn all the
available nutriment, the fungal protoplasm withdraws, leaving an
empty shell (Drechsler, 19336, c; 1935c).

A similar method of capture is used by a species in which the
catching organs consist merely of the ends of hyphal branches, pro-
vided, as on the globose organ, with an adhesive. The penetrating
haustorial tube swells up after entrance, and from the sweUing the
haustorial complex of hyphae grows.

A similar, very striking case of a fungus which catches armoured
Rotatoria, the first of its kind known, was described in 191 1, follow-
ing Zopf's original discovery of a carnivorous fungus in Arthrobotrys,
by SoMMERSTORPF Under the name Zoophagus insidians n. gen., n. sp.
This plant grows epiphytically on Cladophora, and consists of a net-
work of septate hyphae which bear "short" branches scattered at
irregular intervals along them. These short branches have dense
glistening contents, and are the organs of capture. Rotatoria (of
the genera Salpina, Metopidia, Colurus, Monostyla), feeding among
the threads of the algae and associated fungus, take hold of the ends
of the short branches, and remain attached, unable to break loose.
By pulHng ofT a newly captured animal he (Sommerstorff) was able
to determine that the end of the hypha had enlarged, apparently by
the swelling of the membrane, which now took up methylene blue
with avidity. Previous to having captured an animal, there appears
to be no adhesive, since no detritus could be observed sticking to
the ends of the short branches, nor did they take up the stain. Som-
merstorff concluded that the swelling takes place on the stimulation
occurring when the animal takes the short branch into its mouth.
Generally the prey cannot escape, despite his size. But as it has
no other organs of locomotion save the cilia, the 'tail' only is avail-
able for struggling. If he can get leverage with this on a neighboring
algal filament, he may and sometimes does escape. After struggling
ceases and death is intervening, the capturing branch grows into a
penetrating tube which then sends numerous thin-walled haustorial
hyphae to withdraw nutriment.

Francis E. Lloyd — 172 — Carnivorous Plants

This organism was studied later also by Mirande and by Gickl-
HORN. The former generally verified Sommerstorff's observations.
He observed, also, the capture of Stylonychia and of other organisms
than armoured Rotatoria, which have the same manner of feeding.
By staining he thought to have identified a substance in the short
h}phae capable of quick swelling.

GiCKLHORN however paid closer attention to the contents of the
short hyphae, the organs of capture. By means of staining and solu-
bility tests he came to the conclusion that an oval glistening body,
observable in the short hyphae, is a discrete body of callose capable
of great swelling. It is always present in organs ready for capture
as a definite plug. In unstimulated branchlets it is always on the
inside; on stimulated ones it occurs as a mucilaginous cap. Dead
organs are always free of the mucilage and are cut off from the bear-
ing hypha by a partition. After a short time following capture (lo-
30 minutes) the callus plug is emptied into the maw of the prey and
spreads out entangling the whole of its ciliary mouth apparatus.
There is no further discharge of mucilage after the expulsion of the
one shot. These results were indeed foreshadowed by Sommerstorff,
but not proven. His suggestion that the mucilage was provided by
the swelling of the outer membrane Gicklhorn could not verify.
GiCKLHORN on the other hand failed to show how the callus plug
makes its escape, since no pore or other point of exudation could be
observed (Text fig. 4A, b).

Gicklhorn studied the mode of capture and its sequelae, confirm-
ing and amplifying such observations as had been made by Sommer-
storff. He observed in freshly caught animals which had succeeded
in escaping that the ciliary apparatus was in a swollen condition.
He asserts that only certain sorts of Rotatoria are caught {Colurus,
Distyla, Metopidia, Monostyla, Salpina and Squalella species) and
never those which are supplied with a strong ciliary apparatus, such
as Brachionus, Noteus, Anuria, Rotifer and Philodina, all of which
were subjected to experimental observation. He was unable to ob-
serve that infusoria such as Stylonychia, Stentor, Paramaecium and the
flagellates Euglena and Paranema were ever caught. This fact, which
he held to be such, indicates at once that not only is the plant a
capturing one but that the animal must be capturable. He admits
observing some infusoria "caught", but they were stuck to the catch-
ing organs, and this does not prove that they were properly caught
in the manner of Rotatoria. He concludes that instead of speaking
of the animal as being caught one should say that it gets itself caught,
since only those armoured Rotatoria which are able to swallow the bait
can be caught. In a culture with many animals the process was
followed and this account is given. If an animal hits against the
main hypha or against a short hypha sideways, these bend a little
under the impact and then recover. The animal, on hitting, infolds
its ciliary apparatus; if, however, it approaches a short hypha, the
trapping organ, end on, so that it enters the ciliary apparatus, the
latter immediately clamps down on it and draws it in. This is done
repeatedly for 5-10 minutes during which interval repeated attempts
are made by the ciliary apparatus to open, only to clamp down again

Chapter XI — 173 — Carnivorous Fungi

in response to the mechanical stimulus provided by the short hypha.
At the end of this interval it can now be shown by staining methods
that the callus plug has been emptied and that it has swollen and
spread out to involve the entire ciliary apparatus, which is now ren-
dered useless. In this condition the animal finds its weak foot useless
in effecting escape, and in the course of another period of twenty
minutes it ceases to struggle. From these observations Gicklhorn
draws the following conclusions. In the first stage of capture, there
is no adhesive effect on stimulation of the short hypha, as Sommer-
STORFF thought, but repeated mechanical grasping of it by the re-
tractile ciliary apparatus of the animal. Secondarily there follows
the excretion of the mucilage. This is an active process on the part
of the hving short hypha on stimulus, and is not a simple swelling
of the membrane. He tried, with success, to stimulate short hyphae
to throw off their mucilage by stroking them with a fine hair. The
short hypha now begins to send out haustoria which penetrate through-
out the body of the animal. Even at the end of digestion, the mucilage
plug, which can still be seen, is found to have hardened and become
yellow in color, holding the shell of the animal in position. The
growth of the haustoria proceeds till the interior of the body is a mass
of hyphae which send out conidiophores projecting from the animal
and in swarm-spores produced in a sac which escape through the
mouth end.

A plant with a similar method of capturing its prey as that em-
ployed by Zoophagus insidians is Sommerstorffia spinosa, described by
Arnaudow (1923). Both of these species have been collected in
Massachusetts and observed by Sparrow (1929).

An extraordinary group of fungi which prey upon species of Amoeba
and shelled rhizopods has been uncovered and studied by Drechsler.
His accounts include the minutiae of taxonomic interest as well as
the mode of capture. We need not take consideration of the former
here. They are nearly all plants with septate hyphae producing
conidia of various forms, and in some cases the sexual method of
reproduction is known. The method of capture is quite similar in
all cases. The species of Amoeba appear to be large. Amoeba ter-
ricola or related species being often the victim. There is evidence
that certain fungi can attack only one kind of Amoeba and not an-
other. In some fungi an adhesive has been observed, in others not,
leaving it for conjecture that a non-visible adhesive occurs. There
is seldom any preformed structure with the function of capture, but
this occurs in Dactylella tylopaga Drechsler. In this "prolate el-
hpsoidal protuberances" are provided with an adhesive. An animal
sticks to one of these, which then sends out a tube of penetration.
This grows inside the animal into a branching mass of short h>^hae
which absorb the body of the animal.

Pedilospora dactylopaga captures shelled Rhizopods (Drechsler
1934). Eight species of Acaulopage have been described by Drechs-
ler, all of which capture Amoebae in much the same way as Dac-
tylella tylopaga except that there is no special organ involved in capture.
"An Amoeba after capture is always to be seen attached whether
to a mycelial element or as is often the case in some species, to a

Francis E. Lloyd — 174 — Carnivorous Plants

fallen conidium by means of a minute mass of golden yellow adhesive
material." "From the mycehal element or the conidium is thrust
forth a narrow process which passes through the deposit of adhesive
material and perforates the animal's pellicle to give rise inside to a
more or less characteristically branched haustorium or haustorial
system. When the protoplasmic contents of the Amoeba are nearly
exhausted, the protoplasm of the haustorium begins to withdraw
back into the parent mycelial filament. Eventually the haustorium
is completely evacuated and thereupon, like the collapsed pellicle
surrounding it, becomes altogether invisible; so that an instance of
capture is afterwards found recorded, and then usually only rather
dubiously, in an inconspicuous scar-hke or shghtly protuberant modi-
fication of the contour of the hypha or conidium" (Text fig. 4c),
(Drechsler 1935&, p. 183). In the case of another fungus, Endo-
cochlus asteroides Drechsler, the animal is attacked by conidia picked up
in its wanderings. Sticking to the surface of the pelhcle, they form a
small bulbous body, serving apparently as an appressorium, through
which a slender tube punctures the pellicle and enters the animal, pass-
ing in to some distance. There the end swells up, taking in the proto-
plasm of the conidium, which becomes detached and is usually thrown
off by the animal. Sometimes the conidium is ingested, however.
Owing to the fact that the animal may be infected a number of times,
because of the numbers of fallen spores, it may have a corresponding
number of bulbous bodies, each derived from a conidium. After the
conidium with its germ tube is loosened and cast off, the remaining
globular thallus becomes considerably enlarged and turgid. As it length-
ens it curves and with elongation becomes a helicoidal mass. In the
meantime, the animal remains alive and active, so that we are con-
templating here a case of parasitism. The briskness of action per-
sists for some time, until the bulk of the animal becomes reduced,
and it finally succumbs. The inclosed fungus then sends out slender
hyphae which penetrate the pellicle to the exterior, where spores
and sexual apparatus are produced, to produce new hyphae which
begin the cycle again. The same story is presented by Cochlonema
verrucosum Drechsler, and in C. dolichosporum, but in these it^ is
started by conidia which are first ingested by the animal, one having
the same dimensions as Amoeba sphaeronucleus. In Bdellospora heli-
coides Drechsler the infection takes place as in Endocochlus asteroides.
In Zoopage phanera Drechsler the manner of capture is a matter
of inference rather than direct observation. The animal captured
is an Amoeba from 35-110 micra in diameter. An adhesive is in-
dicated, though Drechsler suggests that the small botryoidal or-
gans seen in a captured animal could be taken for grappling organs.
At all events they are very distinctive in form as his figure shows.
In the forms above described it is evident that we are dealing
with organisms that stand between plants which have elaborated
organs designed — if we may use the word — for first trapping an
animal before disposing of it, and those which infect an animal by a
process which must be repeated in a very many cases, as for instance
that of Cordyceps and related plants already mentioned. Carnivorous
the latter are, but they can hardly be regarded as "trapping" plants.

Chapter XI — 175 — Carnivorous Fungi

The significance of carnivory for Zoophagus (and hence by implica-
tion for fungi in general) has been indicated experimentally by Gickl-
HORN (1922). By culturing it for two months in properly prepared
water, free of animals, he found that it could persist saprophytically,
as many other fungi do. Under these conditions, however, it became a
"hunger form", with the "long hyphae", though normal as to branch-
ing, weakly developed, and with depreciated cellular contents. In two
days after the addition of Rotatoria, the hyphae became appreciably
stronger and were well filled with contents, "After such evidence one
can hardly avoid the thought that the capture of animals by Zoophagus
supplied an important source of nutrition and that we have before us a
highly specialized adaptation" {I.e. p. 217).

Literature Cited:

Arnaudow, N., Zur Morphologic und Biologie von Zoophagus insidians Soramerstorff.

Jahrb. d. Univ. Sofia 15-16:1-32, igiS-Cigai) (Bulgarian with German summary).
Mnaudow, N., Ein neuer Radertiere (Rotatoria) fangender Pilz (Sommerslorffia spinosa

nov. gen., nov. sp.). Flora 116:109-113, 1923.
Arnaudow, N., Untersuchungen iiber Sommerstorffia spinosa nov. gen., nov. spec. Jahrb.

d. Univ. Sofia Bd. 19, H. 2, Abt. la, 1923.
Arnaudow, N., Untersuchung iiber den Tiere fangenden Pilz Zoophagus insidians Som.

Flora 118-119:1-16, 1925.
BuDDE, E., tjber die in Radertieren lebenden Parasiten. Arch. f. Hydrob. 18:442-459.
Couch, J. N., The formation and operation of the traps in the nematode-catching fungus,

Dactvlella bembicoides Drechsler. Jour. Elisha Mitchell Sci. Soc. 53:301-309, ^937-
Drechsler, C, Morphological features of some fungi capturing and killing Amoebae.

J. Wash. Acad. Sci. 23:200-202, 1933a.
Drechsler, C, Morphological diversity among fungi capturing and destroying nematodes.

J. Wash. Acad. Sci. 23(3):i38-i4i, 19336.
Drechsler, C, Morphological features of some more fungi that capture and kill nema-
todes. J. Wash. Acad. Sci. 23(5):267-27o, 1933c.
Drechsler, C, Several more fungi that prey on nematodes. J. Wash. Acad. Sci. 23(7):

355-357, igssd.
Drechsler, C, Organs of capture in some fungi preying on nematodes. Mycol. 26:135-

144, 1934a.

Drechsler, C, Pedilospora dactylopaga n. sp., a fungus capturing and consuming testa-
ceous rhizopods. J. Wash. Acad. Sci. 24:395-402, 19346-

Drechsler, C, Some conidial Phycomycetes destructive to terricolous Amoebae. Mycol.
27:6-40, 1935a.

Drechsler, C, Some non-catenulate conidial Phycomycetes preying on terricolous Amoe-
bae. Mycol. 27:176-205, 19356.

Drechsler, C, A new species of conidial Phycomycete preying on nematodes. Mycol.
27:206-215, 1935c. ...

Drechsler, C, A new Mucedinaceous fungus capturing and consuming Amoeba verrucosa.
Mycol. 27:216-223, i93S<f.

Drechsler, C, A new species of Stylopage preying on nematodes. Mycol. 28:241-246,
1936a.

Drechsler, C, New conidial Phycomycetes destructive to terricolous Amoebae. Mycol.

28:363-389, 19366. . ^ , „ . . • . .

Drechsler, C, A Fusarium-like species of Dactylella capturing and consuming testaceous
rhizopods. J. Wash. Acad. Sci. 26:397-404, 1936c.

Drechsler, C, New Zoopagaceae destructive to soil rhizopods. Mycol. 29:229-249, 1937a.

Drechsler, C, Some Hyphomycetes that prey on free-living terricolous nematodes. My-
col. 29:447-552, 19376. _ ., A u n* 1 •

Drechsler, C, New Zoopagaceae capturing and consuming soil Amoebae. Mycologia
30:2:137-157, 1938.

Drechsler, C, A few new Zoopagaceae destructive to large soil rhizopods. Mycologia

31:2:128-153, 1939.

Drechsler, C, Five new Zoopagaceae destructive to Rhizopods and Nematodes. My-
cologia 31 :4:388-4i5> i939- „ , r /rj -7 \

Fresenius, Beitrage zur Mycologie. Frankfurt, 1850-63, p. 18, pi. 3, fags. 1-7 {fide f^ovY).

Geitler, L., iiber einen Pilzparasiten auf Ajnoeba proleus und uber die polare Organisation
des Amoebenkorpers. Biol. Zentralbl. 57:166-175, 1939.

Francis E. Lloyd —176— Carnivorous Plants

GiCKLHORN, J., Studien an Zoophagiis insidians Som., einem Tiere fangenden Pilz. "Glas-

nik" Kroat. Nat. Ges. 34(2):i99-288, 1922.
GiCKLHORN, J., Aphanomyces ovidestruens nov. spec, ein Parasit m den Eiern von Diap-

tomus. Lotos 71:143-156, 1923. , , . T^ . , .

KoNSULOFF, St., Untersuchungen liber Rotatonenparasiten. Arch. f. Protistenk. 30:353-

MiRANDE, R., Zoophagiis insidians Sonunerstorff, capteur de Rotiferes vivants. Bull. Soc.

Myc. Fr. 36:47-53, 1920.
Rennerfelt, E., Untersuchungen iiber die Entwicklung und Biologic des Krebspestpilzes,

Aphanomyces astaci Schikora. Mitt. Anst. f. Binnenfischerei bei Drottninghobn, Stock-

hohn. No. 10, 21 pp., 1936. _ , ^ .„ . j • •

Scherffel, a., Endophytische Phycomyceten-Parasiten der Bacillanaceen und emige neue

Monadinen. Archiv Protistenk. 52:1-141, 1925. _

Schikora, F., Uber die Krebspest und ihren Erreger, Aphanomyces Magnusi Schikora.

Verhandl. Bot. Verein Prov. Brandenburg 63:87-88, 1922.
SOMMERSTORFF, H., Ein Ticrc fangender Pilz {Zoophagiis insidians nov. gen., nov. sp.).

Oest. Bot. Zeitschr. 61:361-373, 191 1.
Sparrow, F. K., Jr., A note on the occurrence of two rotifer-capturing Phycomycetes.

Mycol. 2i(2):90-96, 1929. , , „ , t.-. a i.-

Valkanov, a., Uber Morphologie und Systematik der rotatorienbefallenden Pilze. Archiv

Protist. 74(0:5-17, 1931. . , , ,, J ^

Valkanov, A., tjber die Morphologie und Systematik der Rotatonen befallenden Uomy-

ceten (bulgarisch). Jahrb. Univ. Sofia, 27, 1931. . r- -n •

Valkanov, A., Nachtrag zu meiner Arbeit uber rotatorienbefallende Pilze. Archiv Protist.

78(2) :485-496, 1932.
ZoPF, W., Zur Kenntniss der Infectionskrankheiten niederer Thieren und Pllanzen. Nova

Acta d. Leop.-Carol. Akad. d. Naturf. 52:315-375. 1888.

Chapter XII

DIONAEA MUSCIPULA AND ALDROVANDA

VESICULOSA

Dionaea: general description. — Early discovery. — Original description by Ellis. —
Work of Curtis (1834), Oudemans (1859), Caxby (1868), Darwin, Goebel. — Mor-
phology (Seed and seedling. Structure of mature leaf: trap. Lobes, glands, sensitive hairs.
Internal structure). — Physiology. — Aldrovanda: general description. — Discovery, dis-
tribution.— Morphology (Seed. Germination. Mature leaf. Posture of the trap). —
Physiology.

These two monotypic genera are members of the family Droser-
aceae, and while the former, Dionaea, is well known, it being widely
grown in greenhouses, Aldrovanda is well known chiefly to such bot-
anists as have a special interest in these curious plants. Dionaea
has a very restricted geographical range, Aldrovanda a very wide
one. Though the method of trapping animals is identical, the one
is a land plant, and Aldrovanda a submersed water plant. We con-
sider these separately.

Dionaea muscipula Ellis, Venus' fly trap: — This is a small
plant (77 — 2), consisting of a rosette of leaves three to six inches
across arising from a rootstock growing more or less horizontally.
The rootstock is apparent even in the young seedling (Smith 193 i).
Long scapes are sent up bearing several flowers in a short cyme with
two to fourteen flowers. These are of pentamerous structure, five
small elliptical sepals alternate with five white cuneate and somewhat
oblique petals, usually fifteen stamens. The leaf consists of two re-
gions, a basal "footstalk" as Darwin called it, articulated by means
of a short cylindrical portion (midrib) with the blade which is a trap.
The footstalk is a more or less expanded leaf-like structure, either
broadly obcordate to narrowly obcordate in form, depending on ex-
posure to fight and the presence of surrounding vegetation. The
upper part of the leaf, a "striking and noteworthy" trap, to quote
Goebel, consists of two dished lobes of trapezoidal form. The outer
margins are "ciliated," that is, are provided with a row of coarse
projections, prongs or teeth. Ellis (1770) spoke of the arrange-
ment as "a miniature form of a rat-trap," and Curtis (1834) com-
pared it to "two upper eyefids joined at their bases." Springing
from the upper surface of the two lobes there are six slender, sensi-
tive hairs, three on each side placed in triangular position (in ex-
ceptional cases a smaller or larger number has been noted {iS — i).
The rest of the surface is covered rather densely with two kinds of
sessile glands, most of which under usual circumstances are colored
with brilliant red pigment, giving a bright red tinge to the surface.
When with suitable temperatures the sensitive hairs are moved, the
two lobes swing swiftly on their common axis, and the finger-like
cilia intercross to form a barred cage. Darwin interpreted this ini-
tial posture as an arrangement to allow small insects, relatively value-

I'-.

Francis E. Lloyd — 178 — Carnivorous Plants

less on account of their size, to escape before final closure, a view
for which F. M. Jones has adduced constructive evidence. Subse-
quent movement progressing during >^-i2± hours approximates the
lobes more closely and even causes them to become flatter, bringing
the inner surfaces into closer apposition.

If by this reaction an insect, the normal agent of stimulation,
has been caught, the body may be more or less compressed between
the lobes (17 — 3). The glands then secrete a digestive fluid and
in a few days the insect body disintegrates and the products are
absorbed. In the course of ten days the lobes open again, and are
ready to catch other prey. This may be repeated two or three times
before the leaf reaches its complete maturity, when it dies. Of this
arrangement Curtis (1834) remarked, "if it were a problem to con-
struct a plant with reference to entrapping insects, I cannot con-
ceive of a form and organization better adapted to secure that end
than are found in Dionaea muscipula."

" This plant, which Linnaeus called miraculum naturae, appears
to have first been discovered by Arthur Dobbs, Governor of North
Carohna, and he sent the following account of it to Mr. Collinson
in a letter dated at Brunswick, Jan. 24, 1760. After describing the
Schrankia, he proceeds:— 'But the great wonder of the vegetable
kingdom is a very curious unknown species of sensitive; it is a dwarf
plant; the leaves are like a narrow segment of a sphere, consisting
of two parts, like the cap of a spring purse, the concave part out-
ward, each of which falls back with indented edges (like an iron
spring fox trap); upon anything touching the leaves, or falHng be-
tween them, they instantly close like a spring trap, and confine any
insect or anything that falls between them; it bears a white flower;
to this surprising plant I have given the name of Fly Trap Sensitive.'
Mr. Collinson, in a memorandum, has recorded the death of Gov-
ernor Dobbs in 1765" (Dillwyn 1843).

It may be inferred from the note by Governor Dobbs, written

in 1760, but which did not gain publicity till the appearance of the

Hortus Collinsonianus in 1843, that the Dionaea was well known in

North Carolina when, in 1763, a Mr. Young, "the Queen's botanist,"

had his attention drawn by some friends to a "pecuhar plant" which

he subsequently found in great abundance in North Carohna and

in some parts of South Carohna. Still later Young brought hving

plants to England, he also introduced them in Kew (Sims 1S04) and

from these Ellis, a London merchant, drew his description and

figure sent by him to Linnaeus in 1770 (Young 1783). These were

published in a small volume entitled. Directions for Bringing over

Seeds and Plants from the East Indies and other Distant Countries in a

State of Vegetation, in 1770. Ellis' description was pubHshed in

Latin at a subsequent date in Nova Acta Soc. Scient. Upsaliensis

1:98, 1773. Though Ellis' description was based on living material,

he had just previously received from John Bartram of Philadelphia,

through Mr. Peter Collinson, an herbarium specimen which furnished

material which enabled Dr. Solander and himself to determine that

they had before them a new genus allied to Drosera (Ellis 1770). But

Ellis, who from the dried material got no hint of the motility of the

Chapter XII — 179 — Dionaea and Aldrovanda

leaves, did so when he examined the material brought to England by
Young, and he was the first to publish a definite statement of the
carnivorous habits of the plant. It is this, expressed in the letter to
Linnaeus, which interests us here. Having recalled that Mimosa is
irritable, but shortly recovers from its position of response, Ellis
continues :

''But the plant, of which I now inclose you an exact figure, with
a specimen of its leaves and blossoms, shews, that nature may have
some view towards nourishtnent, in forming the upper joint of the
leaf like a machine to catch food: upon the middle of this lies the
bait for the unhappy insect that becomes its prey. Many minute
red glands, that cover its inner surface, and which perhaps discharge
sweet liquor, tempt the poor animal to taste them; and the instant
these tender parts are irritated by its feet, the two lobes rise up,
grasp it fast, lock the rows of spines together, and squeeze it to death.
And, further, lest the strong efforts for life, in the creature thus taken,
should serve to disengage it, three small erect spines are fixed near
the middle of each lobe, among the glands, that effectually put an
end to all its struggles. Nor do the lobes ever open again, while
the dead animal continues there. But it is nevertheless certain that
the plant cannot distinguish an animal, from a vegetable or min-
eral, substance; for if we introduce a straw or a pin between the
lobes, it will grasp it full as fast as if it were an insect."

The above paragraph, quoted also by Hooker in 1875, was taken
from the original, a copy of which is to be found in the Library of
Congress, Washington. It shows us clearly what Ellis thought
about the plant. Linnaeus, to whom was sent the letter contain-
ing the above quotation, did not, however, fully respond to Ellis'
evident enthusiasm, and merely regarded the movement as a special
case of irritability like that in Mimosa, and believed that on reopen-
ing the captured insect was released {Mantissa plantaruni altera,
Holmiae 1771); nor did he see eye to eye with Ellis about the func-
tion of the "three small erect spines" whether from sagacity, as
Hooker suggests (1875), or as part of his general non-responsiveness
to Ellis' interpretations, it is hard to say. It happens, of course,
that Ellis was wrong; the idea was even fantastic. No less was
that of Erasmus Darwin who wrote: "In the Dionaea muscipula
there is a still more wonderful contrivance to prevent the depreda-
tions of insects: the leaves are armed with long teeth, like the an-
tennae of insects, and lie spread upon the ground around the stem,
and are so irritable, that when an insect creeps upon them they fold up
and crush or pierce it to death." {The Botanic Garden 2 : canto I, p. 39).

E. MoRREN (1875) in a footnote calls attention to a description
by the French encyclopaedist Denis Diderot (1713-1784) of "Plante
de la Caroline appellee Muscipula Dionaea", of which he says, at
the close of his description "Voila une plante presque carnivore".
The essay from which this is quoted is said to be dated 1 774-1 780,
not 1762 as M. Catalan told Dr. Morren. It is of interest to
record Diderot's speculation "Je ne me doute point que la Muscipula
ne donnat a I'analyse de I'alcali volatil, produit caracteristique du
regne animal."

Francis E. Lloyd — 180 — Carnivorous Plants

In 1834 M. A. Curtis (quoted on p. 177), a minister resident in Wil-
mington, N. C, published his observations, which led him to think
that the sensitiveness resides only in the hair-like processes, and
that other parts of the leaf may be touched or pressed without any
response. This is not quite true, as has later been found. Further,
that insects captured are not always crushed on being caught, for
if the trap were opened again they might escape; but that in time
they were surrounded by a mucilaginous fluid by which the insects
were more or less consumed. The fact that a special sensitivity
resides in the six slender hairs of the upper surface of the trap had
been noted by a botanical draughtsman, Sydenham Edwards, em-
ployed by Dr. John Sims in illustrating Curtis's Botanical Magazine.
This observation was recorded by Sims (1804) in the description of plate
785 of vol. 20 which reads: "These small spines are mentioned and
figured by Ellis and supposed by him to assist in destroying the en-
trapped animal; but that they are only irritable points, and that any
other part of the leaf may be touched with impunity, was discovered
by our draughtsman, Mr. Edwards, several years ago, when taking a
sketch of a plant flowering at Mr. Lipman's, Mile End, and has since
been repeatedly confirmed. The same observation was made, without
knowing it had previously been noticed, by our friend Mr. Charles
Konig" (Hooker 1875).

In 1859 OuDEMANS, a Dutch botanist, rediscovered the sensi-
tivity of the trigger hairs, and did a number of experiments which
afforded results which anticipated some of Darwin's. He found
that there is no periodic closure of the trap, as Meyen has claimed,
and that traps, after closure, opened again during the night. Meyen
had said that the closure was too slow to catch insects, to which
Oudemans answered that at sufficiently high temperatures it is rapid,
which we now know to be true. He recorded that the trap does
not open after catching prey until several days after its death. When
the trap does, the prey is found lying in a slimy liquid; and further
that the trap does not remain closed over inanimate objects such as
paper, reopening in 36 hours or less. Though he thought that the
stimulus was transmitted to the mid-vein, he attributed closure to
alterations of strain in the parenchyma.

In 1868 Canby thought it might be that the fluids collecting
in the closed trap might escape and, flowing down the petioles of
the leaves, might enrich the soil at the base of the plant. Experi-
ments showed him, however, that this is not the case, but that, on
feeding the leaf, the insect is entirely destroyed and absorbed, thus
confirming Curtis, thirty-four years his predecessor. He concluded
by saying, "so that, in fine, the fluid (secreted by the leaves) may
well be said to be analogous to the gastric juice of animals,^ dissolv-
ing the prey and rendering it fit for absorption by the leaf."

That, however, the sensitive hairs are the only sensitive spots
in the leaf was shown by Darwin and by Goebel (1891) not to be
true. "It is sufiicient to rub the upper or lower surface, and not
too strongly, with a solid object to procure immediate closure of the
two halves of the trap" (Goebel). This, as will be seen, has later
been confirmed.

Chapter XII — 181 — Dionaea and Aldrovanda

We now direct our attention to certain of the more intimate de-
tails of the life history and structure of the plant.

The seed and seedling (Smith i 931). — The seed is a small pear-
shaped mass with a lid at the micropylar end. The embryo, also
pear-shaped, Hes with its two broad cotyledons in contact with an
abundant endosperm. The cotyledon-ends remain in contact with
the endosperm for an extended period, while their bodies enlarge,
become green, and eventually spread apart, becoming elongate in
form. Meanwhile the primary root has developed, bears a niass of
root hairs, but does not persist. From the plumule there arises at
once a rhizome bearing small leaves which are similar to the mature
leaves in all points, except that they are small (Goebel 1891, Holm
1891), the shape of the trap is more rectangular, and the glands
are fewer in number. This story, as we shall see, is similar to that
of Aldrovanda.

Although, as Miss Smith says, the juvenile leaves (those of seed-
lings) are similar to those of mature plants, there are differences
worthy of note, among which are the following: Traps 2 nrni. long
have poorly developed marginal spines of rather irregular form and
little rigidity, and may arise symmetrically, so that those of the
opposing lobes face each other (/5 — 8); or there may be a conspic-
uous lack of symmetry and a disparity of number (18 on one side
and 13 on the other), some teeth displaying branching to some ex-
tent {i8~g). Traps of larger size (4 mm. along the midrib and 7
mm. along the free margins) have well developed spines with pos-
ture and rigidity comparable to traps on mature leaves. The number
of glands in juvenile leaves is much smaller, while the size of the
glands approaches that of maturity. They are rather widely scattered.
I counted about 70 in the 2 mm. trap {i8~()). There is a wide
zone between the outer limit of this glandular region and the lobe
margin; and in the outer zone of allure, a narrow zone just within
the margin, the nectar (?) glands are only 2 to 12 in number, con-
fined to the outer angles of the lobes in the 2 mm. traps examined;
while as many as 50 were found in the 4 mm. trap. As would be
expected, the structure of the lobes is of much greater delicacy as
compared with that of mature leaves. In the 2 mm. trap the lobe
was found 0.5 mm. thick at the level of an inner sensitive hair {18 —
17, 18). The inner (upper) epidermis cells were somewhat larger than
those of the outer in the ratio of 1.4 to i, and had somewhat thicker
outer walls. The number of parenchyma cells between ranged from
two to four courses with large interspaces, in this feature again re-
sembling the mature leaves of Aldrovanda much more than do
the thicker mature leaves of Dionaea. It is easier, thus, to see the
parallelism of action in these two plants, to which reference is
made beyond. The sensitive hairs are much smaller and simpler
in construction {18 —t,)- They are about 0.6 mm. long, the outer
stiff "lever" being somewhat more than half that. The basal por-
tion is deeply constricted, and the bending cells are relatively large
and impinge on each other in the middle of the hair, there being
no medullary cells. These latter seem, therefore, of no importance
beyond that of a filler in the large, sensitive hairs of the mature leaf.

Francis E. Lloyd — 182 — Carnivorous Plants

In young plants derived from leaf cuttings (seen growing in Mu-
nich) the lobes of the traps were usually quite oblique in form, and
the cilia were very irregular and often very small. Both kinds of
glands were present in small numbers. I counted six alluring glands
clustered near each of the outer angles of the lobe in one, and three
in another trap. The digestive glands were more numerous, but still
few and scattered. The sensitive hairs were absent in one case, and
from one to two occurred in several others. While small, they showed
all the normal histological details.

During development the blade and trap both display circination.
In the latter the two lobes are rolled inwardly and gradually un-
fold, the cilia being the last to unroll. The blade lobes are inrolled
longitudinally at first, but as it develops the axis of the roll grad-
ually comes to lie more transversely, the last portion of the blade
to unroll being the apex. At first also the trap is bent sharply back
on the blade, but lying asymmetrically on its right lobe, which presses
against the blade surface. With further growth the trap swings
forward on its stalk, and now comes to lie more or less on its left
lobe, or otherwise expressed, the trap is twisted more or less to the
left (as seen from above) (iS — 7). This posture is more evident
in small plants, with small traps, than in plants large enough to
produce normal sized traps. It is, however, often quite evident in
the latter, as I have myself observed in plants growing under my
eye. In the account beyond of Aldrovanda, it will be seen that in
this plant the trap always lies on its left side, a position which offers
distinct advantages in the trapping of prey. Much less marked, how-
ever, is the posture just mentioned for Dionaea, and it cannot be
said to be of any significance. I have never seen the trap bending
to the right (as seen from above).

Structure of the mature leaf. — The structure of the leaf has often
been the subject of examination (Oudemans, 1859; Daewin, 1875;
Kurtz, 1876; de Candolle, 1876; Fraustadt, 1877; Goebel, 1891;
Haberlandt, 1901; GuTTENBERG, 1925). The winged form of the
petiolar region traversed by a single vascular bundle, which presents
nothing further of special interest, is regarded by Goebel as a physio-
logical compensation in the interest of photosynthesis, such compen-
sations being generally found among the carnivorous plants. In
seedlings the petiole is relatively much larger.

The outer part of the leaf, the blade, is the trap. This consists
of two lobes, trapezoidal in form, united along the middle line by a
thick midrib, which has often been considered a hinge, but which
has no hinge function {i8 — i). At their bases the lobes have their
greatest thickness, thinning off gradually as the margins are approached
{18 — 4b). Here, however, they are thickened locally by the enlarged
bases of the marginal cilia. These are prominent, tapering, finger-like
processes, evidently emergencies (Solereder), which are so placed
that when the lobes are approximated they interweave like the fingers
of closed hands. The cilia have been thought to be homologous with
the tentacles of Drosera, but, as Goebel pointed out, the comparison
fails in that the ciha show no trace of glandular tissue, and have
evidently a widely different function. The whole trap acts, to quote

Chapter XII — 183 — Dionaea and Aldrovanda

LiNDLEY (1848), like the jaws of a steel trap. The simile may not
be appHed too rigorously since it is not the edges of the lobes which
catch the prey.

The two lobes, when the leaf is widely open, stand at an angle
of 40-50 (Darwin: ''80") degrees to each other, published drawings
being often in error on this point (iS — 40). They are clothed with
a distinctly firm epidermis of straight walled cells, elongated parallel
with the veins, becoming somewhat wavy on approaching the margins,
which lends a surprising stiffness to the trap. When once a lobe is
cut, as in making sections, it becomes evident that the epidermis is
the only mechanical tissue present. The outer surfaces bear scattered
stellate trichomes {18 — 11) which are found also in the bays between
the ciha of seedhng traps, and even shghtly invading their inner sur-
faces. The inner surfaces are supplied with very numerous glands,
all having the same structure. They consist of two basal epidermal
cells, placed parallel with the midrib, and whose walls are thickened
by cellulose ridges, producing an appearance which led Macfarlane
(1892) to take them to be ''intercellular protoplasmic connections."
What he saw, it seems, are only layers of cytoplasm lying between
the cellulose ridges, but this is not to deny that such protoplasmic
connections may not also be present. Surmounting the two basal
cells is a second course of two small cells, with cutinized diametrical
wall, constituting the stalk. This supports a large capital of about
32 cells in two courses, the lower capped by the upper to form a bun-
shaped mass (iS — 12, 13). There are two physiological kinds of
these glands, as evidence adduced by Frank Morton Jones indicates
(1923), namely digestive (and absorptive) and alluring. The former,
thought by Ellis "perhaps" to discharge a sweet liquor, occupy the
major area of the surface and are so numerous that they often crowd
on each other {18 — i). The alluring glands occupy a narrow zone
just within the ciHated margin. Between the two groups of glands
there is a narrow zone quite free of glands (Jones). Though identical
in structure, the digestive and alluring glands display some differences.
The digestive glands {18 — 12, 13) are rendered conspicuous by their
deep red color, due to anthocyanin present in the sap of their cells,
and are responsible for the deep red note of color of the inner surface
of the trap. The alluring glands (in the traps I examined) contain
no pigment. These are imbedded somewhat in the epidermis {18 —
14), while the digestive glands stand out prominently. These also
are larger (0.096-0.1 mm. in diam.) than the alluring glands, which
measure 0.06-0.073 mm. in diameter. That the alluring glands se-
crete a sugar (or something attractive to insects) is supported strongly
by Jones' observations already alluded to: "... these little ants
were observed to occupy a uniform position on the upper surface,
their heads close to the bases of the marginal spikes. As they moved
slowly across this belt of the leaf they made frequent and prolonged
pauses, during which their mouth parts were observed under the lens
to be in motion against the surface of the leaf. A larger and winged
hymenopteron was observed to be engaged in the same performance.
Obviously they were feeding upon some attractive exudation of the
leaf. The behavior of visiting insects is entirely convincing to the

Francis E. Lloyd — 184 — Carnivorous Plants

observer that a baited area extends across the leaf surface just within
the bases of the marginal spines. This baited marginal band is so
situated upon the leaf surface that a visiting insect in length too small
to reach from the halt to the trigger hairs, usually does not spring the
trap. Whether or not these conditions are to be interpreted as ad-
justments to that end, the effect of the arrangement, in conjunction
with the peculiarities of the closing movement by which small insects
are given an opportunity to escape, is to limit the captures of the
leaves to insects approximating one quarter of an inch or more in
length." Jones examined the captures of fifty closed traps and found
that, of all the prey "only one was less than 5 mm. in length and
only seven less than 6 mm.; they were 10 mm. or more in length,
with a maximum of 30 mm." (Jones 1923). In this way was cor-
roborated Darwin's suspicion that the posture first assumed by the
trap on closure in which the marginal spines form a cage is one which
permits small insects to escape. I have observed larger ones, small
centipedes, doing their best to force their way between the spines,
but without success. A wood louse was seen to free itself because
its position was such that its carapace held the lobe margins open
just enough to allow escape, which was evidently facilitated by the
fact that the lateral projections of the carapace allowed leg movement.
Many a wood louse is not so lucky.

The closure of the trap, a seismonastic movement, normally follows
when sensitive or trigger hairs are disturbed as Curtis recorded in
1834. "Each side of the leaf is a little concave on the inner side
where are placed three deHcate, hair-hke organs, in such an order,
that an insect can hardly traverse it, without interfering with one
of them, when the two sides suddenly collapse and enclose the prey
with a force which surpasses the insect's efforts to escape." Though
usually three in number on each lobe, there may on occasion be more
or fewer. When three, they stand at the angles of a triangle
placed in the middle of the lobe with its base nearer and parallel
to the outer cihated margin. If we examine one of these hairs we
find that it is multicellular and displays two distinct regions. The
outer of these is a slender cone {18 — 2) in form, about 1.5 mm.
in length and 0.15 mm. thick at the base. It is composed of elon-
gated, thick walled cells and constitutes a lever; any slight move-
ment causes a bending in the basal region, to which Oudemans (1859)
attributed a special sensitivity. This is only 0.15 mm. in height
and is conspicuous on account of a deep constriction slightly below
the base of the lever, first described by Goebel (1891). This con-
striction, the hinge or bending place (Goebel), is made up of a single
transverse ring of cells of which their outer walls are deeply indented
{18 — 15, 16). Their lateral walls are thick and collenchymatous in
character, their end walls thin. Because of the indentation, the outer
wall is thinner at this point. Within the ring of indented cells there
is a medullary group of elongated cells, tracheidal in character
(Goebel), absent from the small trigger hairs of seedling leaves, which
measure only 0.15 to 0.2 mm. in length, of which the lever occupies
well over a half, the greatest width being 0.03 mm. {18 — 3). The
indented cells are surmounted by three layers of flattish cells under-

Chapter XII — 185 — Dionaea and Aldrovanda

lying the base of the lever, and stand upon a base of about four or
five courses of cells, meeting the general leaf surface. All the cells
have rather thick walls, and there are no intercellular spaces. This
whole basal region is the podium. Its histological character just
mentioned is such as to permit bending. It is, however, the whole
podium which bends, and not merely the cells under the constriction
0/ ~ 5)- I noted while cutting hand sections that the podium readily
stretches and compresses, bending being a combination of these. But
although the whole podium bends notably when the lever is much
displaced, it is quite clear on watching with the microscope that
sHght bending is evident first and at once in the constricted zone,
as GoEBEL recorded.

The hinge cells were thought by Macfarlane (1892) to be de-
void of a cuticle, or to have a very thin one. Haberlandt, however,
denied this. The cuticle is fairly thick and displays a certain amount
of wrinkling, which would allow freer movement of the coUench}^-
matous cell walls beneath. It further appears finely punctate, inter-
preted by Macfarlane as due to the presence of pores. Haberlandt
regarded this to be due rather to denticulations of the inner surface
of the cuticle which prevent loosening of the cuticle under repeated
bendings by anchoring it to the cellulose wall. It is easier to agree
with Haberlandt that the points seen (of which there is no doubt)
are not pores, than that they are extensions of the cuticle into the
cellulose underneath. Nevertheless, it is possible to see minute ir-
regularities on the inner face of the cuticle, so that the interface
between the cuticle and cellulose is greater than it would be other-
wise.

The medullary cells show some peculiarities. In addition to a
fime porosity (Goebel), Haberlandt records the presence of mi-
nute granular inclusions of high refringency in the middle layer
between the walls and between these and the hinge cells. He de-
scribes these as cutinized granules. Some cutinization certainly oc-
curs.

There is no vascular connection between the medullary cells of
Goebel and the leaf, since there is no vein in the hair.

The internal structure of the leaf blade or trap was described by
MuNK in 1876, concerned as he was with the direction of movement
of electrical currents, in much detail as to form and position of the
component cells. We recall that the trap has a massive midrib trav-
ersed longitudinally by a double vascular bundle which gives ofif
branches, running parallel to each other, towards the margins, ap-
proaching which they form a coarse zig-zag network. All the remain-
ing space between the two epiderms is occupied by a thin-walled
parenchyma, of large-sized cells inside and smaller against the epi-
derm, those against the inner epidermis being larger than those against
the outer, where they are much smaller and more numerous. There
are more smaller cells opposite the vascular bundles and more cells
of very large size between the vascular bundles {18 — 10). The
walls beneath the inner epidermis and the cells of the first parenchyma
course are thickened into a collenchyma {18 — 5) and to some ex-
tent also, those between this and the next course. This mechanical

Francis E. Lloyd —186— Carnivorous Plants

element is absent from the outer epidermis region. There is a total
absence of palisade tissue, a feature common to many carnivorous
plants, as Schmid showed. The parenchyma cells are elongated in
different degrees according to position. Below the chief vascular
strands in the midrib their long axes run lengthwise, above at right
angles to it and here are shorter. Those of the lateral regions run
up into the two lobes, and here they attain their greater longitu-
dinal dimensions, the largest, in the middle, being the longest. Ap-
proaching the margins, they become shorter and, as Munk points
out, are shortest, though not round, at the base of the ciha, length-
ening again in the cilia themselves. The intercellular spaces are
large and extensive, and while the protoplasm is very tenuous, chlor-
oplasts are present and much starch may occur, as Brown pointed

out.

Physiology. — If the question is asked how the structure just de-
scribed is related to the movement of the lobes, the answer is indicated
by comparing it with that of the lobes of very small traps found
in seedUngs, and with those of Aldrovanda. They all show the same
capacity for movement, whether the parenchyma consists of one
course of cells only {Aldrovanda) or of a few, as in seedlings traps.
The evidence indicates that the seat of movement resides in the
epidermis first of all. Further complication of structure is connected
with the mechanical strength of the lobes naturally greater in the
massive leaves of adult Dionaea plants. That is, the machine as
such is stronger (involving parenchyma cells) and can exert more
energy in the last, without any difference in the seat or directions
of movement. All movement occurs in the lobes and, as Brown
showed, none in the midrib, which is therefore not a hinge in any
sense. This is indicated in the diagram, borrowed from Ashida, shown
in i8 — 4a, though this does not indicate the extreme possibilities
of closure, better shown in iS — 46.

The stimulus leading to action. — In nature the walking of an insect
across a lobe of the trap almost inevitably results in the disturbance of
the sensitive hairs, ensuring the prompt closure of the trap. This
Curtis (1834) clearly observed. Ellis (1770) had thought the move-
ment as following irritation of the glands by the feet of the prey, and
Broussonet (1784) believed that it was due to the loss of turgidity
caused by the pricking of the surface by insects. But from the time
of Curtis it was supposed that it was necessary only to touch a hair
to bring about closure, until the work of Burdon-S Anderson demon-
strated the fact of summation of stimuh. Macfarlant: found inde-
pendently that under usual circumstances (temperature is important)
in order to effect closure of the trap, it requires two stimuh, either by
touching the same trigger hair twice, or any two different ones with an
interval of time neither too short (about 0.75 sec.) nor too long (oyer
20 sec). This was in 1892. Previous observers, with the exception
above noted, had failed to notice this behavior, very obvious when
once seen. For example, Darwin says: "It is sufficient to touch any
one of the six filaments to cause both lobes to close . . . ," but observed
that an extremely delicate stimulus might be inadequate. Darwin
does remark, however, that " on another occasion two or three touches

Chapter XII — 187 — Dionaea and Aldrovanda

of the same kind were necessary before any movement ensued," but
failed to indicate a general rule.

Following in the trail left by Burdon-S Anderson and Munk,
the method of electrical stimulation has been used by Brown and
Sharp (1910) to study time and intensity relations. They found
first of all that at 15 deg. C. two stimuli were always required, and
must be applied within an interval of from 1.5 to 20 seconds. But
at the higher temperature of 35 deg. C. frequently only one stimulus
was required, while at 40 deg. C. only one stimulus was required
in 50% of the instances. In order to elucidate this behavior, Brown
and Sharp tried the effect of electrical shock in various intensities
and found that the number of shocks required varied inversely with
their intensity. The authors then proceeded to determine the number
of stimuli (by bending the sensitive hairs) required when applied at
various intervals, viz., 20 seconds, i, 2, and 3 minutes, and found
that for these intervals, 2.0, 3.8, 6.2 and 8.7 stimuH (averages of
several tests) were required. It thus appears that "the number of
stimuU necessary for complete response varies almost directly with
the length of the intervals. It would seem, therefore, that the re-
sponse follows on a definite amount of accumulated efiect, "possibly
the accumulation of some chemical substance as the result of ex-
citation." It should be added that the physiological condition of
individual leaves has a modifying effect — at a given time and place
all leaves are not equally sensitive.

In 1873, stimulated by Darwin's studies, J. Burdon-S Anderson
published the first observations on the electric current in leaves as
indicating physiological disturbance, using those of Dionaea. Having
demonstrated that there is a normal current from base to apex of
the trap while there is one in the reverse sense in the petiole, related
quantitatively so that if the petiole were cut off at different lengths,
the current in the trap was increased, he then studied the effect of
stimulating the sensitive hairs on the current. Whenever a fly was
allowed to walk into the trap it disturbed the hairs and at once there
was observed a deflection of a galvanometer. If the stimulus were
repeated, the galvanometer indicator came to rest in a different position
(more to the left) each time. Disturbance of the hairs with a camel's
hair brush had the same effect. Thus the fact that movements of
the sensitive hairs constitute a stimulus was demonstrated by noting
consequent electrical disturbance.

Localization of perception. — It had been generally accepted since
OuDEMANs' time, in spite of Meyen's evidence to the contrary, that
StimuH leading to closure could be received only by the sensitive
hairs. Oudemans could not repeat Meyen's (1839) result, namely,
causing closure by scraping the midnerve. Darwin, however, found
both the area within the triangle formed by the sensitive hairs and
the surface along the midrib to be sensitive, so that when scratched
or pricked with a needle, closure followed. Macfarlane found that
this occurred on pinching the blade of the trap with steel forceps,
but that two stimuli were required. Brown and Sharp confirmed
Macfarlane's observation, only quaUfying the numerical expression
since they found that one only, or two or more pinches might be re-

Francis E. Lloyd — 188 — Carnivorous Plants

quired to procure reactions. They also found that the trap may be
stimulated to close by the apphcation of strong electrical stimuli to
the petiole. That various kinds of stimulating agents (cutting, hot
water at 65 deg. C, various chemicals, electrical stimuli) can effect
response has been abundantly shown by Darwin, Burdon-S Anderson,
MuNK, Macfarlane, Brown and Sharp. But it is easier to stim-
ulate by cutting, etc. the upper face of the leaf than the lower. If
the lower face is cut, it must be cut deeply so as to reach the upper
face tissues (Munk). Stimulation is not procurable by cutting the
outer marginal zone of the cilia (Munk). It appears therefore that
the more ventral tissues (Munk: the parenchyma) are sensitive, not
the more dorsal. But it remains the central fact that normally the
closure of the trap results from the stimulation of the sensitive hairs,
even though very slow closure may take place in response to applied
protein (chemonasty), after the power to react seismically has been lost.
It was natural for earlier observers, from Curtis on, to suppose the
whole of the hair to be sensitive, as did Darwin. "These filaments,
from their tips to their bases, are exquisitely sensitive to a momentary
touch" (1875). Sixteen years previously, however, Oudemans had
succeeded in showing experimentally that the sensitivity resides in
the basal region of the hair, to which Munk (1876) and Batalin
(1877) agreed. Darwin appears to have attached importance to the
flexible base in allowing the hair to bend rather than be broken by
the closing lobes. He saw that there is a constriction about the
base, merely mentioning it. Goebel (1891), in view of the config-
uration of the cells of the constriction (as well as of the organ in
general), believed that these receive, on movement of the "lever,"
a "much stronger stimulus than any other leaf cell." The stimulus
is hindered from moving upward by suberized cells in the two courses
above (in which I find very little if any suberization).

The cells of the constricted zone were regarded by Haberlandt
(1901) as special sense organs. The cells respond to compression,
but not to release from a constrained bent position (Brown and
Sharp), and if the hair is amputated, pressure on the remaining
base will procure closure (Brown and Sharp 19 10).

The mechanism of closure. — The first effort to explain the mecha-
nism of closure in the trap of Dionaea was made by Meyen, in 1839.
To him the spiral vessels of the nerves, because of their spiral-spring-
like structure, seem to afford a suitable mechanism. If the idea is
naive, it still indicates the early desire and effort to answer the ques-
tion.

Ziegenspeck much later (1925) {through von Guttenberg) at-
tributed importance to a hinge mechanism, saying that closure is
due to the loss of turgor by the cells of the tissues above the mid-
vein. Von Guttenberg (1925) showed this to be incorrect.

Darwin described the closure of the trap as passing through two
phases. There is first a sudden response, bringing the edges of the
lobes into some approximation, enough at least to bring the ciHa in
position so as to make a sort of cage preventing the escape of suf-
ficiently large prey, and allowing small ones (ants especially) to escape
(Jones) (// — 6). This is followed by a slow movement during

Chapter XII — 189 — Dionaea and Aldrovanda

which the lobes are closely apposed, pressing together, their mar-
ginal regions being curved outwards. Ashida in his studies of Al-
drovanda has called these the "shutting" and the "narrowing" phases
of closure. We are now to consider the first of these, to which more
attention has been paid than to the latter. Darwin investigated
the mechanism of closure by making marks on the upper surface of
a lobe in the transverse sense before stimulation, viewing the same
through a window cut in the opposite lobe. When closure had been
effected, the marks were found to be closer together and he concluded
from this that closure is accompanied by a transverse contraction
of the more superficial cells of the whole upper surface and sub-surface.
He thought also that the tissues above the midvein took part with a
hinge-like action.

This contraction was attributed by Munk (1876) to the loss of
turgor by more sensitive superficial tissues ("parenchyma") lying
beneath the upper epidermis, accompanied by the active expansion
of the tissues of the lower layers of parenchyma near the under epi-
dermis. De Candolle, from anatomical study, seems to have held
essentially the same view.

Burdon-Sanderson, seeking for a "resistance" which has to be
removed in responding to stimulation, could find it only in the turgor
of the leaf. "In the case of cells which are excitable the immediate
effect of excitation is suddenly to diminish the power (of turgescence)
and thereby produce a diminution of the volume of the cells which
is equal to that of the water (probably holding diffusible bodies in
solution) which is discharged into the intercellular spaces." It was
already known from the work of Bruecke, cited by Munk, that
the only mechanism of the actual movements of the sensitive plant
(Mimosa) was such a diminution of turgor in the sensitive region of
the pulvinus.

Batalin (1877) re-examined the matter and, confirming Darwin's
observations, extended the account to include subsequent opening.
Using the same method as Darwin, namely, measurements of changes
between ink marks during closure, Batalin came to agree with him
that there is a real contraction of the upper side of the lobes and a
concomitant expansion of the lower, both longitudinal and transverse.
He takes issue with Darw^in (and Ziegenspeck, 1925), however,
holding that the midvein takes no part, or at least a very small and
unobservable part. When the trap remains closed, as it does for a
week or ten days (or even longer) if it has been fed a Hving insect,
it enters at once into a second phase of movement. The lobes begin
to compress together mutually, so that in a half-hour (as I have
observed) much of their inner surfaces are in actual contact, leaving
however a space above the midvein. The compression is such that
the margins of the lobes are turned outwards and the cilia come to
lie more nearly parallel to the general plane of the lobes (iS — 4b).
As Batalin observes, the pressure exerted is enough to crush a soft-
bodied insect. Darwin thought that this compression is owing to
the absorption of animal matter. Batalin said that it is caused by
the reduction of the expansion of the lower surface, for he determined
that during the slow compression of the lobes together after the in-

Francis E. Lloyd — 190 — Carnivorous Plants

itial closure, there is actual and measurable shrinkage of that sur-
face, except where the body of the insect propped it out. AsnroA
(1934), we remember, compares this movement with the slow movement
which supervenes on "closure" in Aldrovanda, caUing it the "narrow-
ing" movement.

But the main contention of Batalin, in which he was in agree-
ment with Darwin, was that during closure there is an actual shrink-
age of the upper or inner surface, accompanied by expansion of the
lower. This amounts to saying that the tissues including the epidermis,
contract in the upper region, and expand in the lower. This was
made an issue by Brown (1916). Using again the same method,
he found extension in the lower surface and decreases in the upper.
But the latter amounts were very small, amounting to only 1.5%
of the original distances, while for the lower surface the differences
of dimension range between :^.s to 10, or an average of 6.7%. Fur-
thermore, and this is of prime importance, Brown found that there
is an error of observation due to changed surface curvatures, so that
the actual surface retains dimensions which it only seems to lose,
since what one measures is not the curved surface, but the chord
of its arc. Brown's opinion, based on the measurement of a model,
was that "if there is any change in the area of the upper surface
during closure it is probably in the direction of an increase rather
than in that of a decrease," in this squarely contravening previous
opinions. During subsequent opening, however, the reverse obtains.
The upper surface now expands (to the amount of 9.4% of the orig-
inal measurements) while the under surface maintains its enlargement
merely. True there is a small apparent expansion which is attrib-
uted by Brown to the same sort of error as that detected in the
measurements of shrinkage of the inner surface. It was shown also
that as the result of stimulation the growth of the lobes of the trap
was greater by a good deal than their growth during a long period
when there was no stimulation, from which it appears that stimu-
lation is a Hberator of growth and that, accordingly, the responses
to stimulation become less vigorous if the stimulus is repeated of-
ten. This recalls Batalin's experiments which showed that when
a trap was stimulated seven times on ten successive days, the abil-
ity to respond was not lost, but was progressively very materially
weakened.

What then takes place during the response movement? Macfar-
LANE allowed that the contraction observed by Darwin would be
due to the escape of water through pores in the protoplasm, and
sought for some visible evidence of such. He ventured to suggest
that appearances in the parenchyma cells of the motile tissue, con-
sisting of "rows of extremely minute globules or pores in the proto-
plasm," suggested a parallel with animal voluntary muscle, and that
on ultimate analysis the activity might be explained, as in the case of
plant cells, by water movements. This is not the same as saying
that "Macfarlane beheved that there are structures in the leaves . . .
which resemble animal muscles."

Batalin, not being able to detect any change in translucence of
the tissues, which would be expected if there were any effusion of

Chapter XII — 191 — Dionaea and Aldrovanda

sap into the intercellular spaces, such as is well known to take place
in the pulvini of Mimosa during movement response, denied that
there is any extrusion of water by the cells in the upper moiety of
the trap lobe, and Brown denying the contraction of the upper face
sees no necessity for such extrusion, but falls back on the expansion
of the cells of the lower face. This makes it necessary to find a move-
ment of water from a source sufficient for this expansion. The only
source considered is the parenchyma of the upper face, in which, since
the intercellular spaces are not flooded, water must pass from cell
to cell; the movement would then resemble that of geotropism.
An acknowledged difficulty is seen in the rapidity of the response
which, though often slow enough, is at times and normally so rapid
that complete closure is reached in the space of even less than a
half second (77 — 6). This difficulty must be faced as also that
arising from the attempt to account for the loss of water by some
cells (those of the upper face) by changes in the substances present
in the cells to less osmotically active ones, thus permitting the water
to be drawn off into other cells (those of the lower face) to facili-
tate their expansion. How sufficient water can thus be moved to
procure the recorded amount of expansion of the lower face, with-
out causing a reduction (contraction) of the upper face in even greater
amount (since the latter is shorter, if only slightly), is not clear.

This rapidity of movement seems to demand that there be a con-
dition of unstable equilibrium resulting from growth and residing in
the trap lobes. Batalin advanced this idea but he was, it is recalled,
committed to explain a shortening of the upper face. That tissue
tensions do exist may be taken for granted (Darwin), just as they
exist in the valves of the fruit of Impatiens. In this plant the ten-
sions are held in check by mechanical conditions, namely, the mu-
tual adherence of the valves. This disturbed, the valves spring away
by immediately curving in the same sense as the lobes of the Dionaea
trap. In the latter the lobes maintain their form, unless stimulated,
by the opposition of the two epiderms to their contiguous tissues.
When stimulated, the balance of forces is upset and curvature imme-
diately follows. That is, when stimulation takes place something
happens to release the tensions. What this something is we do not
yet know. If we might postulate chemical changes in the cell
contents from sugar to starch in the upper surface of the trap lobes,
it would serve us with a mechanism for changing the tensions, but
sufficiently rapid changes are not known. Brown's experiment in
which he substituted xylene, in which sugars are insoluble, for water,
are suggestive, but not convincing further than showing that ten-
sions exist which might be released by such a mechanism. This is
in essence the theory put forth by voN Guttenberg (1925) who be-
lieves that the movement is caused, not by any reduction of turgor
in any tissues whatever, but by the drag of the parenchyma on the
two epiderms (upper and lower), the upper being thicker and less
extensible than the lower {iS — 5, 6, 10), as Macfarlane also main-
tained. That this drag is positive, exerting a pull on the epiderms,
is indicated by the fact that if the tissues of the upper epidermis
and the contiguous parenchyma are partially robbed of water by

Francis E. Lloyd — 192 — Carnivorous Plants

the application of a plasmolyte to the upper surface, no closure can
take place on stimulation (application of the plasmolyte to the base
of the sensitive hairs was avoided). Indeed, any experiment in which
the parenchyma is robbed of its turgor renders the valves incapable
of closure. This positive drag therefore, present before closure,
stretches the epiderms as much as it, previous to excitation, is ca-
pable. On stimulation this capacity is increased, the epiderms re-
sponding by expanding differentially, the upper scarcely at all, the
lower 6-7%, in accordance also with measurements by Brov^n and
others. Von Guttenberg then faces the questions, whence the water
necessary to increase the volume of the parenchyma cells, and what
conditions allow the momentary increase of water uptake? To the
former he suggests that the water comes from the vascular tissues;
to the latter that it may be due to the sudden changes of substances
in the sap from a large molecular to a small molecular condition.
Von Guttenberg extends this theory to the case of Aldrovanda,
making the pertinent observation that, in view of the fact that this
trap has only a single course of parenchyma cells, it is unthinkable
that there exists a differential action in the tissues between the epi-
derms, of which more beyond.

Von Guttenberg's difficulties may, however, on theoretical grounds
be avoided. If it be assumed that the response to irritability is con-
fined to the epidermis, we might argue that this response consists
only in the reduction of turgor. True, as von Guttenberg says,
this would be removing one factor in tissue stretching, but as turgor
expands the cells in every direction, the relative amount of exten-
sion depending on the lengths of the walls, its removal would allow
the application of the energy of the turgid parenchyma to the flaccid
epidermal cells, the longitudinal walls of which then would respond
readily to the stretching effort, the amount of stretching depending
only on the physical properties of the walls. In a word, the system
would work like a bimetalhc strip of metals of different indices of
expansion, von Guttenberg's idea, but demanding simply loss of tur-
gor in the epidermis only, and this, as von Guttenberg observes in
regard to Ziegenspeck's theory, is easier physiologically than a rise
in turgor in a mass of tissue. It should be added that the loss of
turgor by the epidermis need not advance beyond an initial stage
of relaxation, just sufficient to allow, without evident effusion of water,
the stretching of the longer walls, which would otherwise be pushed
out laterally by conditions of turgor. Thus the theoretical neces-
sities are reduced to a minimum, and the movement of Dionaea
brought into fine with movements in general in sensitive plants.
As to the bursting of fruits such as Impatiens, Sicyos, etc. we have
to do with change of shape of parenchyma cells without change in
turgor. In Impatiens the two epiderms are of unequal extensibility.

This seems to be the view advanced by Ashida (1934) which,
prompted by his study of Aldrovanda, he applies "by deduction to
the case of Dionaea." He cites Macfarlane's observations that
the lower epidermis has a thinner cuticle than the upper, and is there-
fore more easily distensible, permitting curvature on the relaxation of
the upper epidermis with the effect of closure.

Chapter XII — 193 — Dionaea and Aldrovanda

What is the nature of stimulation is certainly not known. Haber-
LANDT regarded the constricted cells at the base of a sensitive hair
as sense organs which are activated by compression. Brown found
that the sensitive hairs do not respond to decompression procured
by two successive movements of a hair which had previously been
kept in an extreme bent position. Propagation of the stimulus cannot
be dependent upon the vascular tissues, since they are absent from
the trigger hairs; and in Aldrovanda, which has no vascular tissues
except the single strand along the midvein, it is even more obvious that
the path of movement must be found in the parenchyma, but whether
of the epidermis alone or of the internal tissues also, is not yet known.

That response, an event following on stimulation, is accompanied
by electrical disturbances Burdon-Sanderson showed, and the char-
acter of these permitted him to Hken them to those which occur
during muscular contraction, though this is not the same as iden-
tifying the contraction, asserted by Darwin and Batalin, of the
upper surface with muscular contraction (F. Darwin, 1875), espe-
cially when now such contraction has been questioned (Broavn).
The molecular transposition measured by Burdon-Sanderson might
indeed be the expression of sap movements, and such sap movements
need not be great quantitatively to upset an equilibrium and might
constitute a trigger action to start the mechanism a-going.

Whatever the tensions in the open trap lobe may be, they must
be duplicated in the similar trap of Aldrovanda, and when we look
at this beyond, it is a help to comprehend what happens in Dionaea
when such tensions are reUeved.

For Dionaea we may at present say: —

i) During the open condition there are tensions present which
are so distributed that they maintain the trap in an open position,
the lobes standing at an angle as great as 80 deg. (Darwin).

2) When stimulated the lobes close, the ciHa becoming interlaced
like the fingers of clasped hands. The lobes remain concavo-convex,
inclosing a wide space between them. During this closure the outer
face of the lobe expands, the inner remains unaltered, or at least
it does not contract. If the stimulus is prolonged by chemical stim-
ulation (as when an insect has been introduced), the lobes continue
toward a greater mutual compression and thus obliterate to some
measure the inclosed space (the "narrowing" of Ashida). Batalin
thought that this is due to a subsequent contraction of the lower
face of this lobe. It might be due to a passive extension of the upper
face resulting from rapid exudation of secretion, depleting the tissues
of water. The edges of the lobes, which do not actively participate
in the movements, become bent outwards and the cilia now extend
less transversely, so that the two sets become more nearly parallel.

3) With the cessation of secretion and absorption, the lobes re-
open, this being the result of increased growth of the upper faces,
the expansion of the lower faces being maintained.

The mechanically stimulated trap closes, and reopens without nar-
rowing in about 24 hours, when it will respond again. But repeated
daily responses are followed by decreasing sensitivity, probably due to
the completion of growth.

Francis E. Lloyd — 194 — Carnivorous Plants

If closure follows trapping of suitable prey, narrowing (in the
sense of Ashida) takes place. Reopening follows at the end of a
period of days (5-10 or more) when there is evident a diminution of
sensitivity, which however is regained in the course of time (some
days, OuDEMANs).

Digestion. — During all this digestion and absorption have been
taking place. Darwin did a variety of experiments with various
substances. We have seen that the upper surface of the lobes is
crowded with many glands capable of secretion and absorption, as
Darwin stated. These glands remain passive unless some suitable
material (insect, meat, etc.) is inclosed between the lobes. Then
there is a copious secretion of a fluid which has the power of diges-
tion, and which causes the dissolution of the substances acted upon.
"It is so copious that on one occasion, when a leaf was cut open,
on which a small cube of albumen had been placed 48 hours before,
drops rolled off the leaf" (Darwin). The secretion is acid, the pres-
ence of formic acid (Balfour) serving also for the inhibition of bac-
teria, so that, unless too great ''portions" have been supplied, there
is no odor of decay. Balfour found that a strip of meat placed partly
within the closed valves and partly out, showed no bacterial action
within, but did so without. When the rotted portion was placed in
a fresh leaf, the odor of decay disappeared. This contravened the
opinion of Rees and Will, whose experiment seems to have been done
with abnormal plants (Goebel). There seems therefore to be no
doubt of the digestive power of the secretion, though no in vitro
experiments have been done with Dionaea secretion. According to
Darwin some substances are not digested (fats, iibro-elastic cartilage).

If not too great masses of material have been fed, when the trap
begins to open the interior is found to be dry, and the fluid has been
entirely absorbed. Experiments to show the usefulness of the ab-
sorption of proteins, such as those carried out by various authors on
Drosera, Utricularia, have not been done. Our opinion on that must
therefore rest on evident analogy.

Fraustadt thought that during the period when the trap is closed
over an insect photosynthesis stops. But Pfeffer (1877) suggested
that the lowering of the starch content observed by Fraustadt may
be the accompaniment of a change in metabolism while at the same
time photosynthesis may be proceeding. The work of Kostytschew,
if meagre, seems to deny Fraustadt' s belief (see under Drosera,
p. 119).

Aldrovanda vesiculosa L. : — Aldrovanda is a small fresh water plant
{ly — 5) growing in quiet waters, floating just below the surface.
It is quite rootless, and consists of a slender stem, clothed with whorls
of leaves not distantly separated. Each whorl has eight leaves mu-
tually attached at their bases. It branches infrequently, so that usu-
ally one finds only a single stem. The whole plant reaches a length
of 10 to 15 cm. with a width of 2 cm. including the spread leaves
which are refiexed in age. The tip of the shoot is especially con-
spicuous by the numerous bristles which jut beyond the general leaf
profile. The flowers are supported on short stalks, bringing them

Chapter XII — 195 — Dionaea and Aldrovanda

just beyond the extent of the bristles. They measure about 8 mm.
when widely open. The seeds are ovate, clothed with a hard shell
(KORZSCHINSKI 1886).

This unique plant was first seen in India and was cited in 1696
by Plukenet as '' Lenticula palustris Indica" in his Almagestum
Botanicum or Phytographia (4: 211, pi. 41, fig. 6). In 1747 Gaetano
Monti had received a collection of it made by an Italian physician, Dr.
Carlo Amadei, in the DulioH Swamp, east from Bologna. It was
named Aldrovandia by Monti in honor of the Italian naturalist,
Ulisse Aldrovandi, who died in 1605. This plant was identified
by J. J. Dillon with the Plukenet one from India. In 1751 it was
mentioned in a dissertation by L. J. Chenon (1751), a student of
Linnaeus, as Aldrovanda (probably a mistake in copying, thinks
DuvAL-Jou\TE, 1861) and finally published byLiNN.\EUS in the Species
Plantarum 1753, p. 281, as Aldrovanda vesiculosa.

Caspary points out on high philological authority that the Lin-
naean name is ungrammatical. The name Aldrovanda is now generally
accepted in accordance with the International Rules of Botanical
Nomenclature. Another plant from India was described as the species
verticillata by Roxburgh {Flora Indica 1832, 2: p. 113), but this
was shown by T. Thomson not to be distinct, but has been regarded
as a variety. A plant from Queensland, Australia, once called the
var. auslralis, is not distinguishable from the original species, though
Darwin found some difference in size, together with other minor ones,
such as the number of serrations on the bristles.

Aldrovanda vesiculosa ranges from S. France to Japan, south to
Austraha, and in Africa to the southern tropics where it was found
by Miss E. L. Stephens in the Chobe Swamp, ico miles west of
Victoria Falls. This material, together with living plants, has been
studied by me, the latter having been obtained in Silesia and grown
during the summer of 1933 in the Garden of the Botanical Institute
of Munich. Beautiful herbarium specimens in all stages of fruiting
and flowering from Mizoro Pond, near Kyoto, were sent me by Dr.

JOJI ASHIDA.

The morphology and anatomy of the vegetative parts of the
plant were first described by Cohn in 1850, and more completely
by Caspary in 1859 and 1862. Further reference to details was
made by Goebel (1891), Fenner (1904), and Haberlandt (1901).

Like the leaves of Dionaea, those of Aldrovanda consist of a flat-
tened petiole armed at its apex. This appears somewhat truncated,
with four to six, or seldom even eight parenchymatous lobe-Hke bris-
tles, surmounted by a nearly circular leaf blade, 4 mm. wide. When
mature the petiole is wedge-shaped, broader at the apex, 6 mm. long
and 4 mm. wide. The bristles extend another 5 mm. The midrib
of the petiole with its vascular tissue continues into the blade, which
has the form of a steel trap, as in Dionaea.

Seedling {ig — 1-5). — The elliptical seed has a snout at one end,
plugged with a cap, under which lies the root end of the short hypo-
cotyl. Surmounting this are the two broadly conical cotyledons
pressing against the large endosperm, much as in Dionaea (Smith).
In early germination the hypocotyl protrudes, pushing off the cap

Francis E. Lloyd —196— Carnivorous Plants

and carrying it forward for some time, till indeed the hypocotyl reaches
its fullness of development, with a length of 3 mm. (79 — 1-3). By
this time the petioles of the cotyledons have emerged, and, just above
the plumule, expand to form a sack-like expansion surrounding it.
Above this they are suddenly constricted, the isthmus entering the
seed and connecting with the expanded ends of the cotyledons which
form a haustorium. The developing plumule breaks out of one
side of the surrounding cotyledonary envelope and progresses toward
forming the plant. The leaves of the first whorl are slender ligulate
and taper to a fine point, or may be variously laciniate to some de-
gree. There are usually five in the whorl. The next whorl, raised
on an evident internode, shows still more laciniations, but does not
yet produce traps (79-5). These, however, usually appear in the
fourth whorl. Subsequently the mature condition is gradually estab-
lished. The hypocotyl ends without forming a root cap, and initial
cells appear never to be established after the primary condition has
passed {iq — 4). This, Korzschinski, who described the course of
germination, did not see, and this lack was indicated by Goebel.
The structure of the seed and seedHng in its primary condition is
quite similar to that of Dionaea, as described by Smith (193 i), dif-
fering however in a few details, notably in the greater expansion of
the cotyledonary petioles to embrace the plumule, and in the failure
of root growth.

The leaf of maturity. — This consists of a wedge-shaped petiole
(regarded by Nitschke as the leaf base {fide Troll, 1939) and the nar-
row isthmus between it and the trap as the petiole, a view now re-
garded as untenable), somewhat truncated at the apex, where it bears
four to six, occasionally more (eight, Caspary) serrate bristles, and at
its middle point a leaf blade in the form of a trap.

When four bristles only are present they appear to stand two
on each side of the trap, but the inner two, as revealed during de-
velopment, stand somewhat behind the insertion of the trap, and
overlap it {19 — 6). If a fifth occurs, this quite evidently stands
behind the trap, and therefore does not, as Caspary noted, arise from
the end of the petiole, but from its dorsal surface. The bristles cannot
therefore be regarded as lobes of the leaf, as Cohn thought, nor
as stipular appendages (Nitschke, 1861), but rather as emergences.

The structure of the petiole {19 — 9) in general is that of water
plants; there are wide intercellular chambers of pentagonal, hexagonal
(along the midrib), or elongated form (along the margins), separated
by partitions one cell thick. Fenner, who has more than anyone
else described the minutiae of the plant's structures and their de-
velopment, errs in showing large hexagonal chambers over the midrib,
and in fact the figure of his transverse section does not consist with
that of the leaf en face, the former being correct. The epidermis is
scantily clothed with two-armed trichomes {19 — 16), standing on
two very short stalk cells with cutinized walls, these in turn on two
epidermal basal cells. The arms of these hairs may be short or, es-
pecially along the margins, twice as long. The bristles taper grad-
ually from their broader bases and are serrated irregularly by
projecting unicellular trichomes, ending in a similar spinous one.

Chapter XII — 197 — Dionaea and Aldrovanda

The trap stands at the apex of the petiole, the midrib, carrying
a single annular vessel with an ample phloem, being continuous from
one to the other. But it always stands asymmetrically, resulting
from a twist in the stalk, in such a manner that the mutually appressed
lobes in the young trap are turned with their free margins to the
left (as viewed from above) through an angle of about 90 degrees.
In addition to this torsion, the trap is bent backwards (i.e., to the
right as seen from above) through an angle of 30 to 40 degrees (77 —
5). Monti evidently refers to this posture when he said, "In bar-
bularum medio folliculus oblique appenditur." It is thus brought
about that, when the traps are open, their openings face outwardly
away from the stem, instead of tangentially. The course of devel-
opment of the leaf is here worth a glance {ig — 6). In its earliest
stage, the whole leaf consists of a mere conical protuberance from
the stem apex. Soon it becomes apparent that the basal half is broad-
ening to form the fiat petiole, while the now more cyhndrical end
is to become the trap. Very soon this begins to show torsion which
progresses until, when the leaf is approaching maturity, the trap
comes to lie in its definitive position. In the meantime the bristles
have developed, first the outer followed by the inner. At an early
stage it can be clearly seen, as it was by Caspary, that the trap and
inner bristles do not he in the same plane. In maturity, the bristles
project much beyond the trap and so produce the bristly appearance
of the plant.

The position of the mature traps resulting from torsion and bend-
ing may be regarded as a distinct adaptation, since their mouths,
when open, are all placed so as to avoid obstruction from neighboring
leaves, which in view of their numbers and crowding, is obviously
advantageous for the easy approach of prey.

In describing the action of the trap, whose pecuharities of posture,
much less pronounced but present in Dionaea, have just been de-
scribed, a special terminology is required, proposed by Ashida. It
is clear that if the trap is twisted 90 degrees to the left (in the sense
above indicated) the one side or lobe of the trap must come to lie
against the bristles {ig — 5). This Ashida calls the bristle-side lobe.
The other lobe is the free-side lobe. The importance of this distinc-
tion lies in the fact that both lobes are concavo-convex and lie dished
the one into the other (ig — 7). That is, the outer surface of the
bristle-side lobe and the inner surface of the free-side lobe are convex,
the other two concave. Since this has been brought about in the
course of development, the two lobes acquire a different set, the
effect of which will be clear when the action of the trap is described.
The trap has a unique structure, which we shall now describe.
Morphologically it is a leaf blade, each half being nearly semi-circular,
the circle being subjected to some degree of skewing. Each, of course, is
attached to the midrib, which is the thickest portion. Fenner,
CzAjA and others have called the midrib the "hinge," but as the
proximal parts of the lobes do not move at any time, this is a mis-
nomer. It is true that textbook figures taken from earlier authors
would indicate the contrary, but they are certainly wrong, as Ashida
has clearly shown.

Francis E. Lloyd — 198 — Carnivorous Plants

Each lobe, when the trap is mature and is in the set posture
(ready to catch prey), is concavo-convex from within out. But the
curvatures are not simple spherical ones. It will be seen by the fig-
ures herewith that two oval zonal regions, one on either side of the
midrib, are fiat (ig — 8, 21-23) and are subject only to slight cur-
vatures under stress during the closure of the trap. From this flat
middle region, the lobes spread with a maximum curvature along the
transverse middle line. Here the curvature is much like that of the
ribs of a vessel amidships. This is the principal region of motion
during closure (2, ig — 22-23). The next region is one of compara-
tively Httle inward curvature (5, 19 — 23) as far as the margin, which
is sharply bent back inwardly to form a valve edged with a row of
sharp teeth. The whole looks hke a widely opened clam or mussel
(ig — 21a). This is what one sees looking merely at the outer form.
When the thickness of the lobes is examined, the following is found.
The inner half, along the midrib, is thick and relatively rigid. This
half (the thick region) includes the place of greatest curvature, to-
gether with a measure beyond (again as seen in a transverse section
normal to the midrib at its middle point) (1-3, ig — 23). Anatom-
ically it consists of three courses of cells, the two slender celled epi-
derms which are thin, enclosing a single course of very large
thin-walled cells, the long axes of which run transversely the leaf
(ig — 10, 11). This structure is continued around the sharp bend
of the motile region, which is somewhat thinner than elsewhere, into
the sides of the trap somewhat less than half-way to the free mar-
gin. At this point the lobe suddenly thins, the middle course of large
cells ceasing. The lobe then consists of a very thin membrane con-
sisting of only the two epiderms juxtaposed, and so it continues
quite to the inturned margin which forms the valve. The valve
itself is thicker again, due to the enlargement of the epidermal cells,
giving it a useful firmness to make it eftective. Since only the mid-
dle transverse structure as seen in section has been examined, some
details concerning the curvature of the lobe margins must be men-'
tioned.

The stiff region of the trap wall along the midrib does not extend
its full length, so that, beyond certain points, the proximal and distal
parts of the walls are thin and, when the trap is closed, readily ap-
proach each other so as to lie juxtaposed. As the marginal valve
does not reach the midrib — it becomes narrower as it approaches
it and quite ceases 0.75 mm. away — there are left two spaces, one
at the apex and one at the base of the blade, which, when the trap
is closed, can allow the escape of water, while elsewhere the valve
acts to prevent the escape of prey during the whole course of closure
of the trap. This escape of water is necessary to permit the two
thin regions of the lobes to approach and to become mutually ap-
pressed. This is possible because the thin region of the free-side
lobe inclines to bend when pressed against the bristle-side lobe,
due to its set acquired during development, so that when the
trap is fully closed, the thin regions of the two lobes dish into one
another as during development {ig — 21, 22), crowding the prey, if
caught, into the digestion cavity. Before the act of closure is looked

Chapter XII — 199 — Dionaea and Aldrovanda

into more carefully, further details of anatomy will be examined.
The thick region, as above said, is composed of three cell layers, the
two thin epiderms sandwiching a middle course of large thin-walled
cells of cyhndrical form. In these three courses the cells are elon-
gated at right angles to the midrib, and have straight walls, excepting
that the outer epidermis beyond the motile zone, to be delimited
later, has wavy-walled cells. In passing over into the thin region
of the lobe, the middle course of cells ceases entirely, so that there
remain only the two epiderms juxtaposed {ig — 19). In the inner
zone of this region the cells are elongated and have straight walls,
but there is a gradual transition to irregularity, when the walls be-
come wavy. The thinnest part of the valve is toward the outer
edge, where it is reduced to 0.5 mm. and is here only one cell in
thickness. This is accounted for by the fact that, as Cohn and
Caspary observed, the cells of the two epiderms become mutually in-
tercalated, the cells of the inner course protruding between those of the
outer course to occupy part of the general outer surface and vice versa
{ig — 12, 13, 17). The margin of the thin region is reflexed to form
the valve and has a greater thickness, namely about i mm. Along
the edge of the valve stands a row of sharp, stiff, unicellular hairs
which, when the trap is closed, intercross to prevent any escape
of prey between the valves (Cohn) {ig — 18), recalling the analogous
arrangement in Dionaea, but in a reverse sense. This the trichomes
accomplish more by numbers than by strength, which is indeed not
great, as Darwin observed. His doubt on the usefulness of the de-
vice is, however, scarcely justified.

The cells of the two epiderms of the thick regions differ in size.
The outer epidermis per unit of measurement is composed of more
and therefore more slender cells, than the inner, in the ratio of about
7:5. The cells of the middle course are longest near but not next
the midrib. The structure of this is seen in ig — 10, 11, 23.

The inner and outer surfaces of the trap are supplied with a variety
of trichomes with various functions. On the outer surface there are
squat, two-armed hairs {ig — 16) similar to those found over the
general plant surface. Their capital cells are devoid of cuticle, and
they secrete mucilage. On the inner face of the lobes are to be found
three kinds of hairs. On the surface of the distal zone of the thin
portion there are four-armed hairs resembling superficially the "quad-
rifid hairs" (so called by Darwin) in the interior of the traps of many
species of Utricularia {ig — 15). Aside from having four arms, which
lie prostrate against the surface of the leaf, they are otherwise of
the same structure as the two-armed hairs, and like them are devoid
of cuticle, and secrete mucilage (Goebel). They are distributed in
a broad zone lying adjacent to the valve (ig — ^8). The innermore
region is devoid of them. On the inner surface of the thicker region
of the lobe occur bun-shaped glands which may be regarded as di-
gestive and absorptive in function {ig — 14). They arise from two
(Fenner) epidermal cells, on which stands a short stalk of four cells
which expand into balloon shaped upper ends, clothed with a dozen
or more cells to form the capital. They too lie in a zone of much
density toward the outer margin of the thick region, and are few

Francis E. Lloyd — 200 — Carnivorous Plants

and more scattered nearer the midrib, on which, however, there is a
dense row of them. Darwin regarded the "quadrifid" or cruciform
hairs within the trap as absorptive, but Duval-Jouve (1876), because
of their occurrence on the outer surfaces of the petiole, etc., considered
them as of identical nature with the latter.

In this region, to which, thought Mori (1876), irritability was con-
fined, there are also about 40 (20 on each lobe) long, very slender hairs,
described by Goebel, Haberlandt and Fenner, analogous to the
normally six sensitive bristles which occur on the lobes of the Dionaea
trap. In Aldrovanda, however, they are of a much simpler though
equally effective structure. They are about 1.3 to 1.5 mm. long and
0.05 mm. thick except at the base, where they are a bit wider. ^ They
are very slender shafts, arising from a four-celled base lying in the
epiderm, and projecting slightly therefrom. On these is surmounted a
length of two courses, each of four long, slender cells. These bear the
super-sensitive cells, four in number, though sometimes there appear to
be only two. Haberlandt does not state the number. They are
short, thin-walled and form a sort of joint or hinge where the otherwise
stiff hair can bend sharply, thereby compressing the cells on the con-
cave side (Goebel, 1891). Above there are two courses of slender cells
of two each, gradually tapering to a sharp, sometimes forked, end
(ig — 20). They are arranged and postured in such fashion that,
contrary to the impression given by some authors, they are not bent on
the closure of the trap. They stand upright on the flat region of the
trap on either side of the midrib, where they have plenty of head-room
when the trap is fully closed, but obliquely on the sides so that, though
long enough to reach beyond the fully closed digestion chamber, they
lie sandwiched between the thin regions when approximated without
being bent. Disturbing these hairs results in the closure of the trap,
one touch of a bristle of a young leaf sufficing, but as the leaf grows
older two or even many more become necessary. Quite old leaves,
appearing at first to be beyond response, showed action when a lot of
the sensitive hairs were disturbed by a sweeping motion of a needle a
considerable number of times.

But though I did 300 experiments I found it diihcult to make a
very definite rule. There is a good deal of difficulty, of course, in
getting a clear-cut result when one is dealing with so small an object
as the Aldrovanda trap which has so many deHcate bristles close to-
gether. In cases where the results were quite clear-cut, the data
were contradictory. Thus in one case a young trap responded to
one touch of a single hair while another one, of similar age and ap-
parently ready for action, being widely open, required seven stimuh
appKed to a single bristle. Another required even more, caused by
a sweeping of a number of hairs after six single stimuH. In some-
what older leaves, two stimuh only were frequently required to effect
closure, but this also was by no means constant. Older traps behaved
often in a singularly refractive fashion, but yet were found to respond
at last. One case only: a single inner (on the fiat region) bristle
was bent 10 times; a second was bent 10 times; several bristles
were then bent by a sweeping motion 10 times; then several outer
bristles were swept ten times, and finally a single inner bristle was

Chapter XII — 201 — Dionaea and Aldrovanda

bent twice, followed by the closure of the trap. On the other hand
an old trap closed with one stimulus only, seen by Dr. E. Merl
and myself, as we were working jointly at the time. Another dis-
tinctly old trap responded to the eleventh stimulus, ten on one hair,
the eleventh on another. Many did not respond at all. Ashida
made quite similar observations. De Lassus (i86i) had already
observed that young traps are somewhat more sensitive than older
ones. It became apparent that this lack of uniformity, while a fact,
does not mean lack of dependability of the trap in nature, since
the prey which ventures into an open trap must needs stimulate
many hairs many times if it moves about. If the trap closes par-
tially (see below) so that the prey cannot escape, the continued move-
ments insure a further stimulation, and complete closure is assured.

The mode and mechanics of closure may now claim our attention.
We have seen that each lobe displays two concentric regions, an inner
thicker, and an outer very thin and pliable, and edged with a valve.
If a relatively weak stimulus is applied, the lobes close till their
free edges meet. Unless additional stimulus is added, in the course
of a short time (20 to 30 min.) the lobes begin to open, and shortly
resume their original postures, at some 45 or 50 degrees from each
other. If, however, a sufificiently strong stimulus, or repeated stimuli
be used, the lobes continue to close still further. This is possible
because the free-side lobe flexes under pressure against the bristle-
side lobe, at first just inside the valvular edge, the flexure extending
until most of the two regions are mutually appressed {ig — 21).
The two marginal valves become bent under this mutual pressure,
the teeth intercrossing so as to prevent prey from escaping when
the lobes are first closed. Resulting from the whole movement, the
thick regions have moved together and a space has been inclosed by
the meeting of their outer Hmits, forming a smaller but more ines-
capable prison {ig — 22). Here the digestive glands begin their
work of digestion, and in the course of time the prey is disintegrated
and the products absorbed. If a plant is Hfted out of the water,
the water films stimulate the traps to closure, and in closing, air
is entrapped. The idea that the traps were hollow, closed organs,
held by Monti, led him to use the descriptive name ''vesiculosa.'''
CoHN (1850) and de Lassus (1861) found this to be a mistake.

The mechanism of movement. — The sensitivity of the trap was
first observed by Auge de Lassus, who was cognizant of the facts
regarding Dionaea and Drosera in 1861. The fact was rediscovered
by B. Stein in 1873 (mentioned by Cohn in 1875) who found that
it is the slender hairs which are capable of receiving stimulus, and
recognized the analogy in this detail with Dionaea. Additional con-
firmation was offered by Mori (1876). Goebel showed more com-
pletely this analogy by demonstrating the hinge of the sensitive
hair. Czaja (1924) studied the effect of various kinds of stimula-
tion. He incorrectly regarded the midvein as a hinge about which
the valves rotate to approach each other in closure.

It has remained for Joji Ashida to make a studious attempt to
elucidate the mechanism of response, following that of Brown and
Sharp for Dionaea. Ashida first made clear where the exact re-

Francis E. Lloyd — 202 — Carnivorous Plants

gion of active bending is. To determine this, he devised a method
of imbedding the open trap in agar jelly, transferring it from warm
water to still fluid agar at the same temperature, low enough to do
no harm. On setting, the agar with its imbedded trap could be cut.
It was noticed that on cutting the leaf it would react, and in doing
so, it would withdraw from the agar on the outside, so indicating
the zone of maximum bending. This was found, as already shown
above, to be in the flanks of the thick region, between the flat part
next the midrib and the outer rib-like region (2, ig — 23). How
is this movement accomplished? In a complete response the amount
of movement is sufficient to bring the edges of the thick regions
in mutual apposition, thus inclosing an ellipsoidal shut-off space.
Meanwhile, as already said, the thin regions dish the one into the
other as the result of mutual pressure brought about by the thick
regions.

Precisely what happens to procure the bending is more obscure.
An observation made by Ashida is, if substantiated, of prime im-
portance. It is that the outer epidermis of the motile zone, when
in the state of open rest, is undulated, and in this condition not
in a state of extension, whereas the inner epiderm is, if not fully ex-
tended, at least more so than the outer, since very hght, if any, un-
dulation is to be seen. During closure the undulations disappear,
due to stretching of the tissues. As I have already suggested (1933),
the two epiderms act after the fashion of a bi-metallic spring. As-
suming this to be the case, two questions arise. What condition
of the tissues operates to keep the outer epidermis lax? And what
happens to procure the changes from the lax to the taut condition?

In addition to this undulation of the outer epidermis, the motile
region is thinner than the non-motile parts of the thick region. And
if the opposing lobes are cut away so as to exclude their mutual
pressure during closure, it is ascertained that the lobes can curve
far more than they do otherwise, as is the case in Dionaea; and fur-
ther, that the free-side lobe bends, during closure, more than the
bristle-side lobe. Ashida has also demonstrated to his own satis-
faction that the outer epidermal walls are the more easily extensible,
the outer subepidermal walls less easily, while the two inner walls
are least extensible. This conclusion is regarded as flowing from the
observation that, if a trap is plunged in acetone or alcohol, under
the internal pressures induced by the entrance of these fluids into
the cells, vesicles arise, but only on the outer face of the motile zone.
The vesiculation is caused by the rising of the cuticle and the breaking
of the radial walls of the epidermis. Evidently the outer epidermal
walls are readily extensible, but, since they do not retract when the
vesicles are reduced, they are thrown into folds. Ashida argues that
the walls are plastically, not elastically, extended. That the motile
zone is weaker than the lobe is elsewhere was shown by a tearing test,
the result being that the lobes always tear at the motile zone. Again,
from observing the movements of the intercellular air on the en-
trance of alcohol, the inference was drawn that the walls in the motile
region are more readily penetrated than elsewhere.

Before discussing the mode of operation of the motile mechanism

Chapter XII — 203 — Dionaea and Aldrovanda

in Aldrovanda, for the purposes of comparing it with that of Dionaea
(there is a definite analogy between them), Ashida draws the following
parallel between them. The motile zone of Aldrovanda is but three
cells in thickness; that of Dionaea composed of several to many cells
in thickness, according to age of plant.

Aldrovanda Dionaea

The outer epidermis
The outer epidermis The mass of parenchyma beneath

(either one or both)

The middle course of parenchyma The parenchyma as a whole

The inner epidermis
The inner epidermis The parenchyma beneath

(either one or both)

The actual leaf movement embraces two phases of motion which
Ashida calls (a) the shutting movement, to the "shut" stage, when
the rims of the lobes just meet; {19 — 21b) and (b) the narrowing
movement leading to the "narrowed stage" {ig — 2id, 22). In re-
covery the opening passes through the "rebulging movement" from
the narrowed stage to the merely shut stage and the reopening move-
ment, completing the opening. To avoid confusion these terms will

be used.

In the shutting movement the margins of the thick region ap-
proach sufficiently to bring the margins of the thin regions together.
This follows on the application of a weak stimulus, but proceeds no
further. In time reopening occurs. If, however, the stimulus is suf-
ficiently strong, this posture is passed through, the edges of the thick
region approach mutually still further, the thin regions press on each
other mutually, and the free-side lobe buckles, dishing itself in against
the more rigid bristle-side lobe. Ashida maintains that these two
movements are not simply a continuation the one of the other, as will
be seen.

The rapid shutting movement is caused by the loss of turgor by
the inner epidermis. This allows the other two layers to expand and
the curvature ensues mechanically. The membranes of the outer epi-
dermis are stretched irreversibly. The undulations described by
Ashida disappear, having previously been maintained by the outward
pressure of the inner epidermis. The posture thus attained now
changes to that of the narrowed posture by the narrowing movement.
This is a slow movement, accomplished by the slow elongation of
the outer epidermal cells, that is, by growth. Resulting is the mutual
appression of the two thin regions, during which water must escape
from the inclosed space. Ashida tried to determine by means of
colored fluid where the water escapes but did not get any very convinc-
ing evidence. It has been suggested above that the escape is between
the non-valvular parts of the lips at the forward and rear ends of
the margins. Ashida, by means of ingenious optical apparatus, was
able to record photographically the advance of the whole movement.
The shutting movement is very rapid, occupying about one fiftieth of
a second, following on a latent period of 0.09 seconds. This rapid
movement involves the expulsion from the trap of water, the pres-

Francis E. Lloyd — 204 — Carnivorous Plants

sure of which must be overcome. As in the case of Utricularia, the
energy expended is sufficient to cause a trap lying free in the water
to close with a sudden jerk, displacing it, just as a Peclen swims.
There seems to be a slight difference in the behavior of the two lobes,
the free-side lobe moving a bit more rapidly than the other. The
difference is very httle, however.

Advance from the shut to the narrowed posture is slower, more
irregular, complicated by conditions. The slowness depends in the
first place on the slowness of the mechanism causing it, namely the
absorption of water and extension by growth of the outer epidermis.
It seems not unlikely, however, that closure is impeded by the pre-
vention of the escape of water by the mutually appressed valves of
the lobe margins and the probably tight appression of the non- valvular
portions. This, of course, insures in nature the retention of prey.
To test this point, Ashida made a hole in the bristle-side lobe to
allow the free escape of water, when the record indicated that the
inclosed water in an uninjured trap does indeed offer impedance to
narrowing. The free-side lobe, however, due to its measurable rigidity
offers resistance to buckling and by itself produces irregularities in
the rate of narrowing, which commences in any event, if the stim-
ulus is sufficient, in about 30 min. after the shut stage has
been reached. The narrowed condition in the case of strong stimu-
lation, but in the absence of prey, is maintained for a period of from
6 to 12 hours.

In the return to the widely open condition, the trap passes through
the reverse of the two phases of movement seen during shutting and
narrowing, that is, during a first period the rebulging of the free-side
lobe takes place, followed by the reopening of both lobes when the
trap is again ready to react if stimulation is applied. All this is
ascribed to the growth of the inner epidermis. During its progress
irregularities of rate of movement can be ascribed to the resistance
to the inflow of water into the narrowed trap and the elastic action
of the thin regions added to the action proper to the thick regions.
It is not known just how the water enters, but it may again be sug-
gested that the valve-free parts of the lips of the thin regions may be
the place of entrance, as well as of exit.

To recapitulate. — The rapid shutting movement is caused by the
response of the inner epidermis of the thick region in loss of turgidity.
The slow narrowing movement is brought about by the growth of the
outer epidermis, following its stretching in the curving of the lobe.
The movements of recovery are due to the growth of the inner epi-
dermis, following the restoration of turgidity. The shutting move-
ment is facihtated by the circumstance that the walls of the outer
epidermis are at open rest, not stretched to their full capacity, and
that these walls can be stretched plastically. A feature peculiar to
Aldrovanda is the fact that the loss of turgor by the inner epidermis
causes curvature of the single large celled middle layer, the walls
attached to the inner epidermis shrinking and those to the outer
expanding. In a more anatomically complex organ, such as the trap
of Dionaea, the same must be true of all the parenchyma, but the
difference of extension between the outer and inner walls of any

Chapter XII — 205 — Dionaea and Aldrovanda

individual cell must be less, since the total difference as between the
inner and outer epidermis is distributed throughout the tissue of, it
may be, some dozen cells in thickness. It is quite possible, therefore,
to extend the explanation given above for Aldrovanda to Dionaea, at
the same time excluding the loss of turgidity from the parenchyma, and
refer the whole movement to the action of the epidermis alone. This,
of course, does not square with Brown's explanation, but it neverthe-
less deserves consideration.

Reference has already been made to the fact of sensitivity, its
seat and the varying response of leaves of various ages. We inquire
now more particularly into the responses to various types of agents,
whether stimulatory or otherwise. Under conditions of nature, within
the ordinary limits of temperature during the growing season, it has
been found that stimulation through pushing against the sensitive
hairs by animals, such as water fleas, spiders, etc. of small size, pro-
cures closure (shutting and narrowing). The trap then usually has
caught a small total amount of food material which is digested. In
the course of a few days (5-6, Czaja 1924) the traps reopen and are
ready to act again. This may be repeated by the same trap several
times, the number depending on the size of the prey caught and
the amount of undigested remains. The possible activity in repeated
response and reopening is certainly not so limited as thought by
Fenner.

If, however, the prey is large and fills the digestion cavity (Czaja
used for experiment pieces of flatworm), the trap may never open
again. This may be due to the time involved, so that the trap passes
through its growth period and loses its sensitivity, or the production
and accumulation of substances having a poisonous effect. Too much
feeding is known to have a deleterious effect in other carnivorous
plants. This argues little or nothing in regard to the total value of
the process, since one long feeding may be of as much use to the
plant as several short ones.

The observation of Burdon-Sanderson on Dionaea, that response
can be obtained by electrical stimulation, was the beginning of a num-
ber of studies of interest in the field of general plant and animal
physiology, leading to the examination of various agents on the ac-
tivity of the trap. By means of the electric current it has been pos-
sible to analyze the response into time phases. Czaja determined
the intensity of threshold stimuli to be 0.91 • lo"*^ Coulomb for an
opening shock, and 0.24-10"^ for a closing shock. By repeated ap-
plication of smaller shocks he found that there is a summation of
stimuli. AsHroA used this method for further analysis of the response,
and found that for fully opened leaves the direction in which the
current engages the trap has its effect, which is greater when applied
transversely than longitudinally, from which it is inferred that the
stimulus is more effective when running parallel to the long cells of
the motile zone, than across, and that the latter is more sensitive to
this stimulus than are the sensitive hairs themselves. This may be
related to the various resistances offered by the tissues concerned and
the direction of the current through the individual cells. The possi-
bility of controlling the intensity of stimulation by means of the elec-

Francis E. Lloyd — 206 — Carnivorous Plants

trie shock has further permitted the examination of the behaviour of
the trap under special conditions of temperature.

Within the permissible temperature limits (extreme temperatures,
it goes without saying, are finally damaging) Czaja found that sensi-
tivity increased with higher temperatures, as determined by observa-
tion between 15 to 35 deg. C. Raising the temperature gradually to
45 deg. was followed by spontaneous closure of nearly all the traps
(that is, excepting some of the oldest). Opening again on reduction
of temperature to 20 deg. they again closed on gradual lowering to
10 deg., the older traps responding in this way on reducing the temper-
ature further to 5 deg., all due to the reduction of sensitivity, as shown
by appropriate trials (Czaja 1924).

AsHiDA went further, and found that sudden changes of tempera-
ture (he used changes of 10 deg. C.) in either direction would cause
closure. From his data the curious fact emerges that sudden reduc-
tion of temperature beginning with any workable levels from 10 to
40 deg. C, is more effective than sudden rise in temperature at these
levels. Further, the higher the initial level of temperature the more
sensitive is the trap to rises, and the less sensitive to drops. Exam-
ples of stimulation to both rise and fall of temperature are not lacking,
e.g. changes in the growth rate of coleop tiles (Silberschmidt), nastic
movements of leaves (Stern and Bunning), the curhng of tendrils
(MacDougal), cited by Ashida, offer some analogy. Protoplasmic
movement is retarded only by a fall in temperature, the Aldrovanda
trap being stimulated by both rise and fall, but more by the latter.
Aldrovanda appeared therefore to Ashida to be unique in the quantita-
tive aspects of behavior in this regard. Metzner (1920) had, however,
already shown that bipolar-flagellated Spirillum sp. show a reversal of
movement due to thermotaxis both on increase and decrease of tem-
perature. But such rises and falls of temperature as can be experi-
mentally imposed can scarcely be expected in nature except as slow
changes; they can hardly be regarded as affecting appreciably the
general economy of the plant.

In a third paper Ashida has given the results of studies of response
of Aldrovanda traps of different ages to weak and strong stimulation
applied in the form of constant currents of 30.6 and 70.1 volts. With
the strong current all traps of various ages and at different temper-
atures (10 to 40 deg.) close promptly in the same time interval. With
the weak current, however, the responses were scattered, a frequency
polygon expressing the results having its highest node at about the
same position as in the case of the strong current stimulation, with
secondary nodes scattered to the right of gradually lessening height.
The explanation of the delayed responses lies, Ashida beheves, in
the distinct and different sensibility of the hairs and the motile zone,
diverse excitability of different cells and other possible causes.

Responses to chemical stimuli. — Both Czaja and Ashida have
studied the behavior of Aldrovanda to a series of chemical agents:
narcotics, electrolytes (acids, alkalis, salts), non-electrolytes (sugar,
glycerine).

Sugar, glycerine. — It is difficult because of the impenetrable cu-
ticle to plasmolyze the cells of the trap unless a cut is made to allow

Chapter XII — 207 — Dionaea and Aldrovanda

the approach of the plasmolyte, as Czaja also found for Utricularia.
At the most, the cells display a systrophic contraction, but there is
no quantitative relation to the concentration of the plasmolyte
(Ashida). By watching the creeping in of the plasmolyte through
the cut ends of cells, Czaja estimated the osmotic equivalent to be
about m/3 KNO3.

A plasmolyte has the effect of immobilizing the trap by the with-
drawal of water. For sugar (sucrose) immobilization occurs in from
40 seconds to 7.5 minutes for solutions of the concentrations 0.50 M.
to 0.1 1 M. according to the age of the leaf, the younger being more
easily affected. They react on stimulation till immobilization sets
in. On immersion in the solution (0.2 M. sucrose), the trap opens
a little beyond the normal, due perhaps to the withdrawal of water
weakening the bending force of the outer epidermis and the middle
layer. The springiness of the walls is, however, retained to some
extent, and they will spring open if closed by force, unless too far.
When left for some hours (six) in a 0.15 M. glycerine solution the
power of movement is recovered, due perhaps to the penetration
of the solute into the cells. Rapid changes of concentration in either
direction can stimulate, causing partial closure, but it is evident
that the immobilising effect and that of stimulation are antagonis-
tic. Only traps which are not completely immobilized can react at

all.

Neutral salts. — Czaja found only a "narcotizing effect." Ashdda,
however, found also that salts in solution can stimulate, for, though
strong solutions may quickly immobilize, their first effect is stim-
ulation and the traps close. Even in a saturated solution of KCl
the traps reacted within 1.6 to 2.6 seconds according to the age of
the trap, the younger the quicker. Similar behavior was found for
some other salts.

Acids and alkalis. — When Ashida exposed traps to low concen-
trations of acetic acid and of HCl (0.005 to 0.05 N.) they closed after
various rather irregular periods from 29 to 2 minutes, respectively.
Since osmotic pressure is regarded as not entering in, Ashida tried
combining an acid with a non-electrolyte (acetic acid with sucrose)
and found that the reaction time was reduced, and that the more
sucrose is present, the shorter the reaction time. Two possible ex-
planations present themselves. Sucrose, even much below the concen-
tration which can stimulate osmotically, may help the stimulating
effect of the acid, an additive effect; or the permeability of the proto-
plast to acid may be increased during partial plasmolysis. Ashida
favors the latter alternative, for the additive effect is not observed
in young traps, in which osmotic and chemical stimulation alone
procure quicker responses. The second alternative also receives support
from the observations of Scarth (1927) that acid dyes penetrate
Spirogyra cells more readily when the protoplast is changing its vol-
ume during plasmolysis.

Where now is the stimulus perceived? From applying acid to the
different leaf surfaces, Ashida found that when it was put on the
upper surface the response was obtained more quickly, indicating
that the joint cells of the sensitive hairs are the points of perception,

Francis E. Lloyd — 208 — Carnivorous Plants

though the difficulty of confining the acid to the hairs alone, which is
obvious, throws some doubt on the conclusion.

Other organic substances: — When exposed to commercial for-
malin, the trap becomes immobilized in 35-45 seconds. During
this period, if stimulated mechanically the traps will close. In some
cases the traps close on direct stimulation by the reagent about 15-30
seconds after immersion. Because overtaken by immobilization, the
closure is never complete. In dilute formalin, the closure may be
complete, since immobilization does not overtake the traps quickly
enough. This is explained by the toxic effect overtaking the epi-
dermis cells before the reagent can enter by way of the sensitive
hairs. Ashida also reports the recovery of sensibility in 20 minutes
after immobilization by exposure to concentrated formalin for 45 sec-
onds, indicating that the injury to the epidermal cells is to some
extent reversible. Such traps may close and narrow, perhaps as an
after effect of adherent formalin for i minute, hence completely im-
mobilized traps close and narrow in water 5-10 minutes later in spite
of complete immobilization. If entirely killed by longer exposure, no
movement occurs.

Such experiments are puzzling, but indicate at least that the
toxic effects are realized somewhat slowly, and that the stages of
turgor reduction realized in the meantime are such as to allow the
working of the mechanism of closure, partially or completely. The
slow penetration must be due to the resistance of the cuticle. Ashida
recognizes two effects, stimulation and immobilization, on the rate of
the latter depending the ability to respond, through the action of the
sensitive hairs, into the hinge cells of which the reagent can penetrate
more quickly.

In ethylalcohol 10-40 percent by volume, traps close spontane-
ously in from about i to 90 minutes, according to age. Curiously
enough, in solutions stronger than 40% closure occurs in two steps,
both sudden, separated by a pause.

In saturated chloroform-water, most traps close within i minute,
some quickly, some slowly and some irregularly. Restoration to water
procures no further activity, and they die. Czaja had previously
obtained similar results. He found also that, after treatment with
ethyl ether at similar concentrations, return to water restored the
normal activity. The effect of narcotics, he says, is at first stimulative,
then destructive. He obtained similar results with methyl and ethyl
alcohols.

The addition of peptone or egg-albumin to the culture medium
causes the traps to close and narrow, as also if a fragment of fish
or meat is fed to the trap. Closure follows then from mechanical
stimulation, and narrowing results from the chemical stimulation.
Gelatin and coagulated egg-albumin do not stimulate beyond the
shutting stage, since they do not stimulate chemically (Ashida).

Digestion and absorption. — There is not much doubt that Aldro-
vanda digests prey and absorbs the products of digestion. All ob-
servers have seen that the bodies of prey, except the hard parts,
disappear, and Darwin did a few experiments which convinced him
that absorption takes place. The evidence is, however, not complete,

Chapter XII — 209 — Dionaea and Aldrovanda

and we base our opinion with Darwin on the obvious analogy with
Dionaea. The closure of the trap into the narrowed posture reduces
the volume of fluid containing the digestive ferment, if present, thus
rendering it more effective (Darwin). Fermi and Buscaglione
(1899) offered some evidence that Aldrovanda is capable of digestion
by placing traps on sterilized gelatin, and finding that it was rendered
fluid.

To the earliest observers the traps of Aldrovanda were thought
to be vesicles. The lifting of the plant from the water released the
traps so that air was inclosed, as in the case of Utricularia, but, as
CoHN pointed out, this air was purely accidental. That it enabled
the plant to float in the water, or at least assist in this when the
plant becomes loaded with prey (Fenner), in view of the fact that
it floats whether air is present in the trap or not, was seen to be a
gratuitous explanation. More recently Fenner advanced the idea
that gas is present as the result of chemical activity (digestion of
one kind or another). But this also has been questioned (Czaja
1924). That the presence of an air bubble assists in digestion
by reducing the water content of the digestion cavity, thus procuring
a more concentrated solution of ferments, that the water is absorbed
by the epidermis cells, and that the air bubbles assist in opening by
outward pressure (Fenner), are ideas which are superfluous in view
of the fact that adequate observation shows that the inclusion of
air is accidental. Fenner reports also that in vigorous leaves, after
the capture of prey, not only do the digestive glands, but also the
valve trichomes and the quadrifid hairs show signs of activity by
exhibiting changes in their contents. This requires further exami-
nation. It is true that Darwin had made similar observations indicat-
ing, it seemed to him, that the valve teeth and quadrifids do absorb,
but this evidence was regarded only as indicatory.

When small prey are captured, the trap, stimulated by the vigorous
movements in the attempt of the prey to escape, remains closed for
some time according to the mass of the substance to be digested.
If the trap is overfed, as when supplied with large pieces of a flat-
worm (Czaja) it remains permanently closed, possibly because poi-
soned by the overplus of deleterious products, or perhaps because
growth had ceased. Reopening may be repeated several times under
favorable conditions of sufficiently meagre feeding. As already said
heavy feeding may, however, advantage the plant as much as several
smaller feedings.

According to Schenk (on the authority of Cramer, 1877) Al-
drovanda was grown by him for two years in an inorganic salt solu-
tion without, apparently, any deleterious effect. Pfeffer cites this
in support of the non-obhgate character of carnivory in this plant.
This experience of Schenk's is surprising in the light of the expe-
riences of AsHiDA, who found it very difficult to grow the plant ex-
cept under rather special conditions. "No inorganic culture medium
could be found which would keep the plant in the normal form even
for a week". Aldrovanda grows in shallow water between the stems
of Typha, Zizania, Phragmiles, etc., which Hausleutner regarded as
merely protection against sun and wind. Asheda took a hint from

Francis E. Lloyd — 210 — Carnivorous Plants

these relations, and found that by introducing into the water re-
mains of the associated plants mentioned above, Aldrovanda can be
made to grow quite satisfactorily even to flowering and seeding.

Literature Cited:

AsHTDA, Joji, Studies on the leaf movement of Aldrovanda vesiculosa L., I. Process and
mechanism of the movement. Mem. Coll. Sci., Kyoto Imp. Univ. B. 9:141-244, 1934;
II. Effects of mechanical, electrical, thermal, osmotic and chemical influences. Ibid.
B. 11:55-113, 1935; III. Reaction time in relation to temperature. Bot. Mag. Tokyo
51:505-513, 1937.

Balfour, T. A. G., Account of some experiments on Dionaea muscipida (Venus's Fly-
Trap). Trans. Proc. R. Soc. Edin. 12:334-369, 1875.

Batalin, .a.., Mechanik der Bewegung der insectenfressenden Pflanzen. Flora 35:54-58;
105-111; 129-154, 1877.

Broussonet, P. M. A., Hist. & Mem. de I'Acad. des Sci. 1784, p. 614.

Brown, Wm. H. &L. W. Sharp, The closing response of Dionaea. Bot. Gaz. 49:290-302,
1910.

Brown, Wm. H., The mechanism of movement and the duration of the effect of stimula-
tion on the leaves of Dionaea. Am. Journ. Bot. 3:68-90, 1916.

Burdon-Sanderson, J., Note on the electrical phenomena which accompany stimulation
of the leaf of Dionaea muscipida. Proc. R. Soc. London 21:495-496, 1873.

Burdon-Sanderson, J., Venus's Fly-Trap {Dionaea muscipida). R. Inst. Lecture, 5 June,
1874. Nature 10:105-107; 127-128, 1874.

Burdon-Sanderson, J., On the electromotive properties of the leaf of Dionaea in the ex-
cited and unexcited states. Phil. Trans. Roy. Soc. London 173:1-53, 1882. Second
paper, ibid. 179:417-449, 1888.

Burdon-Sanderson, J. & Page, Paper read before the R. Soc. London, 14 Dec. 1876 (n.o.).
Results set forth in Burdon-Sanderson, 1882.

Canby, W. M., Notes on Dionaea muscipida ElHs. Meehan's Gard. Mo.: 229-231, 1868.

DE Candolle, C. p., Sur la structure et les mouvements des feuilles du Dionaea muscipida.
Arch. Sci. Phys. Nat. 55:400-431, 1876.

Caspary, R., Aldrovandia vesiculosa Monti. Bot. Zeitung 17:117-124; 125-132; 133-139;
141-150, 1859.

Caspary, R., Aldrovandia vesiculosa. Bot. Zeitung 20:185-188; 193-197; 201-206, 1862.

COHN, F., t)ber Aldrovandia vesicidosa Monti. Flora 43:673, 1850. {through Cohn 1875).

Cohn, F., tJber die Function der Blase von Aldrovandia und Utricidaria. Cohn's Beitrage
Biol. d. Pflanzen i(3):7i-92, 1875.

CoKER, W. C, The distribution of Venus's Fly Trap {Dionaea muscipida). Journ. Elisha
Mitchell Sci. Soc. 43:221-228, 1928.

CoKER, W. C, A new locality for the Venus' fly-trap {Dionaea muscipida). Science 88:188,
1939.

Cramer, C, tJber die insectenfressenden Pflanzen. Lecture, Zurich 1877.

CuRTis's Botanical Magazine, Vol. 20. Dionaea muscipida. Venus's fly-trap. Plate
785 and description, 2 pp., London, 1804.

Curtis, M. A., Enumeration of plants around Wilmington, N. C. Boston Journ. Nat.
Hist. 1, 1834 {Dionaea muscipula, pp. 123-127).

CzAjA, A. Th., Reizphysiologische Untersuchungen an Aldrovandia vesiculosa L. Arch. f.
d. gesamte Physiologie d. Menschen u. Tiere 206:635-658, 1924.

Darwin, C, Insectivorous Plants. London 1875.

Darwin, E., The Botanic Garden. London 1791.

Darwin, F. in Darwin, C, 1875, ed. of 1888, p. 257.

Diderot, Denis, Oeuvres de Diderot. Ed. by .\ss6zat, 1875. {re Dionaea, 9:257) {through
£. MoRREN, see under Drosera).

Diels, L., Droseraceae. Das Pflanzenreich, IV, 112:1-136, 1906.

DiLLWYN, Lewis W., Hortus CoUinsonianus. An account of the plants cultivated by the
late Peter Collinson, Esq., F. R. S. arranged alphabetically according to their
modern names, from the catalogue of his garden and other manuscripts. Swansea,
W. C. Murray and D. Rees, 1843.

DuvAL-JouvE, J., Letter to de Schoenfelt. Bull. Soc. Bot. de France 8:518-519, 1861,

Duval-Jouve, J., Note sur quelques plantes dites insectivores. Bull. Soc. Bot. France
23:130-134, 1876.

Ellis, John, Directions for bringing over seeds and plants from the East Indies and other
distant countries in a state of vegetation together with a catalogue of such foreign plants
as are worthy of being encouraged in our American colonies, for the purpose of Me-
dicine, Agriculture and Commerce. To which is added The Figure and Botanical
Description of a New Sensitive Plant called Dionaea muscipida or Venus's Fly-trap.
London, 1770. Sub-title: A botanical description of the Dionaea muscipula or Venus's
flytrap; a newly discovered sensitive plant: in a letter to Sir Charles Linnaeus

Chapter XII — 211 — Dionaea and Aldrovanda

Knight of the Polar Star, Physician to the King of Sweden, and Member of most of

the learned societies of Europe, from John Ellis, Fellow of the Royal Societies of

London and Upsala (with a colored plate of Dionaea muscipida).
Fenner {see imder Nepenthes).
Fermi & Buscaglione {See under Utricidaria).
Fraustadt, a., Anatomie der vegetativen Organe von Dionaea muscipida Ellis. Cohn's

Beitrage Biol. d. Pflanzen 2:27-64, 1877.
Gardiner, \V., On the changes in the gland cells of Dionaea muscipida during secretion.

Proc. Roy. Soc. London 36:180-181, 1883.
Goebel, K., Pflanzenbiologische Schilderungen. S. 57-72, Marburg 1891.
GoEBEL, K., Organographie der Pflanzen. S. 741-742, 1898-1901.
Guttenberg, H. von, Die Bewegungsmechanik des Laubblattes von Dionaea muscipida

Ell. Flora 18-19:165-183, 1925.
Guttenberg, H. v., Zur Kenntnis lebender Bewegungsmechanismen. Planta i :666-678,

1926.
Haberxandt, G., Sinnesorgane im Pflanzenreich. S. 108-117, Leipzig 1901.
Hausleutner, Cultur der Aldrovanda. Bot. Zeitung 1850, S. 831 {through Ashida).
Heinricher, E., Zur Kenntnis von Drosera. Ferdinandeum Zeitschr. Innsbruck 46:1-25,

1902; 47:300-307, 1903.
Holm, T., Contributions to the knowledge of the germination of some N. American Plants.

Mem. Torr. Bot. Club 2:57-108, i8gi.
Hooker, J. D., Address to the Dept. of Bot. and Zool., B. A. A. S. Belfast meeting 1874.

London 1875, pp. 102-116 {Dionaea, pp. 103-105).
Jones, Frank Morton, The most wonderful plant in the world. Natural History 23:589-

596, 1923 (Contains a facsimile of part of a letter from Charles Darwin to Wm. M.

Canby, in which the phrase forming the title of Jones' paper is quoted. The letter

was dated froni Down, Kent, Feb. 19, 1873).
KoRZSCHiNSKi, S., tjber die Samen der Aldrovandia vesiculosa L. Bot. Centralb. 27:302-

304; 334-335> 1886.
Kostytschew, S., Die Photosynthese der Insektivoren. Ber. d. D. Bot. Ges., 41:277-280,

1923.
Kurtz, F., Zur Anatomie des Blattes der Dionaea muscipida. Arch. Anat. und Physiol.,

Lpz. 1876:1-29.
DE Lassus, Aug6, Analyse du memoire de Gaetan Monti sur V Aldrovandia suivie de

quelques observations sur I'irritabilite des follicules de cette plante. Bull. Soc. Bot.

de France 8:519-523, 1861.
Lindley, John, An introduction to botany. 4th Ed., London 1848.
Lloyd {See under Heliamphora).

MacDoug.'VL, D. T., The physiology of tendrils. Bot. Centralbl. 66:145, 1896.
Macfarlane, J. M., Contributions to the history of Dionaea muscipula EUis. Contrib.

Bot. Lab. Penna. 1:7-44, 1892.
Metzner, P., Die Bewegung und Reizbeantwortung der bipolar begeisselten Spirillen.

Jahrb. wiss. Bot. 59:325, 1920.
Meyen, F. J. F., Neues System der Pflanzenphysiologie 3:543-550, Berlin 1839 {through

V. Guttenberg).
Monti, Gaetano, De Aldrovandia novo herbae palustris genere. Comment, de Bononiensi

Scient. et Art. Instituto et Academia 2(3)404-411, 1747. Monti's paper itself bears

no date, but must have appeared between 1737 (it contains a quotation from Lin-
naeus which appeared in that year) and 1747 (de Lassus, 1861). A translation into

French appeared in the Collection Academique 10:401-407 (foreign part), Paris 1773

{through DE Lassus).
Mori, A., Nota suU'irritabilita delle foglie dtW Aldrovandia vesiculosa. Nuov. Giom. Bot.

Ital. 8:62, 1876.
Munk, H., Die electrischen und Bewegungserscheinungen am Blatte der Dionaea muscipida.

Arch. f. .\nat., Physiol, u. wiss. Med. 1876:30-203.
NiTSCHKE, 1861 {See under Drosera).
Oudemans, C. a.. J. .\., Over de prikkelbaarheid der bladen van Dionaea muscipida. Versl.

Mededeel. K. Akad. Wet. 9:320-336, 1859.
Pfeffer, W. {see under Drosera).
Plukenet, L., Almagestum Botanicum seu phytographiae Plukenetianae onomasticon

etc. London 1696. 2d. ed., London 1769:211-212.
Rees and Will {See under Nepenthes).
Sanderson, J. Burdon, see Burdon-Sanderson.

Scarth, G. W., The influence of internal osmotic pressure and of disturbance of the cell-
surface on the permeability of Spirogyra for acid dyes. Protoplasma 1:204, 1927.
Schenk, cited by Cramer, 1877.
Silberschmidt, K., Untersuchungen iiber die Thermowachstumsreaktion. Ber. d. D.

Bot. Gesell. 43:475, 1925 {through Ashida).
Sims, John, Dionaea Muscipula. See Curtis's Botanical Magazine.
Smith, Cornelia Marschall, Development of Dionaea muscipula, I. Flower and Seed.

Francis E. Lloyd — 212 — Carnivorous Plants

Bot. Gaz. 87:507-530, 1929; II. Germination of seed and development of seedling to

maturity. Ibid. 91:377-394, i93i-
SoLEREDER, H., Systcmatlsche Anatomie der Dicotyledonen. Stuttgart 1899.
Stein, B., 1873. Mentioned by Cohn, 1875, p. 73.
Stern, K. & E. Bunning, Uber die tagesperiodischen Bewegungen der Primarblatter von

Phaseolus viultiflorus, I. Der Einfluss der Temperatur auf die Bewegung. Ber. d. d.

Bot. Gesell. 47:565-584, 1929 {through Ashida).
Thomson, T. and J. D. Hooker, Praecursores ad Floram Indicam. Journ. Proc. Linn.

Soc. Bot. London. 2 :83, 1858.
Troll, 1939. See under Nepenthes.

Young, Wm., Catalogue des Arbres, etc., d'Amerique. 1783, p. 34 (n.v. through Coker).
Ziegenspeck, H., tJber Zwischenproducte des Aufbaues von Kohlenhydrat-zellwanden

und deren mechanische Eigenschaften. Bot. Arch. 9:297-376, 1925 {n.v., through v.

Guttenberg, 1926).

Chapter XIII
UTRICULARIA, BIOVULARIA AND POLYPOMPHOLYX

Form, habit and habitat. — Distribution. — Embryology. — The various types ar-
ranged according to the character of their traps. — The vulgaris type (Freely floating
forms. Semi-submersed, submersed but anchored forms. Terrestrial and epiphytic forms).
— Biovularia, a floating type. — The purpurea type, a floating form. — The dichotoma-
monanthos type (Floating form. Terrestrial. Polypompholyx: annual, monaxial). — The
cornuta type. — The caerulea type. — The capensis type. — The orbiculata type. — The
longiciliata type. — The globulariaefolia type. — The nana type. — The Lloydii type. —
The Kirkii type. — The simplex type.

Of the Lentibulariaceae there are recognized by Kamienski five
genera, Pinguicula, Genlisea, Polypompholyx, Utricularia and Bio-
vularia. The first two, Pinguicula and Genlisea have already been
considered. It now remains to treat Utricularia, Polypompholyx and
Biovularia. These plants are freely floating or anchored aquatics, or
are epiphytic, in wet moss, or are terrestrial in wet to moist sandy
soils. The largest of these are among the aquatic and epiphytic
forms. The former are well exemplified by U. vulgaris, among the
best known, which is a lax floating plant of several feet in extent,
but of little mass. The epiphytic U. reniformis is more massive
and makes a brave showing as a greenhouse plant, frequently found
in the good company of the orchids because of the similar habitat
requirements and the showiness of their flowers. It is indigenous in
Brazil. One species, U. nelumbifolia, finds its home habitually in the
water-containing leaf rosettes of large Tillandsias, whence it sends
out runners which reach into the urns of neighboring rosettes
(Gardner 1842, Ule 1898). Showy species are U. Humholdtii, U.
longifolia (both S. America), U. Endresii (Costa Rica) and the related
U. Dusenii which, though small, has a flower of the same type. Other
of the larger species are the tropical American terrestrial U. globu-
lariaefolia and amethystina (Trinidad), and a few others chiefly no-
ticeable because of their tall inflorescences. But among the aquatics
are found also the smallest species, as e.g. U. cymbantha (Africa) and
a related unnamed species of still smaller size, the flower being only
2 mm. long, and the stolons mere threads, collected by Miss E. L.
Stephens in Portuguese East Africa. As small are the two species re-
ferred by Kameenski to Biovularia, B. minima and B. olivacea. With
a few exceptions such as those above noted, the terrestrial species
are smafl. When in flower and in numbers, they are conspicuous
enough, but if not in flower they may be found only with very as-
siduous hunting. How one would find U. simplex (S. W. Australia)
unless in flower is past guessing, so minute are the leaves.

But large or small, massive or delicate, perennial or evanescent
annuals, they present a complex and puzzling morphology. They
are entirely rootless, even in the embryonic condition. The distinc-
tion between stem and leaf is vague. Only in the inflorescence and
in certain shoots (air-shoots of U. vulgaris etc.) is the morphology

Francis E. Lloyd — 214 — Carnivorous Plants

easily recognizable. But most to be wondered at are the traps which
present an astounding degree of mechanical delicacy depending on
a fineness of structure scarcely equalled elsewhere in the plant king-
dom. Moreover they occur in an unexpected variety of form. But
withal they are small, the largest scarcely exceeding 5 mm. in great-
est extent, the smallest 0.3 mm. The prey caught by these traps
are small — water fleas, minute larvae often of mosquitos, very young
fish (Dean) and small tadpoles, fish and tadpoles, however, being
ensnared only by being caught by the tail or head in the mouth
of the trap but not entirely engulfed as is smaller prey (Dean,
Lloyd). Larger prey may, however, be finally "entirely absorbed"
according to Matheson (1930), as e.g. in the case of the larva of
Brachydeutera argentata. Matheson sees the difficulty of explain-
ing this, since the greater mass of the animal's body is outside of
the trap. An explanation is, however, at hand and will be later of-
fered.

The observation that mosquito larvae are often caught in the
traps has led some investigators to hope that the floating large trapped
species of Utricidaria would be useful in aiding to control those pests
(Matheson 1930). It is well enough known that multitudes of
mosquito larvae are in fact captured, the number being limited only
by the number of traps available (Franca).

The flowers are of the personate type, two-lipped, the throat
usually closed by a palate, and the lower Hp provided with a spur
of various shape, except in a few species, in which it is a saccate
enlargement, as in Biovularia, in U. cymbantha, U. Stephens ae (Ms.
name), and in U. minor, a fact which tends to invalidate Kamienski's
genus. U. purpurea and its co-species have flowers with a peculiarly
laterally saccate lower lip. Most extraordinary are the two species
U. capilliflora F.v.M., and U. Dunstani Lloyd (Lloyd 1936^) (Text
FIG. 5). The former has the upper lip drawn out into two very long
attenuate lobes, the lower lip into five slender finger-like lobes. In
U. Dunstani, the upper lobe is rounded and entire while the lower
lip is five lobed, the middle adjoining two lobes small, and the two
lateral lobes very long and attenuate and to the casual observer iden-
tical with the long tips of the upper lip in U. capilliflora. The flowers
are very small; had they been large they would have excited as much
admiration as the flowers of some orchids or that of an Aristolochia
with long appendages. The showy flowers of some S. American spe-
cies have been mentioned. These are large and are rendered conspicu-
ous by the wide lateral lobes of the lower lip particularly. Yellow
is the most prevalent color, but white to purple flowers are also fre-
quent with admixtures in the form of yellow or reddish markings
on white or blue grounds; or red on yellow, especially on the palate,
are often met with. Generally the pistil has a unilocular, globular
ovary with a two lipped stigma, the lower being considerably larger
than the upper. The seeds are usually numerous on a globular ba-
sifixed placenta. In certain species, however, two to three seeds
only are produced on a rounded placenta {U. cymbantha) or two only,
back to back, in Biovularia, the placenta reduced. The release of
the seeds may be by circumcissile or irregular fracture, by a trian-

Chapter XIII

— 215 —

Utricularia, Biovularia

gular window {U. subulata), or by a longitudinal slit {U. orhiculata).
There are only two stamens bending upward along the curve of the
ovary wall.

The sepals are usually two in number, upper and lower. In Poly-
pompholyx there are four, hence de Candolle's justification of his
name Tetralobus, of slightly later date than and therefore superceded
by Polypompholyx Lehmann. Polypompholyx as a generic name
should, according to Barnhart, be in turn superseded by Cosmiza
Raf. 1838, since both Polypompholyx Lehmann (Feb. 1844) and Telra-
lohus DC. (Mar. 1844) are later names. Lehmann's name is at the
moment generally used.

U. fimbriata superficially resembles Polypompholyx in having its

Fig. 5. — Utricularia capilliflora (277) and U. Dunsiani (272); up, upper corolla lobe;
us, upper sepal; aus, abnormal upper sepal. The two figures on the right are shadow
prints from herbarium specimens, natural size.

two fimbriate sepals backed up by two fimbriate bracts, and has been
incorrectly referred to Polypompholyx. Barnhart calls it Aranella
fimbriata (Barnhart 191 6).

The Utriculariae are of world wide distribution. Of the widest
distribution are the submersed or semi-submersed plants of the type
of U. vulgaris. They are found occupying a circumboreal zone char-
acterized by certain species peculiar to it, throughout N. America
including Greenland, Europe and Asia. Related species extend
throughout the tropics and into S. America, S. Africa, Australia and
New Zealand. The submersed floater U. purpurea and its co-species
are purely American, extending from Maine, possibly Newfoundland,
down the Atlantic Coast, west, to Indiana, and into E. South Amer-
ica. Another submersed form, U. tubulata, which is related to quite
another group peculiar to Australia (the U. monanthos type), is found
in N. E. Australia (Queensland).

The terrestrial forms are widely distributed in the tropics of the

Francis E. Lloyd — 216 — Carnivorous Plants

Old and New Worlds, with northward extensions along the Atlantic
coast, through Central America, and along the Pacific Coast of Asia.
These include a considerable number of distinct types, some of which
are pecuHar to America, others to the Old World, and still others
common to both (£/. suhulata). They are to be found also in Aus-
tralasia.

So far as is known, Utricularia is not found on oceanic islands.

Embryology. — The embryological story of Utricularia has not yet
been worked out in its fullness, but certain striking features have
been observed and have been recorded by Lang, Merl (Genlisea),
Merz, and Wylie and Yocom. Merz pointed out that the vascular
tissues supplying the large placenta end there and do not enter the
funicles of the ovules. As a compensation for this, it would appear,
special masses of nutritive tissue arise in the chalaza and in the pla-
centa {Utricularia vulgaris, etc.) or what amounts to the same thing,
in a "funicular hump" (to quote Wylie) (Poly pomp holy x Lehm.).
These islands of food materials are made use of for the embryo by
the penetration into them of peculiar haustorial extensions of the
endosperm. Transitory starch appears in these haustoria. These nu-
tritive masses are finally cut off from the ovule, as the embryo ap-
proaches maturity, by the development of diaphragms of suberized
cells. Since according to Merz the end of the suspensor extends
into the placental nutritive tissue, this is cut ofT at the same time.
Wylie and Yocom do not support this. In the related Genlisea,
nutritive islands occur entirely within the ovule (Merl).

The ovule has but a single integument, which is common enough
among the sympetalae, from the outer cell course of which a thin
testa is derived. The tegmen, membranous and delicate, may repre-
sent the remaining tissue of the integument, since the nucellus is
absorbed (Merz). The mature embryo has neither root nor punctum
vegetationis (Merl, Lloyd).

Seed, embryo and seedling. — The seeds, in most cases, develop
on a large globular, central placenta, and are crowded suf^ciently
in some species so that by mutual compression, they become angu-
lar, approximately hexagonal, with fiattish outer chalazal and inner
micropylar surfaces, the latter the smaller (Warming, Wylee). The
planes of the sides radiate as if from the center of the placenta {U.
vulgaris, U. resupinata), or the ovules overlap and become winged
with circular wings {U. oligosperma, U. exolcta) the whole dished
into a concavo-convex lens shape. In other species, chiefly terres-
trial, the seeds are minute (down to 0.2 mm. in diam.) and globular
or oval, occasionally lobulate {U. reniformis, U. Dusenii, U. pur-
purea) or have polar clusters of trichomes {U. brachiata) (Compton)
or scattered glochidia covering one end {U. orbiculata).

The testa is composed of an epidermis of darkened (reddish or
brown), relatively large cells with their inner and radial walls thick-
ened, the outer thin and collapsed in maturity. Their shape varies
from round angular to elongate, many and relatively small or few
and relatively large. In a few instances they become mucilaginous
when wet, in species which grow with the vegetative parts submersed
and attached to a stony substrate in running water {U. rigida,

Chapter XIII — 217 — Utricularia, Biovularia

and probably U. neottioides) . In a very few instances the seed has
no hard testa, but simply a thin lax membrane easily torn, and the
seed seems to be released by the rotting away of the capsule {Bio-
vularia) .

Embryo. — The embryo is a mass of scarcely differentiated cells
containing much starch and oil, and is either bun shaped with a
depression at the vegetative pole {U. vulgaris) (Warming) (2/ — 6)
or flattened by lateral compression, with the growth pole on its edge
{U. emarginata and exoleta) {21 — 7), this position and form being due,
according to Merz, to rotation of the embryo-sac and embryo through
90 deg. following the abscission of the egg-pole endosperm; or again
more or less oval or spindle shaped (a host of species, e.g. U. bifida)
(22 — I, 18). In a few there is a depression at the root pole (Poly-
pompholyx) {22 — 26). Its structure is very simple: it is a mass
of rounded parenchyma clothed with an epidermis, the cells of which
covering the growth pole are small, columnar and highly protoplas-
mic. The primary organs usually originate at this pole, but may
occasionally appear elsewhere {U. bifida) {22 — 19-22). There is an
entire absence of a root. Even at maturity the embryo has as yet
no lateral organs. Goebel reported them in the embryo of U . or-
biculata, but I have failed to find them. In U. reniformis and U.
nelumbijolia primary leaves have been seen, and these species may
display vivipary (Merl).

While the growth pole produces lateral organs, two in many cases
(cotyledonoids) or various numbers in others, a punctum vcgetationis
(the primary primordium of a shoot) is never present in the embryo,
nor is ever developed. Shoots when present in the plant are always
produced as lateral organs.

Germination {22 — 1-28). — The events of germination show that
there are three types of seedling: (i) a simple, in which there are
two cotyledonoids (since one of these is a stolon); and (2) a com-
plex, in which there are 6 to 13 (Warming, Jane) cotyledonoids,
or indeed only two; and (3) a type in which there are no cotyledonoids
at all, with three primary shoots only. The simple type is displayed
by all the terrestrial species (so far as known), such as U . capensis, U.
bifida, U. monanthos, Polypompholyx, which have been studied. The
complex type is seen in U. vulgaris (with many cotyledonoids) and in
U. exoleta, U. emarginata etc. (with two cotyledonoids). U. capensis
among others has been studied by myself (19376). The seed is minute
and oval in shape. From the growth pole emerges at first a leaf
followed shortly by a stolon. Between them in the expected position
there is no punctum vcgetationis. The leaf extends upwards, the stolon
downwards. The first evidence of further growth is the appearance
of a trap, on the base of the stolon, a little away from the middle
Kne (22 — 14), or rarely, as in U. cleistogama, asymmetrically near the
leaf base. At the base of the trap-stalk a bud arises which pro-
duces an ascending stolon with more or less crowded leaves. This
becomes thicker with upward growth, finally becoming an inverted
cone, a protocorm bearing leaves, traps and stolons. Their vascular
tissues form within the protocorm a loose stem-Hke structure, similar
to that described by Warming for Genlisea and by myself for Poly-

Francis E. Lloyd — 218 — Carnivorous Plants

pompholyx. At its top is formed at length a large radially sym-
metrical bud, which becomes the inflorescence {22 — 11, 25). Of the
stolons, some are anchoring in function (rhizoids) while others be-
come runner stolons, bearing leaves along the upper surface, and
facing toward the corm, in the adaxial axils of which other scapes
may arise. Thus is produced a spreading plant. Scapes may also
arise from buds adventive on the leaves, this among many irregu-
larities of growth. Sometimes the primary leaf may produce a num-
ber of traps in the adaxial axils of which buds are formed, as in the
case of the axils of leaves produced on stolons.

This course of events, thus briefly stated, is followed by all other
species studied: U. bifida, U. monanthos, with evidence from a con-
siderable number of others. U. bifida was also studied by Goebel
who seems to have thought that the first bud was between the cot-
yledonoids, and suffered later displacement, as his drawings indi-
cate. But the statements of both Goebel and Merl (U. longifolia)
admit an asymmetrical position. My own studies of U. bifida have
revealed a strong tendency on the part of the embryo to produce
at first a leaf only, raised more or less on an elongate extension of
the upper part of the embryo body (a podium) and for the primary
stolon to be produced at its base from a distinct bud there seen {22 —
i-ii).

If the protocorm produces no runners, as in U. capilliflora, U.
violacea, U. Hookeri, Polypompholyx, an inflorescence can be produced
only from the top of the protocorm (22 — 25, 26), or by secondary
branching from the scape itself. This development is called abrupt.
Such plants do not spread by runner stolons. The behavior of U.
capensis etc., which spreads by runners, is diffuse.

In all these forms, especially in the abrupt kind of germination,
the primary bud arising from the first stolon can elongate very con-
siderably to form a naked ascending stolon producing leaves only at
length. The curious asymmetrical relations of embryo and its primary
organs than becomes quite apparent. I have seen this in U. violacea,
U. Hookeri and Polypompholyx, in which the events of germination
are preserved in the mature individual, for the embryo may persist
throughout the life of the plant {22 — 26).

The complex kind of germination is displayed by a lot of spe-
cies represented by U. vulgaris, which has been the subject of much
study by Warming, Pringsheim, Kamienski, Goebel, Jane and
Lloyd, and by U. oligosperma (Goebel). The condition found
in U. exoleta (Goebel) and U. emarginata though simple in the sense
indicated by the number of cotyledonoids, is merely a special case
evidently related to the vulgaris condition.

In U. vulgaris the embryo is a prolate spheroid. From a low dome
arise several leaf buds in an outer circle, within which several other
primordia stand, with no apparent order, though it has been stated
otherwise (Goebel). The outer, and some of the inner primordia
become awl-shaped leaves, but to pick out two of these as cotyledons
is impossible (Goebel). Occasionally one of them ends in a trap
(Goebel; Jane). Of the inner primordia, some become leaves, usually
one a dorsiventral shoot, and one a primary trap. Sometimes leaves

Chapter XIII — 219 — Utricularia, Biovularia

are produced which are evidently partly shoot in structure, the dis-
tinction between leaf and shoot being here vague. The apex of the
dome never develops {21 — 1-5)-

In U. exoleta and U. emarginata the embryo is laterally compressed
so that the growth pole is on the edge. Here two leaves are produced.
These may occasionally become shoots {21 — 7D). Near their bases,
but displaced to one side are usually two buds, sometimes a third.
These become dorsiventral shoots.

Aberrancies of development are not at all unusual. For example,
one embryo of U. vulgaris produced only two leaves and two shoots
which became at once dorsiventral "water shoots."

In U. nelmnbifolia, otherwise similar to U. vulgaris, the primary
leaves are widely forked (Merl) and those of U. reniformis are broadly
spatulate (Goebel). According to Goebel, the embryo of U. mon-
tana has no primordia before germination, and on it there are pro-
duced two primordia (the cotyledonoids) of which one becomes a
spatulate leaf, the other a trap, on the side of the "vegetation point"
which grows directly into a radial structure from which stolons etc.
grow. The interpretation here is to be questioned, as it is doubt-
ful if Goebel's identification of the vegetation point is correct.

The third type of germination is seen in U. purpurea. The embryo
is ovate, with a flattened broad end. From this there normally arise,
in succession, three shoot buds standing in a triangle with respect to
each other. The first is dominant in growth; the second to appear
may form at first a close second, but at length lags behind in growth;
while the third may not do more than put in an appearance, and then
fail to develop (Text fig. 6, 1-7). In one case a fourth shoot bud
appeared early (Text fig. 6, 4 and 5), but its subsequent development
showed it to be a branch at the base of the first bud, and in other
seedlings it appeared later, but constantly. The development of the
first shoot has been followed until it produced five whorls of branches
(leaves in the taxonomic texts). The earlier two whorls have but two
members bearing traps, placed symmetrically attached to the upper
moiety of the stem. On the upper surface of the stem as a member of
the whorl may appear a bud of unlimited growth, destined in its turn
to produce whorls of secondary members (Text fig. 6, 7). In the
third to fifth whorls, two additional members, placed in the wide dorsal
space, appear. These, however, do not bear traps. In the fourth
whorl, in one case observed, one of the ventral pair produced a branch
which, however, bore no trap. In one instance a seedling was found in
which the first and second buds were fasciated (Text fig. 6, 6). It is
evident that in this type the seedling at once takes on the morphologi-
cal character of the mature form, which, however, has whorls of four
{U. elephas) or six {U. purpurea) branched members.

It would be most interesting to know the course of germination of
U. tubidata, but of this we are yet ignorant.

Types of Utricularia. — In the following account, the various spe-
cies, so far as they are here included, are arranged according to the
character of their traps.

I . The Utricularia vulgaris type. — All the plants of this type have
traps which are closely similar in the mechanical details of structure

Francis E. Lloyd — 220 — Carnivorous Plants

to those of U. vulgaris, differing among themselves only as regards non-
essential, small details. A description of the trap is deferred till later.

Freely floating forms. — U. vulgaris itself is the best known and
most widely distributed, and will serve well as the type of numerous
species, world wide in distribution.

The plant consists of a cylindrical or laterally compressed axis
which may reach a length of 300 cm., probably more, supporting
two lateral rows of divided "leaves," and dying off behind as it grows
from the tip. The leaves are very crowded toward the growing cir-
cinate end. The whole plant is lax and lies in the still water in which
it grows entangled, one plant with another, forming often dense mats.
Toward the end of the growing season the more terminal internodes
become very short and the leaves densely packed to form a resting
bud (turion), which sinks or floats, according to circumstances, and
may be frozen in the ice. In spring the tips of the leaves or the
chief axis proliferate, giving rise directly to new plants with leaves of
simpler structure at first.

A leaf arises from a single lateral outgrowth from the prostrate
stem, remaining single in U. oligosperma, or forking as it develops
(Pringsheim, Goebel) to form two lobes {21 — 8), at whose bases
may arise secondary outgrowths, one on each of which can develop into
additional lobes. The third and fourth lobes are not at all or only
weakly developed in some species {U. vulgaris) but are strongly de-
veloped in others, e.g. U. Thonningii. In others fifth and sixth small
lobes are formed laterally to the third and fourth, and are known as
auricles. In U. Thonningii (Angola) the auricle is a fan shaped pro-
duction, with many rays from its edge, all armed with stiff bristles,
occasionally bearing a trap. In U. stellaris (Asia, Africa) and flexuosa
(Singapore) the auricles are deeply subdivided, the divisions more
or less curved and crowded upon each other. Each leaf lobe is pin-
natifid, the pinnae being, however, alternate, the internodes so
placed as to produce a zigzag axis, appearing monopodial {20 — i).
The lateral divisions (pinnae) have the same disposition of parts.
The ends of the divisions are usually armed with stiff bristles, either
singly or in bundles, and these afford a taxonomic character.

The traps are borne on the leaves in such position as to suggest
that each represents a leaf division (Goebel, Meierhofer). The
details of the structure of the trap will be given fully beyond. Here
is only to point out that the trap is always placed with its sagittal
section transverse to the plane of the leaf, the mouth facing the
apex of the shoot {20 — i). The interior surface of the trap is mor-
phologically the upper surface of the leaf division which it repre-
sents (hypopeltate, Goebel). In some species only one lobe of a
leaf produces traps, the other (upper) half being wholly photosynthetic
in function.

In certain species leaves of a highly speciahzed kind occur on
the basal part of the flower stalk. In these the midrib is much in-
flated by the enlargement of intercellular spaces while the lateral
divisions are much reduced. These are disposed in a whorl ("false
whorl") and act as floats to support the inflorescence above the
water surface {U. radiata, U. inflata, U. stellaris) {20 — 12). The

Chapter XIII

221

Utricularia, Bioviilaria

size and shape of these floats are characteristic of different species
possessing them. In U. stellaris they are short and of relatively
wide diameter. In U. injlata they are long (4-5 cm.) and clavate.

Fig. 6. — 1-7, Utricularia purpurea. — i, Early stages of germination of a seed {2>-2>
mm long) from which only one growing point arose (to be followed by others later in all
probability); 2, Three figures in a row, three views of an early stage of germination, a
later stage of which is shown in 3, in which two young growing points show circination;
4 and 5, Two following stages in the germination of a seed which produced three growing
pomts all of nearly the same age, with a fourth, secondary to the middle growing point;
6, A case in which fasciation occurred, the two figures on the right show an early stage
of germination, a much later stage is shown on the left, in which it is seen that one of the
growing points had divided, an abnormality; 7, An advanced stage in germination (15 mm
long), one of the three growing points still quiescent; five whorls of branches (the maximum
seen) were produced, as shown on the longer stolon of this figure. At the first whorl of this,
the bud of a branch stolon of indefinite growth is seen. No traps are produced on the two
dorsal branches of the third and fourth whorls.

8-10, Utricularia cleistogama. — 8, Early stage of germination showing primary stolon
and primary leaf, with the primary trap on the leaf near its base; 9, A later stage in detail
showing the origin of the trap from the leaf base; 10, A more advanced stage of the seed-
ling in which a second leaf arose in the place of a primary trap.

Still another form of leaf occurs on short stolons at the base of
the scape of the inflorescence (Buchenau 1865), the so-called rhi-
zoids, several of which are usually present. Their leaves are much
reduced in size and have very small pinnae which are curved and
crowded into claw-Hke masses. More or less of each pinna is densely
covered with mucilage glands quite like those scattered over the
whole plant surface {U. vulgaris, etc.).

Francis E. Lloyd — 222 — Carnivorous Plants

In the species U. oligosperma there is a pronounced dimorphism
of leaves, each kind occurring on separate branches. Those of one
kind bear none, or very few poorly developed traps. They are finely
divided with very long terminal divisions in the form of fiat, ensi-
form divisions, and are very crowded, so that the leafy branches
appear as dense tufts strongly contrasted with those which bear the
second kind of leaves which are trap bearing. The traps are very
numerous and crowded, and the leaves which bear them have only
a single lobe, whose divisions all He in the plane of the axis which
bears them.

The leaves in this and similar species {e.g. U. mixta) are provided
on one side only with a small fohose appendage consisting of about
four radiating accuminate members each bearing a trap, the whole
having a stipulate appearance (2j — 3). These may be similar to
the minute axillary shoots mentioned by Pringsheim in U. vulgaris
{21 — 10), different only in position.

A pair of similar appendages (we may call them tentatively dwarf
shoots) occurs also at the sides of the base of the air shoot, but here
they often have more divisions, each carrying a trap (23 — 2, 6),
They are found also in U. mixta. To the unaided eye in such cases
there appears at the base of the air shoot a tight grape-like cluster
of as many as a dozen, or as few as two to four traps. Without
careful examination it would seem as if the air shoot were supplied
with stipules, but the organs in question arise directly from the axis.
They, however, derive their vascular strands as branches from the
main strand entering the air shoot, though exceptions have been
noted. So far as I am aware these pecuHar dwarf leaves are to be
found only on those species which have strap-shaped stolons, the
greater longitudinal plane being vertical, the upper edge somewhat
narrower than the lower. They cannot readily be homologized with
the auricles of some species because of the distinct origin from the
stolon.

A large number of species of the U. vulgaris type {U. exoleta,
emarginata, gibba, cymbantha, etc.), plants with thread-Kke stolons,
have leaves which may be no more than a single fork, two slender
segments arising from the base, only the lower one bearing a single
trap (27 — 17, 18); or they may be variously more complex, but
always much simpler than in U. vtdgaris.

Branching; origin of the inflorescence. — It was quite apparent to
the earlier observers (Irmisch 1858; Buchenau 1865) that branch-
ing in Utricularia does not follow a pattern common to the flower-
ing plants. The observations of these students, with those of
Pringsheim (1869) and Goebel (1891), afford the available basis for
description to which my own studies have been added. From these
there emerges in fairly clear form the pattern peculiar to these very
aberrant organisms. We consider U. vulgaris which Pringsheim
studied developmentally, in contrast with Irmisch and Buchenau
who examined only the mature condition.

U. vulgaris. — The growing point (2/ — 8) of the horizontal shoot
shows upward circination (Pringsheim, Goebel) and that (i) a row
of peculiar branches arises in a line on the upper surface of the stolon

Chapter XIII — 223 — Utricularia, Biovularia

in no relation to the leaves except that they are usually nearer the
nodes than to the middle of the internode, and either in front or
behind them. They do not arise in a leaf axil. These, first seen
by BucHENAU and thought by him to be roots, are the "tendrils" of
Prlngsheim and "air shoots" of Goebel (23 — i). Gluck ques-
tions their usefulness as air shoots. They are long and very slender
with two lateral rows of "mussel-shaped" leaves with stomata, and
are circinate forwards. They are absent from many species, including
all of the U. exoleta type. They have the abihty to transform them-
selves by apical growth into ordinary shoots (Goebel, Glxjck).
(2) Lateral branches (st2) arise near the upper edge of the oblique
leaf insertion, but not in the leaf axil. They are of occasional oc-
currence only {21 — 8, st2). (3) The inflorescence (sc) arises in as-
sociation with a third stolon branch (sta) in the axil of sta, the latter
being the larger at first, the scape arising from its base. As Goebel
observed in U. flexuosa, "Never does one find the inflorescence iso-
lated, but always combined with a leafy branch springing from its
base." The question which of the two is primary cannot at the mo-
ment be settled. (4) In some leaf axils ("older ones," Pringsheim)
buds may arise, which do not center on the middle of the leaf axil.
Here again are two, one arising from the base of the other, and they
are a leafy shoot and a scape (the smaller) just as when an st2 branch
is present, but now in a leaf axil, not in the branch axil {21 — 9).
Pringsheim calls them dwarf or aborted shoots. (5) In the axil of
most leaves there is a cluster of about 4 traps arising at the middle
point. Goebel interpreted Pringsheim's observation and drawing to
mean that this group of traps arises from a short branch, while
Pringsheim thought that the branch arises from a trap stalk. Goebel
is probably correct {21 — 10).

U. stellaris, U. inflata and the Hke appear to conform to the above
description. In U. oligosperma and U. mixta somewhat more special
conditions prevail. Instead of one branch of st2 rank, there are two,
coordinate in development, one opposite the single leaf (undivided at
the base) and one below the axil. The lower of these {U. oligosperma)
bears leaves with very many traps, the upper is almost devoid of
traps. Above the axil of the upper branch arise the twin branches,
one a leafy branch, the other a scape, conjoined at the base {21 —
11). Thus in maturity there is a cluster of stolons radiating from
the base of an inflorescence, the forward and backward extensions
of the chief stolon, and three leafy branches. In both species the
air shoots are prominent and have laterally placed at their bases
clusters of traps which have been referred to already as dwarf leaves.
They occur also at the leaf bases, either axillary or at one side. There
is evidence that these are dwarf shoots (Goebel) bearing traps with
broadened stalks, not trap stalks bearing shoots, as Pringsheim
thought.

V . minor {21 — 16, 17) was studied by Irmisch and by Buchenau.
It is a smaller and more slender plant, the leaves, in lateral rows, be-
ing placed aslant, facing upward, the upper edge being therefore farther
from the apex than the lower. The branching is essentially as in
U. vulgaris, with the difference that the chief branch (st2) arises at

Francis E. Lloyd — 224 — Carnivorous Plants

the lower edge of the leaf, and is circinate upwards. In the axil of
the leaf there are two shoots, a leafy one (sta) circinate towards st2
and a scape (sc) circinate towards sca. sca and sc may, however, occur
without SC2 which is much less frequent in incidence. In the normal
condition therefore when the scape is developed there are at its base
two stolons, one from the chief axis and one from the base of the
scape. An additional one, apparently seen by Buchenau, may arise
from the scape base. My material came from Eire through Professor
H. H. Dixon.

In U. gibba {21 — 12, 20), on the other hand, the first branch
arises at the upper edge of the leaf base. A pair of mutually facing
and circinate branches then arise, as in U. minor, to produce a second
branch and scape. Secondary scapes arise in close apposition to the
primary as branches of the bearing stolon. A plant sent me by
Dr. F. W. Went from Pasadena behaves similarly as do also U.
emarginata and U. exoleta {21 — 18, 19). In these species the scape
produces near its base numerous branches, not in any leaf axil,
which bear much reduced and very glandular leaves. These are rhi-
zoids {2 J — 8). Secondary scapes may also arise in the axils of these
{21 — 20).

It is apparent that in the vulgaris type of Utricularia the branch-
ing has distinct pecuHarities. One sort of branch, the air-shoot, never
arises in any relation to a leaf. The chief stolon branch (st2) arises
near one edge of a leaf base, more or less overlapping the axil, but
never centered on it. At its base, opposite the leaf axil, arise two
buds, one a stolon and the other, on or near its base, an inflorescence.
In some species a dwarf shoot bearing only traps arises in the leaf
axil, behind the scape and its companion shoot.

The rhizoids are absent from some species. They are regarded as
anchoring in function, but are only very ineffectively so in the floating
species (Glijck). They are much better developed and are much more
numerous in the exoleta type.

In the mature condition the original position of the primordia is
usually completely obscured by the enlargement and mutual distor-
tion of the adjacent parts. The embryonic condition was studied
by Pringsheim and by Goebel, both of whom recognized the ori-
gin of the ''tendrils" or "air-shoots" and of the chief stolon branch.

The immediately above mentioned species {U. gibba, exoleta, etc.)
are, in contrast to vulgaris, very slender plants with thread-like stolons
and simple leaves, once or twice divided, or even thrice {U. eniar-
gitiata) {20 — 2). The internodes are long. Obviously closely related
to these are two African species, U. cymbantha OHver and U. Ste-
phensae (in Ms.) which deserve special mention. These are minute
plants with single flowered scapes. The method of branching is
simple. A stolon branch (st2), always single, and without axillary
buds, arises near the upper edge of a leaf, but more or less axillary,
while the scape arises from or near to the upper surface of the stolon,
and near or somewhat distant from a leaf and certainly in no definite
relation to it {23 — 12, 13, 21, 22).

Submersed, semi-submersed but anchored forms of the vulgaris type. — ■
These fall into two groups: those which grow (i) submersed but send

Chapter XIII — 225 — Utricularia, Biovularia

out shoots of two kinds, one chiefly trap-bearing; or (2) on
the surface of the wet substrate, sending out branches which pene-
trate the substrate, bearing traps and reduced leaves. These emerge
eventually. To the former belong such species as U. ochroleuca, U.
Bremii (Europe) and U. minor (in both hemispheres). To the latter
belongs U. intermedia (20 — 4; 21 — 13, 14). According to Glijck
they exhibit a good deal of polymorphism in response to environmental
differences. So far as they have been investigated the method of branch-
ing shows no peculiarities. For full accounts the reader may be referred
to Glijck's book.

Here may be mentioned the peculiar U. clandestina Nutt. (23 — i).
This is a lax floater of the general appearance of an undernourished
condition of U. vulgaris. It is provided with special branches with
reduced foliage and traps, these occurring sparingly if at all on the
leaves of the main stolons. The scape of the inflorescence arises in
connection with a branch (the latter in or near the axil of a leaf)
and, in addition to the normal inflorescence bearing normal flowers,
bears at its base usually two flowers in the axils of scales. The ped-
icels of these flowers nod downwardly and produce seed abundantly
by close pollination (presumably). Sometimes these cleistogamous
flowers (which never emerge from the water) are produced without an
accompanying scape bearing proper flowers. The presence of scale
leaves allows no doubt that the spur on which they are borne rep-
resents an undeveloped inflorescence. Its position with relation to
a leaf and branch are the same as above described for a normal in-
florescence.

In this species also air shoots are to be found, usually emerging
from the upper surface of the chief axis rather near to an inflores-
cence. They are absent from others of this group above cited.

A few forms are, with the exception of their inflorescences of course,
not only completely submersed, but their chief stolons are buried in
in the substrate of sand or mud, and their leafy branches or merely
their leaves emerge into the supernatant water. Among these may be
counted U. resupinata (N. America), U. biloha (Australia) and U.
paradoxa (in Ms.) (Angola).

U. resupinata {23 — 8, 9). — ^ The body of the plant consists of
horizontal stolons bearing terete, tapering leaves on the upper surface,
with branch stolons emerging laterally, a pair at each node. This
is a wide departure from what we have seen above, and foreshadowing
what we shall see in the terrestrial forms. The leaves are circinate
backwards, that is, away from the apex of the bearing stolon, as
first observed by Goebel in U. orbiculata. The inflorescence arises
as a bud in the forward leaf axil, flanked usually by stolon buds.
From the base of the scape a number of rhizoids spring out and
penetrate the substrate. The method of branching is the same in
U. hiloha and U. paradoxa {21 — 22), the differences being in their
leaves. In the latter they are much as in U. vulgaris but emerge
from the substrate and appear as little trees in the water. The
traps are borne chiefly on the stolons. In U. hiloha the leaves are
articulated, segment with segment {21 — 21). Sometimes a seg-
ment becomes a stolon, iUustrating the indeterminate morphological

Francis E. Lloyd — 226 — Carnivorous Plants

character of these parts often referred to in the literature (Goebel).
The traps occur on secondary stolons, rarely on the leaves.

Utricularia grows generally, when submersed, in still waters.
There are two very striking exceptions to this in U. neottioides and
U. rigida, the former South American, the latter from Africa, both
tropical. They grow in running streams, attached to the rocky bot-
tom, recalling the Podostemonaceae. Creeping on the rock surface and
tightly clinging to it are numerous fleshy, coral-like stolons. From
these arise branches which are leafy, bearing traps (Luetzelburg
1910) and finally flowers. The traps diverge from the vulgaris trap in
being streamlined — to yield to the vocabulary of the moment — being
spindle shaped, the stalk at one end and the mouth at the other. Ac-
cording to O. Staff (1906), U. rigida, which closely resembles U.
neottioides, has no traps. None of the Kew specimens showed any, and
though the herbarium specimens of U. neottioides did not show them,
Luetzelburg found them. The material I examined was collected by
him, and preserved in Goebel's collection.

Terrestrial and epiphytic forms of the vulgaris type. — Of the
strictly terrestrial species are, e.g. U. suhulata L. (W. Africa, Amer-
ica) and U. Rendlei Lloyd (Victoria Falls). These grow in a wet sub-
strate of sand or sandy soil, and consist of very delicate thread-like
stolons sending up simple spatulate or ligulate leaves of very small size
and often difJEicult to see when collecting, and having dehcate scapes
with yellow flowers. U. suhulata shows a cleistogamous condition in
Nova Scotia (Fernald). The method of branching is the same as that
in U. resupinata. The traps are numerous on the stolons, in lateral
rows and one row along the upper surface and along the leaf margins.
The leaves face away from the apex of the bearing stolon.

The epiphytic species are usually large and bear showy flowers,
and are often grown as greenhouse plants among the orchids with
which some of them vie in beauty. Mentioned here may be U. reni-
formis, U. nelumbifolia, U. montana, U. Humboldtii, U. longifolia,
U. Endresii and the small but often large flowered species growing
in the soil such as U. Dusenii and U. Campbelliana, all from Central
or South America. Some species grow in the water held by the leaf
rosettes of Tillandsias, e.g. tj. nelumbifolia in the Organ Mts. of Brazil
(Gardner 1846), and U. Humboldtii, on the Kaieteur Savannah,
British Guiana (Im Thurn 1887), both of which grow in the axils
of the leaves of bromeliads (Brocchinia spp.). These, and especially
the forms which grow in wet moss {U. reniformis), are conspicuous
for their thick, coral-like stolons, the anatomy of which has been de-
scribed by Hovelacque. The method of branching differs with dif-
ferent species.

U. reniformis is on the evidence of its branching related to the
terrestrial types, e.g. U. suhulata. From the 6 mm. thick stolon
the leaves arise in a row on the upper surface. These are circinate
backward, and have a reniform blade 15 cm. in diameter. The bud
of the scape arises in the leaf axil on the proximal side but not al-
ways in the middle point indicating a degree of obliquity in the
position of the leaf (Goebel). The branch stolons arise in single
lateral rows (21 — 15). Though the plant is of stately proportions,

Chapter XIII — 227 — Utricularia, Biovularia

the traps are small. There are no rhizoids judging from Hoehne's
figure.

U. montana has leaves and branch stolons alternating irregularly
in lateral rows, all lying in the horizontal plane of the bearing sto-
lon. Their axillary buds, however, He obliquely (Goebel). Branch
stolons may also arise from the upper chief stolon surface. The
condition here recalls that of U. cornuia, a small terrestrial species.

In U. longifolia leaves may occur both laterally and on the up-
per stolon surface (Goebel). U. Dusenii Sylv. is a small delicate
plant resembling U. reniformis in habit and flower structure, and
has the same disposition of lateral organs. Instead of a leaf with
attendant stolons, the node may bear three leaves. Rhizoids are pres-
ent.

The traps of all the previously mentioned plants adhere strictly
to the kind found in U. vulgaris. Some slight differences occur, but
these will be better described in a following chapter devoted to the
structure of the trap.

2. The Biovularia type. — In the only two known species of Biovula-
ria, the general morphology aligns itself with that of the vulgaris type
while that of the trap stands closer to the U. purpurea type. The
species are B. olivacea (Wright) Kam. and B. minima (Warm.) Kam.

Utricularia olivacea was described by C. Wright in Grisebach's
Plantarum Cubensium . . . (1866) and was regarded by him as closely
related to U. gibba, which it is not. Kamienski, who also related
it to U. gibba, made it the type of a new genus, Biovularia, based
on the number of ovules present in the ovary, namely, two, arising
from the bottom of the ovary and not from an enlarged central pla-
centa, as in Utricularia. We call it therefore Biovularia olivacea.

The plant consists of extremely delicate axes bearing traps on
long stalks in the place of leaves (23^14-18). The latter are ab-
sent, but Wright described the plant as having them. To quote
him: "utriculis obovoideis ad segmenta folii capillaceo-divisa spar-
sis." This error seems to have arisen either from admixture with
other floating species or from the fact that the long stalks frequently
shed the traps at the outer end, and thus appear as leaves. Still
they are not divided. When branching occurs one or two branches
may arise from a node. In herbarium material (Cotype, Herb.
Smithsonian Institution) I could get no evidence bearing on the se-
quence of development.

The inflorescence arises as a branch near the axil of a trap. This
branch assumes considerable thickness, and dominates, in the mat-
ter of size (diameter), the mother stolon. From its base arise two
branches, with a 120 degree angle of divergence, one somewhat higher
up than the other. From the apex of this short thick spur springs
a flower pedicel, which is surrounded at its base by an enveloping,
involucral scale leaf. Just within this may arise a second pedicel
in the axil of a second enveloping scale. In exceptional cases the
second pedicel may arise from a point on the first formed pedicel
a considerable distance above the base. A third pedicel may arise
from the second. We are evidently dealing here with a compound in-
florescence in which the chief axis is suppressed.

Francis E. Lloyd — 228 — Carnivorous Plants

The ovate sepals continue development during the growth of the
capsule, becoming deeply denticulate along their margins, and form
a graceful vase-like involucre about the ripened clavate capsule.

Usually only one seed develops. The narrowly ovate embryo (0,28
X 0.15 mm.) conforms to type, there being no organs differentiated.
It is invested by a loose and papery covering which probably remains
attached to the capsule.

Biovularia minima growing in Lagoa Santa, Brazil was sus-
pected by Kameenski (E. und P. VI, Lentihulariaceae) to be spe-
cifically identical with B. olivacea, both subsumed by him under
Biovularia. With the courtesy of the Botanical Museum, Copen-
hagen, I have been able to examine the original Warming type mate-
rial, and am now in a position to say that the two plants are specifically
distinct on evidence of flower structure.

3. The purpurea type {20 — 3). — To this type belongs a small
group of highly distinctive plants found only in the New World. They
are so far as known freely floating plants, and have no terrestrial
analogs.

The plant body consists of stolons which send out at the nodes
6 or 7 cylindrical branches forming very regular whorls. The whole
displays a minor degree only of dorsiventrality which, however, is
more evident at and near the growing apex, where the stolon apex
is upwardly strongly circinate, and the branches develop at une-
qual rates, faster below, slower above. The cylindrical branches in
turn can produce branches of the third order, more below near their
bases (about 4), fewer above (two), also in whorls, but unevenly
spaced. These branches are more definitely and evidently dorsiven-
tral than the chief stolon, and are of limited growth, and each branch
is constricted at the end into a slender stalk, bearing a trap. At
the base of each branch there is an abscission zone, as there is also
at the base of each branch of the third order and at the base of the
trap.

Barnhart correctly described this plant as having no leaves,
these being represented by verticillate branches. Luetzelburg, how-
ever, regards them as leaves. He studied a species which he called
U. elephas, which differs from U. purpurea in having only two to
four branches instead of 6-7. He examined the growing tip and
believed that he could see that the pair of lateral "leaves" were
united in the early stage of development. Goebel accepted Luet-
zelburg's interpretation. Had Luetzelburg examined U. purpurea
in the same way, the evidence would probably have given him pause,
since six or seven "leaves" would have to have been accounted for.
I have studied both U. purpurea and U. elephas in the same way,
and can find no evidence that any of the branches are fused at first.
This is borne out by the distribution of the vascular strands, which
radiate separately out from the central cylinder. Add to this the ver-
ticillate arrangement of the branches of the third and fourth order,
and it is clear that we have to do not with leaves (even in the re-
stricted sense this term is used when speaking of Utricularia) but with
branches.

In U. elephas as in U. purpurea the scape occurs in the axil

Chapter XIII — 229 — Utricularia, Biovularia

of a more or less aborted branch arising on the upper surface of the
chief stolon. This branch always remains delayed in development.
The scape produces no rhizoids, nor any scale except in the inflores-
cence.

4. The dichotoma-monanthos type. — To this type belongs a goodly
number of species which are purely Australasian, and so far known
only from Tasmania, Australia proper and New Zealand. This type
is not present in the recent Brass collections of New Guinea plants
at the Arnold Arboretum of Harvard University. They are at once
recognizable by their winged traps.

The series includes one freely floating form, U. tubtdala, and while
the terrestrial forms are readily divided into two groups, those with
runner stolons {U. dichotoma, U. monanthos, etc.) and those which
have only anchoring stolons {U. Menziesii, U. violacea, U. volubilis,
U. Hookeri), never runners, and which are confined to the extreme
S. W. of Western Australia. These have been regarded by Goebel
as primitive forms, but the only fact to which this view can be tied
is the absence of runner stolons. Allied and included with these is
the genus Polypompholyx, with 2 (or probably 4) species.

Freely floating species. — The only freely floating species of this
group known, and that only from herbarium specimens in the Mel-
bourne Herbarium (paratypes at Kew and at the British Museum of
Natural History), to which I had access, is U. tuhulata. It was col-
lected in 1875 by W. E. Armit in "mountain swamps near Cash-
mere, 40-50 mi. west of Rockingham Bay" in Queensland, but never
since. In general appearance it resembles U. purpurea, but only
superficially owing to the whorled position of the leaves (j(5 — 10, 11).

The "rather long" stolons bear leaves and traps in whorls, in each
whorl four leaves alternating with four traps on long stalks, so ori-
entated that usually two of the leaves he on one side and two on
the other side of the stolon, the traps being then one dorsal, one
ventral, and one on each side. Occasional departures from the rule
may be observed when two traps may stand side by side, or two
leaves. In the mature condition the leaves and traps are joined
at their bases to form a complete ring of tissue surrounding the node
from which they arise. A dissection of several terminal buds showed
clearly that the primordia of the lateral organs are all quite distinct
at first, so that the ring supporting them is secondary. The primor-
dia appear at first as low mounds of tissue in transverse series of
eight, at first indistinguishable from each other. At about the sixth
node the leaves elongate somewhat, overpassing the traps in growth.
The apex of the axis is long, naked and slightly circinate. The pri-
mordia are not at all crowded. In the axil of a leaf a bud which
develops into a branch stolon may arise. Traps with their stalks and
leaves attain a length of 2 cm. The leaves are flat, hnear and apicu-
late, the trap stalk foliose (Lloyd 1936c).

According to von Mueller, the scape is terminal on a chief
shoot and such evidence as I was able to obtain bears out this view.
I dissected one terminal bud to find that it was indeed an inflores-
cence with a lateral vegetative bud. Slender at the base, it swells
considerably at or above the middle to form a spar-buoy float. The

Francis E. Lloyd — 230— Carnivorous Plants

scales are basifixed. Mueller's description of the flower does not
help us much, but the few specimens I saw in Melbourne indicate
clearly that the flower with a widely spreading lower lip conforms
to that of U. dichotoma. Its color is bluish ("albida-caerulescente").
U. iuhulata as a floating plant appears to stand alone in regard
to the morphology of the leaves and traps. If the upper and lower
traps of a whorl were absent, we would be tempted to homologize
the two lateral leaves with the trap between them with the condition
found in U. gibba, but that would be pressing the matter too far.

U. dichotoma; U. monanthos (23—19, 20). — These and other
related species are characterized by the fact that the stolons dis-
play well marked nodes and internodes, the latter usually quite naked,
though in some cases {e.g. U. dichotoma) traps (facing backward)
may arise from the upper surface of the internode. At the node a
leaf, its upper surface facing backward, arises from the upper sur-
face and from each side of the stolon near the leaf base a branch
stolon and a trap. From the proximal leaf axil two traps and a bud,
which becomes a second leaf, usually spring, and from this axil also
a scape can arise. Thus these forms align themselves with the ter-
restrial forms in general, but are striking for the more readily ob-
servable emplacement of their parts. The traps are generally long
stalked, and in U. dichotoma often emerge slightly from the surface
of the wet but firm substrate, covered by a water film in normal
times. Hundreds of traps could be seen dotting the ground at Nar-
rabeen, N. S. W., using a lens of course. An additional feature of
interest in this group is the widely lacunate structure of the stolons
and petioles, which consist of scarcely more than the epidermis and
the vascular strand with a few collapsed parenchyma cells clinging
to it (much as in Genlisea). In U. monanthos, which grows in shallow
water, both stolons and petioles are much puffed up. U. dichotoma
has very small spatulate leaves and a tall scape; U. monanthos rela-
tively large leaves and a short scape. The scapes produce anchoring
stolons, leaves and traps on their bases.

Another group of species of the dichotoma type is composed of
plants devoid of runner stolons and consisting solely of a corm-like,
vertically growing axis springing directly from the seedHng (22 — 25,
26). The corm is very slender at the bottom where it emerges from
the seedling, widening toward the top, having below the structure
of a stolon, becoming more and more stem-like as in Genlisea (Warm-
ing). This puts out anchoring stolons, traps and leaves, and ter-
minates in an inflorescence. They are either annuals of small, very
deUcate structure {e.g. U. capilliflora, U. Dunstani, U. albiflora)
which grow in wet places during the rainy season, chiefly in N. W.
Australia; or much more sturdy plants, but of the same plan of
structure, all but one {Polypompholyx tenella) of S. W. Australia.
These latter may be annuals found in wet sandy soil {P. latifolia,
P. tenella) or in very shallow water {U. violacea, U. Hookeri); or peren-
nials in wet clay-sandy soil {U. Menziesii) or in water {U. volubilis).
All these with the exception of U. volubilis and U. Menziesii con-
form in morphological features to Polypompholyx, and are sufficiently
indicated in the figure of this genus (22 — 25, 26). General descrip-

Chapter XIII — 231 — Utricularia, Biovularia

tions have been given by Goebel in his Organographie. The perennial
species U. Menziesii and U. volubilis require some further description.

U. Menziesii {20 — 8, 9; 2j — 23, 24) was seen growing near
Perth, W. A. The plant body consists of a minute corm which grows
upwards, dying off below. From it spring hundreds of minute long
stalked traps penetrating the soil in all directions, those growing up-
ward coming close to the surface. The latter are covered by a ro-
sette of long petioled spatulate leaves, from the middle of which
emerges early in the wet season the scape (one or two) with unique,
conspicuously brilliant red, large-spurred flowers. My material allows
the inference that the plant begins its course by forming from the
seedling primary stolon an oval tuber penetrating deeper into the
substrate. From a lateral bud on this a small corm is formed, which
again produces penetrating tubers. At length a substantial corm is
formed which produces near the apex only laterally borne tubers,
two to four in number, regarded by Goebel as water storage organs,
tiding the plants over the dry season, which they undoubtedly do.
They contain some starch. The scape is always borne laterally,
and is not, as in the annual species, a finial of the corm.

U. volubilis {2j — 25). — I found this growing near Albany, W. A.
among the fibrous matting of a wet swamp. The young stages are
not known. The plant body is a stout upright corm, which grows
at the top and dies behind. It bears numerous fihform leaves about
3 cm. long and numerous long stalked traps, often with leaf-Hke
stalks. The scape is terminal, but the corm is continued by a large
lateral bud at the base of the scape. There are also produced long
anchoring stolons of strong texture, bearing traps in groups of three.
The scapes are very long, and twine about supporting reeds.

In the foregoing pages an account of the general structure of
the plant body has been presented, which practically covers all the
major varieties of habit. It is still insufficient for our present pur-
pose in not embracing all the types of Utricularia as indicated by
their kinds of traps. Those still to be mentioned include the spe-
cies U. cornuta, caerulea, capensis, orbiculata, longiciliata, and sim-
plex, taken as typifying large or small groups of species having traps
of pecuUar structure, to be mentioned beyond. While some of these
grow in shallow water, most of them grow in wet sandy soil, and all
have in common the general structure above indicated for U. sub-
ulata, with some slight exceptions, most of which need not here be
amphfied upon.

U. cornuta is an American plant and was described by Schimper.
Its leaves and branch stolons are borne laterally on the runner stolons,
with no very regular alternation.

U. caerulea represents a large number of Asiatic and African
species with leaves bearing numbers of traps and branch stolons.
Goebel (1891) has studied this type. I refer to the plant studied
by Goebel. There is doubt about its proper specific name
(Barnhart), but as I had access to the same material as used by
Goebel, I continue to use the name he used.

U. capensis is a good representative of a number of African and
Asiatic species, the latter including U. rosea and U. Warburgii, both

Francis E. Lloyd —232— Carnivorous Plants

studied by Goebel. This group is represented in the New World
by U. peltata which, Hke some cognate species in Africa, has peltate
leaves, their petioles bearing traps.

U. orbiculata, U. striatula, U. brachiata, U. multicaulis and pos-
sibly some others are minute Asiatic and African species, most epi-
phytic in wet moss on leaves, tree-trunks or on rocks. It was on U.
orbiculata that Goebel first noted the peculiar back-facing position
of the stolon leaves and the coordinate position of the axillary buds
(Goebel, 1891). This species produces a number of minute spherical
pearl-like tubers strung along its stolons, probably for water storage
(Goebel). U. longiciliata (America) is unique so far as the structure
of the trap is concerned. It is a typical small terrestrial species sim-
ilar in habit to U. subulata.

U. simplex (S. W. Australia) exhibits a rare peculiarity^ of pro-
ducing its scape directly from the margin of a leaf, a habit which
it may have in common with its relatives U. lateriflora (S. E. Aus-
traha) and U. calliphysa (Borneo) (Staff 19 14), another species
still undescribed from India, and one from Ceylon. Of over a hun-
dred complete plants of U. simplex {23 — 7) exhumed from a sandy
substrate not far from Albany, W. A. not one showed a different
origin of the scape, though specimens of U. lateriflora, from near
Sydney, N. S. W., showed the primary origin to be from the seedHng
in the usual way, as in U. Barnesii mihi in ms. {22 — 27). Their
traps are minute, of similar structure, but display specific difierences.

U. globidariaefolia and a few similar species are American. They
are terrestrial, rather large and become perennial by their stout,
tough stolons. Aside from their considerable stature, growing as they
do among the grasses and reeds of such habitats as the Aripo Savannah
of Trinidad, they display no striking pecuharities beyond the pos-
session of distinctive traps.

U. Kirkii is an African species with apparently few associates,
if any, and has a distinct form of trap. It is of the usual terrestrial

habit.

U. nana and U. Lloydii, small terrestrial species, are unique as
regards the character of the traps. Both S. American, each appears
to share its peculiarities with no other species yet known.

It will be the purpose of a succeeding chapter to consider the
mechanism of the trap and the various peculiarities of the various
kinds of traps characteristic of the above mentioned types.

— {References on p. 26 f) —

Chapter XIV
THE UTRICULARIA TRAP

General description of the trap ; terminology. — Historical account. — Anatomy and physi-
ology of the trap. — Two mechanical types of trap as regards the posture of the door (the
vulgaris-biovularia-purpurea type. The capensis-caerulea-cornuta type. The monanthos-
dichotoma-Polj'pompholyx type). — The variety of traps. — Digestion; the fate of prey.

In the account of the character of the various sorts of Utricu-
laria already given, their arrangement in groups or types was based
on the character of the traps. In order to explain the workings of
all of them we shall begin by a detailed examination of the longest
studied and best known, that of Utricularia vulgaris and its close
relatives. What we learn of this we may then use as a basis for com-
parison with other types.

The traps have been called urceoli, ampullae, vesiculae, utriculae,
pitchers, bladders, or traps. The most widely accepted term, blad-
der or vesicle, or the Latin, vesiaila, is not so bad as it seems, since
a bladder has an opening guarded by a valve in the form of a sphinc-
ter muscle which keeps it closed except under certain physiologi-
cal conditions when the muscle is temporarily relaxed. It was called
a pitcher by analogy with other carnivorous plants (Staff), but
this suggests a passive trap, and it is anything but that. Utricle
(utriculus), a small bottle with yielding sides (of skin or leather),
presumes a stopper. "Trap", used in this work, precisely fits, be-
cause the mechanism is that of an elaborate trap which is set automat-
ically and, after capturing prey, resets itself repeatedly, by observation
as many as fourteen times, and this is certainly not the limit.

A description, in general terms, of the trap. Terminology. — The
vulgaris type of trap is a small flattened pear-shaped hollow body
attached to the plant by means of a stalk placed laterally, and trun-
cated obliquely across the narrow end, where occurs the mouth of
entrance. The stalk side is ventral; the opposite dorsal. The edge
of the mouth carries in most cases a pair of branched antennae, and
the sides some slender elongated bristles (27 — 7). These form a sort
of funnel leading to the entrance, acting as guides for prey. In some
species these appendages are absent or much reduced in size {U. oli-
gosperma). Because of the flattened shape we may speak of the sides
and the edge of the trap. The sides may be convex or concave,
as first clearly recorded by Erocher, according to physiological cir-
cumstances. When the trap is set, they are concave; after action, they
are less so, and the trap has now a more rounded form. Scattered
over the outer surface there are numerous small spherical glands, de-
void of cuticle, which give off mucilage. These glands are common to
the whole plant surface.

The entrance {26 — i, 2) is guarded by two valves, a larger,
the door, and a smaller membranous one, the velum (Lloyd 1929).
The door is attached to the trap along a semicircular hne on the

Francis E. Lloyd — 234 — Carnivorous Plants

dorsal part of the entrance, its free edge hanging and in contact with
a firm, semicircular collar or threshold, against which the door edge
rests. The convex outer surface of the door bears a lot of longer
or shorter stalked mucilage glands, throwing off mucilage and sugar
(Luetzelburg), which have been said to be attractive to small animals
(Cyprids, Daphneae, etc.) and so to act as a lure. In addition it
bears four stiff, tapering bristles, based near the free, lower door
edge. These are the tripping mechanism. The surface of the thres-
hold, against which the door edge rests, is covered with a "pave-
ment epithelium" of glandular sessile cells secreting mucilage. Along
the outer edge of this pavement there is attached a thin but firm
transparent membrane, the velum, which lies against the lower edge
of the door, filling in the chink between this and the threshold.

The internal surface of the trap carries many glandular hairs
{26 — 9-13), with two or four projections, the former on the inside
of the threshold, the latter everywhere else {26 — 2). Darwin called
them bifids and quadrifids. The capital cells are devoid of cuticle.
The rest of the surfaces except at these points is cuticularized.

In size the traps, at their largest, are usually not more than 5
mm. long; in the majority of species, 3 mm. long and less. Their
small size has militated against readily understanding them.

An ample, partly incorrect description of the trap was furnished
by Benjamin in 1848. He recalled the more important earlier ob-
servations: Meyen had thought the traps open in the mature plant,
ScHLEiDEN thought the entrance was merely guarded by hairs.
Treviranus realized that the tightly closed door prevents the es-
cape of air when inclosed within the trap. De Clair ville said that
the door opens outwardly, but, as Benjamin pointed out, he failed
to see that, if this were the case, air could escape, but nothing could
enter. Benjamin himself clearly demonstrated that the door opens
inwardly — he could push it in with a needle — but not outwardly
— for if you push a needle against it in this direction it is torn. The
function of the traps — he called them bladders — he thought to be
connected with the supply of air to the plant. They were to him
air reservoirs, getting it from the water through the four-armed hairs.
De Candolle (1832) and van Tieghem (1868) believed that they
had to do with the floating and sinking of the plant in spring and
autumn. As pointed out by Goebel, the plants float just as surely
after the bladders are removed. What had not then been observed
is that normally the traps hold no air, but that this enters when
the plant is raised out of the water. As Cohn remarked, the
failure to understand the traps arose out of a wrong point of view.
He and Darwin adopted another only in turn to prove wrong. Cohn
recorded finding various forms of Daphnia and Cyclops in the traps
of a herbarium specimen. He then put a living sprig in an aquar-
ium where it grew rather feebly for some time. There was no prey
in the traps — none in the water. He then added some Ostracods
from a culture, and next morning many of them had been caught
in the traps. But Cohn's observations did not stand alone; the
brothers Crouan (1858) had recorded the presence of small beasts
in the traps. In America, in 1873 Mrs. Mary Treat and a coworker

Chapter XIV

235

The Utricularia Trap

found entomostraca in the traps of U. deistogama. Prompted by
this she made a careful examination to see if she could observe the
method of capture. She thought in 1875 that the animals "open the
door and walk in", agreeing with Darwin and with Cohn that prey
push in the door, which then closes and prevents escape. In her 1876
account she re\ased her conclusions, for she then found, in U . purpurea,
that prey is suddenly engulfed, as if drawn into a "partial vacuum".
Not seeing that the trap walls change their posture, she was ignorant
as to how the vacuum could be achieved; yet her idea foreshadowed
the discovery to be made in 191 1 by Brocher. Mrs. Treat learned
through Dr. Asa Gray, in correspondence with Charles Darwin,
that the latter was making similar studies; so that it is of interest to
see that Cohn, Darwin and Mrs. Treat, whom Darwin later quoted,
were arriving at similar conclusions at the same time independently.
She further saw evidence that larvae were digested in the course of
48 hours. '"'I was forced to the conclusion that these httle bladders
are in truth like so many stomachs, digesting|and assimilating animal
food", she remarked.

Fig. 7. — Copies of the original drawings of Cohn (left) and of Brocher, of the en-
trance of the trap of Utricularia vulgaris.

Cohn's and Darwin's conceptions of how the trap works were
identical as is shown by their descriptions. Cohn said that "the
valve is held against the threshold by a pressure of water within the
trap, but that it is easy to open by pushing it inwards. This ar-
rangement makes it understandable that living water animals, en-
tering the peristome, lift the valve and without difficulty enter into
the hollow cavity of the bladder, whence they cannot escape since
the valve opens only inwardly, not outwardly." And Darwin spoke
in the same manner, saying that "animals enter merely by forcing
their way through the slit-Hke orifice; their heads serving as a wedge."
GoEBEL accepted this explanation, as did Meierhofer and Luet-
ZELBURG. An impressive drawing by Goebel as well as that by
Cohn (Text fig. 7), though incorrect, are still used as illustrations.
It is clear that up to this time the trap was regarded as a passive
mechanism, the animal caught having to do the work of forcing en-
trance. We must add however that it was thought that the door
was either forced against the threshold by a "m a tergo", the water

Francis E. Lloyd — 236 — Carnivorous Plants

pressure (Cohn), or by its own property of elasticity, the latter im-
plicit in all Darwin's statements. It remained thus till 191 1 when
a Swiss entomologist, Brocher, became interested in Utricularia.
Pointing out that the view just above expressed is but an hypothe-
sis, since no one had actually observed what happened, Brocher
tried to do this.

In a series of experiments (Czaja does him the injustice of say-
ing he did none) Brocher established the following points, to his
own satisfaction. When an animal is caught it always disappears
very suddenly. Darwin and Busgen (1888) had all seen this
and recorded their observations, but had drawn no correct inference
therefrom. Further that, at the moment of this disappearance, the
trap gives a spasmodic jump, and widens a little, from which Brocher
concluded that the trap sucks in the prey (57 — 4, 6, 9). This observa-
tion was of fundamental importance. He was able then to explain
why, when a leaf is raised out of the water, the traps are often found
to contain air bubbles, whereas before they were absent. On lifting
a plant from the water he could hear a " crepidulation" (an observa-
tion made independently by others) explained by the swallowing of
air by reacting traps. He saw that in traps which had not reacted,
the sides were concave, but after reaction were flattened or slightly
convex. Finally he found that he could cause a trap to react by
''titillating" the door bristles with a needle point, and that when
this was accomplished there was each time a spasm of movement,
and a change in profile. These observations by Brocher, made with
exactitude, furnished a point of view which finally led to the correct
explanation of the workings of the trap.

Passing on to hypothesis, he supposed that the collapsed form
of the trap is explained by the principle that the rate of develop-
ment of the tissues, being quasi superficial, is greater than the rate of
expansion of the volume. To the extent that the walls are depressed,
the tensions of their tissues are augmented and thus they try the
more to take a normal position, that is, to obliterate their re-entrant
curvatures. The walls are therefore in a position of unstable equilib-
rium, during which the interior is in a state of "negative", that is,
reduced pressure. The proof of this is the fact that, when punc-
tured, the walls take up the normal position, dilating to a maximum.
This could not be possible if the structure of the trap is as repre-
sented in the textbooks, he remarked at this point, since a simple
check valve could not preserve the reduced pressure. He further
supposed that the door is strongly curved, especially transversely,
and that, because of the curvatures of the wall, it is held firmly against
the lip, and, with the addition of mucilage, is thus rendered a water-
tight valve (Text fig. 7). In order that the equihbrium thus pre-
served may be upset, Brocher assumed that the door is endowed with
a certain "sensibility" and "contractihty", so that, on touching the
bristles, it can shrink a little, and thus allow the water pressure to exert
its force. An animal doing this would be swallowed with the inrushing
water. That minute fish are sometimes caught by the tail shows that
it is not because they try to get in, but that merely by the flick of the
tail, they have stimulated the trap. The action of the door or "oper-

Chapter XIV — 237 — The Utricularia Trap

culum" is so rapid, Brocher observed, that it closes before the walls
can more than partly expand, so that the trap may act again, but
this remained questionable. It was admitted that the contents of
the vesicle might be absorbed by the "rhizoids" (quadrifid hairs),
in which case the walls would again be drawn in, and the trap re-
sensitized. But Brocher, not being a botanist, was too modest to
undertake to solve this part of the problem.

Ekambaram (1916, 1918, 1926) in India made observations on
the traps of U. flexiwsa (similar to U. vulgaris), which substantiated
those of Brocher above mentioned, though apparently in ignorance
of this author's work. That is, Ekambaram recognized the two states
of the trap, one with concave and one with convex sides, and that
in the latter, when the "irritable" hairs are touched by a prowling
animalcule, it is sucked into the trap with the inflowing water. It
had been noticed by him also that when the whole plant is lifted
from the water there can be heard "light crackling sounds like the
ticking of a watch" and this was referred to the action of the traps
when released, presumably by water films. When pushed in by the
water, the door becomes inverted and boat shaped, with the "irrita-
ble" hairs folded up into the groove {26 — 7). The movement of the
door he considered to be due to the momentary loss of turgidity, as
quickly regained leading to closure, but he does not offer any evidence
for this. The irritable hairs he mistakenly thought to have the same
structure as those of Aldrovanda.

Ekambaram was able to reset the trap by carefully pressing out
the water by compressing its sides, but it does not appear that he
understood that the trap can automatically reset itself. The escape
from the walls of intercellular air during this operation must have
been accident, and can have no bearing on the matter. Merl found
the contrary.

At about the same time Withycombe, a British student, announced,
in 1916, "that the bladders of Utricularia ... 2lXQ not passive traps,
but that they capture prey by active movement in response to stim-
ulation. A bladder becomes sensitive to contact after its walls be-
come concave on each side. Then, on touching certain short hairs
at the mouth of the bladder, the lateral walls spring outwards, be-
coming somewhat convex, and so drawing a current of water into the
bladder which swept with it, of course, any body sufficiently hght to
be sucked in." Again, this observation was made quite independently,
as Withycombe learned of Brocher's work only ca. 1922 through
Merl. Nor did he yet know of Ekambaram's observations. In his
paper of 1924 Withycombe, thinking Brocher's explanation of the
working of the trap valve inadequate, agreeing, however, about the
matter of "negative" pressure and its results, advanced the idea that
the edge of the valve or door, instead of being merely pressed against
the collar or threshold (Brocher), is caught in a groove from which
it can be released only by an upward movement. This groove stands
in front of the zone of specialized cells (see beyond) and is as deep
as these. "A certain amount of mucilage is secreted apparently by
the middle layer, and this makes a complete watertight fitting of the
valve." Here is a specific attempt to account for the hermetical seal-

Francis E. Lloyd — 238 — Carnivorous Plants

ing of the door mechanism, and though as it will appear a mistaken
one, the idea was correct. The internal water is absorbed by the quad-
rifid hairs, so that the trap can be reset by setting up anew the
strains expressed in the convexity of the side walls. In his experience
this required about 30 minutes. The action of the door is due, he
says, to its irritabihty, and the slender four hairs inserted in the door
are the only organs which can be stimulated. Irritability, however, was
not proven to exist.

Merl's work, above mentioned, appeared two years before
Withycombe's second paper. He set out from Brocher's important
observation that on stimulation the walls of the trap expand, drawing
in water in the capture of prey, but further showed that the operation
can be repeated again and again. During three days he observed the
trap to act thirteen times. Merl correctly determined also that
traps which contain some air can react, contrary to Brocher's view
(but this could happen only if the bubble of air in the trap is not too
large!). It is only if, owing to the shape of the trap, the bubble can
be moved or distorted, that this can happen. The time required for
resetting in U. flexuosa was found to be a minimum of 15 minutes,
but full resetting requires about 30 minutes. In U. purpurea {jy — i)
it takes upwards of two hours (Lloyd 1933a). It was shown by Merl
that the full expansion of the trap sides takes place when the door is
forced open or when the wall is punctured. The reverse of this, the
sucking in of the side walls, is more pronounced the longer a period
of non-stimulation, until of course the cohesion of the internal water
sets a limit. Apropos of Brocher's note to the effect that on removal
of a plant out of the water a clicking sound was noticed due to the
swallowing of air by the traps, Merl was able to do this without
setting off all the traps. Some of them did not react and remained
unaffected under a bell glass. He was then able to procure the re-
action by touching the bristles. Aside from furnishing him an argu-
ment against Brocher's theory that the compression of the trap walls
is due to "atmospheric and hydraulic" pressure (Merl's statement
concerning this view seems incorrect) the experiment shows that it is on
general grounds not surprising that some species are not submersed,
species the traps of which normally exist and act in moist air, sur-
rounded by wet moss, detritus or sandy soil. The action of the traps
on lifting from water is therefore due, it is suggested by Merl, to the
action of water films on the bristles of the door and not to the mere re-
lease from water pressure.

Merl then tried to determine whether the reaction of the trap,
or specifically of the door, is an irritable response. He could not
procure reaction by wounding or by electrical stimulation. As to the
temperature relations he found that the traps reacted as long as they
remained alive, and that by chemical means no condition of rever-
sible inactivity (rigor) could be induced. Incidentally he found that
the trap is so completely sealed by the door that there is no entrance
even for dyes, such as eosin and methylene blue, so long as the dyes
do not induce death of the trap. Nevertheless Merl could not quite
rid himself of the feehng that the mechanism is irritable, and would
have adopted this view if so much evidence "had not spoken against

Chapter XIV — 239 — The Utricularia Trap

it." Among this evidence, he found that during action there was no
disturbance of the air in the intercellular spaces, which would occur
if there was an extrusion of water into them such as occurs in irri-
table tissues. In spite of inimical evidence, however, Merl inclined
to think that the bristles are irritable hairs analogous to those of
Dionaea and Aldrovanda (as had Ekambaram and Brocher). He
proposed, however, the only alternative theory, a purely mechanical
one. The four-armed hairs withdraw the water from the interior
of the trap, thereby setting up a tension, the walls responding to
the draft by cohesion of the water. The highly elastic door, the free
edge of which rests firmly against the threshold, opposes this draft
and comes into a position of labile equilibrium, which must be dis-
turbed "by the slightest movement or by shrinkage" of the (door)
cells, to allow the walls to retract into their relaxed position. Even
now he could not quite exclude a certain irritabihty as a capstone
of the bridge. This view was to be championed later by Kruck.

Working at the same time, independently of Merl, Czaja exam-
ined the problem of the Utricularia trap. His publication was but
a trifle later than Merl's. Proceeding from the same point of at-
tack, Czaja agreed with Merl that the trap could repeat its action,
and could reset itself in a short period of 15 to 30 minutes and that
the reaction (on suitable stimulation) takes place very suddenly.
The concave sides then became much less so. The door in this re-
action opens to a narrow slit, and closes as suddenly as it opens
(neither of which, however, is quite true) allowing the entrance of a
stream of water. The process is released by touching the bristles.
By chemical means Czaja could not decide definitely on the nature of
the mechanism, and this left him for the moment at the same point
as it did Merl. With respect to the anatomy of the trap, he exam-
ined first the closure of the trap by the door, in order to settle the
question of the path of the internal water when the trap is exhausted
as it must be when the walls pass from the less to the more concave
posture. He determined that the entrance is hermetically sealed. The
proof consisted in inserting a fine hair beneath the door edge, when
the trap could not again set itself. When the hair was withdrawn,
again the trap became effective. Further proof was supplied by the
fact that Congo red and methylene blue never entered healthy, but only
damaged, traps. All this he beHeved points to the membrane, or rather
the wall of the trap, as important.

As had been demonstrated by Cohn in 1875, the walls of the trap,
if set free to act by removing inhibiting structures (the threshold and
contiguous walls), will expand. Because of their structure and the tur-
gidity of their cells they always strive to take an outwardly convex
form. Substances which can reduce their turgor (5% KNO3) put the
trap out of commission. On the other hand substances which cannot
penetrate but which withdraw the water from the trap (glycerine,
cane sugar) can up to a certain limit of concentration reset the trap
but if in too great concentrations, cause its total collapse. The reset-
ting of the trap is therefore the result of withdrawing water from its
lumen, and not of direct participation of the walls which would in-
volve turgor changes.

Francis E. Lloyd — 240 — Carnivorous Plants

This is made possible by the tight application of the door edge
to the threshold (specifically, to the layer of epithelium on the top
of the threshold) enhanced by the mucilage which has an added
sealing effect. Since the cells of the wall are not plasmolysed by
glycerin etc., water is not withdrawn from these cells but only from
the lumen without changes in turgor. This is what happens in na-
ture. The setting of the trap results from the withdrawal of water
from the lumen. The only agent for this is the action of the four-
armed hairs. It is allowed that some water may penetrate through
the walls inwardly but at a slower rate than that at which it is thrown
off, for otherwise it would lead to overtension, and this, he held,
would bring the trap into a condition unfavorable for prompt action.

CzAjA holds that the withdrawal of water results not only in the
change in position of the walls, but that this results in turn in a
cramping effect on the door, forcing it against the threshold more
tightly and so effectively increasing its watertightness, an idea held
by Brocher, but which is untenable in the hght of the structure of
the walls, which are thin, acting as hinges near the threshold {26 —
3). When the trap is in the set posture, the walls concave, and the
door tightly in contact with the threshold, the bristles stick out at
an angle in such position that on touching them the edge of the door
is disturbed and a narrow opening is formed between the door edge
and the threshold, through which water is drawn in by the expanding
walls. The action is mechanical. In support of this Czaja records
that it is easier to fire the trap if the bristles are swept from above
downwards than transversely to this direction. This can mean only
that the leverage is more effective in disturbing the door edge when the
levers are moved in one direction than another and rules out mere irri-
tability. Firing the trap is due to the deformation of the door edge
and the consequent lifting of it from the threshold allowing water
pressure to act. For the rest, Czaja did much experimentation show-
ing that the trap is surrounded by a selectively permeable membrane
but Prat (1923) found that the entire plant is protected by this
membrane.

Czaja was the first to take a definite stand that the trap action
is mechanical, aside of course from the water-extruding power of the
walls, and the general condition of turgidity. That is, the springing
of the trap is purely mechanical. This was opposed to the views
of Brocher, Ekambaram and Withycombe, and to Merl insofar
as he allowed the question to hang in the balance. Hegner (1926)
(not knowing of the work of Brocher, Merl or Czaja) made inde-
pendently the observations as to the method of catching prey re-
corded by Brocher, noting its rapidity, but did not venture into
the question of the method of function of the bristles. Thus Czaja
was left the sole champion of the view that the whole capturing
action of the trap is mechanical, but he was not to go unchallenged;
for Miss M. Kruck in 1931 undertook to prove the contrary, but,
to state it abruptly, she quite failed (Lloyd 1932&). In the first
place her presentation of the structure of the trap was askew and
it was patent from her figures that she did not grasp the anatomical
facts. The drawings showed initial and final positions of the door,

Chapter XIV — 241 — The Utricularia Trap

after stimulation and reaction, which simply do not occur. The
physiological evidence consisted in the observation of the extent and
position of intercellular spaces in the tissue of the door before and
after response. To meet this I reproduced photographic evidence
which showed clearly that such changes do not occur, though it is
evident that changes in the mere distribution of air might occur
without vitiating my evidence. Again Kruck stated that the cells
of the door change their shape, but this was found equally illusory.
In the course of response she stated that the bristles lose water,
supplying a stimulus to the neighboring cells of the door which re-
spond in like manner, with the result that the shape of the cells
of the door changes through loss of turgor, but evidence for this
was quite absent. She further allows 15 minutes for the restitution
of irritabihty, in this agreeing with Czaja and Merl, both of whom
allowed this as the time necessary for the withdrawal of sufficient
water to set the mechanism. If Kruck was right, it is not clear
why the restoration of irritabihty should not proceed when a trap
is punctured, but this never occurs. Most impressive is the fact shown
first by Ekambaram, repeated by myself, that by careful expulsion
of the water from a trap, it may be reset repeatedly without allow-
ing time for the restoration of irritability, unless, to be sure, an im-
mediate restoration is predicated. The claim that the bristles are
irritable was shown to be not true by first killing them with iodine,
after which they could procure response on touching (Lloyd 1932ft).
It should here be recalled that Withycombe observed that this re-
sponse could take place even in traps which had lain for a half-hour
in Bouin's picro-formal solution. It seems clear that Kruck failed
in supporting her contention. The evidence points to the contrary,
that the action of the door is purely mechanical, always granting
the turgidity of the component cells, devoid of which they could not
give to the door the necessary properties.

The walls, because of their activity in excreting water from the
trap lumen, are an important part of the mechanism. The total
amount which a trap throws out amounts to 88 % according to Hegner
(1926), much less according to Nold. Such figures are in any event
not important since the total amount excreted depends on the type
of trap. In U. purpurea it must be much more than in U. vulgaris.
That they do excrete water is all we need to know to explain the
action of the trap, and this was first demonstrated by Brocher,
later independently by others, Ekambaram, Withycombe, Hegner,
and possibly Hada (but who had seen Hegner's paper). It may be
emphasized, however, that this action can go on when, as a result
of the introduction of much food material, including salts, in the
form of the bodies of water animalcules, the osmotic pressure of the
internal fluid reaches a considerable but never measured figure. This
cannot be overdone, however, for if glycerine be introduced (Merl)
water is then drawn into the lumen. Experiments show that the
trap works within wide limits in nature. Nevertheless the phys-
iological properties of the walls remained a subject of inquiry, and
this has been pursued by Czaja and by Nold. Czaja's conclusion
was that the trap is surrounded by a selectively-permeable mem-

Francis E. Lloyd — 242 — Carnivorous Plants

brane, the cuticle, which excludes solutes. The four-armed hairs of
the internal surface absorb water more rapidly than it can find its
way in. Nold (1934) had advanced the theory that the potential
difference existing between the outer and inner surfaces of the traps
accounts for the movement of water outwardly. He localized this
difference in the only parts of the walls free of cuticle, namely the
outer spherical and the inner four-armed glandular cells, the only
ones through the walls of which the water can pass. That this shall
pass outwardly, he believed, is assured by the difference in pH at
these places. At the inner surface this is 6.2 and for the outer 6.6,
determined by the Folin colorimeter; or 7.5 and 8.2 with the quin-
hydrone electrode, differences which seem non-significant for water
movement. Nold seems to have shown, however, that the loss of
water from the trap increases inversely with the pH. of the outer
medium, the normal behavior taking place at ^H 5-7. The traps
are damaged at lower and higher pH values. Yet it has been shown
that Utricularia can prosper in water of ^H 4 (Emil Wehrle,
1927) and U. minor in "weakly alkaline water" (Nold). In any
event, since a difference of potential between inner and outer sur-
faces is known to cause a water loss, but since also organs are known
which show such differences without water movements, it is scarcely
possible to regard Nold's hypothesis as proven. This judgment is not
weakened by inspection of the evidence advanced.

There is a further point in the mechanism of the trap about which
opinions had been expressed previous to 1929, without the provision
of proof. I refer to the method by which the watertightness of the
door is procured. That watertightness is a necessary condition for
the successful action of the trap was first recognized by Brocher,
and by his successors in investigation, all of whom placed faith in
the contention that it is due to the tight appHcation of the door
selvage to the threshold, aided by the mucilage present. Withycombe
realized the inadequacy of this explanation, and supplemented it by
arguing that the door edge rests against the outer edge of the "middle
layer," the pavement, seeing in this a valvular seat. An examination
of the action of the trap and certain details of the emplacement of
the door led me to suspect that the explanation was a lame one.
This led to the discovery that the entrance of the trap is guarded,
not merely by one valve, the door, but by two, the door and a second
valve, the velum, attached to the threshold and finding its seat against
the door selvage, thus blocking the chink. This second valve has
been seen in some 75 species, in slightly various form to be sure,
but always present {24, 25). This discovery led to a minute examina-
tion of the structure of the trap in all material available from various
parts of the world. The results of this survey, made on many living
species and on still more preserved ones, he in the field of anatomy,
which in the presentation thus far has received only minor mention.
This is now to be taken up in the following, in which it will emerge
that Withycombe was quite right in principle if wrong in his under-
standing of the mechanism. I myself erred similarly in 1929.

The physiological anatomy and histology of the trap of Utricu-
laria vulgaris and closely related forms will now be considered. Within

Chapter XIV — 243 — The Utricularia Trap

the limits here imposed it will be practically impossible to show in
detail the contributions of the several investigators to our knowl-
edge in this field, and it must suffice to indicate critical observa-
tions. It may as well be said that the study of the anatomy of the
trap is by no means easy, if we desire to have exact knowledge of
the emplacement of the various parts. This is because on cutting
the trap, the tissue tensions are disturbed and the parts (especially
the door) disarranged; and it is necessary to know the exact rela-
tion between the valves (door and velum) and the threshold. This
cannot be finally determined by the study of the traps which have
been cut, though useful evidence can be got this way, but only by
the examination of the entire, healthy organ. An accurate descrip-
tion must be based on living turgid material, and errors have been
made by placing faith on paraffin sections. Again, the presence of
mucilage makes the trap slippery, and the knife, which must be very
keen, readily slips, so that to make a true sagittal section is diffi-
cult and this has led to mistakes. The much used figure first pub-
lished by GoEBEL in his 1891 paper is wrong for this reason, and the
figure used in a recent (German) edition of the Bonn textbook is
equally wrong.

Accounts dealing with our knowledge of the anatomy of the trap
(U. vulgaris and closely allied forms) are those of Benjamin (1848),
CoHN (1875), Darwin (1875), Hovelacque (1888), Dean (1890),

GOEBEL (1891), MeIERHOFER (1902), LUETZELBURG (1910), EkAM-
BARAM (1916), FRANfA (1922), MeRL (1922), CzAJA (1922), WlTHY-

COMBE (1924), Lloyd (1929), Kruck (1931) and Nold (1934)- Dur-
ing the prevalence of the earlier view that the role of the trap was
wholly passive, the results of investigation fell far short of adequacy
in the presentation of details of structure later found to be im-
portant. This period ended with Luetzelburg in 1910. With
Brocher's discovery in 191 1 attention was concerned more and more
with these details, though not always with sufficiently critical ob-
servation, and sometimes with the entire lack of it. This appHes
particularly to the entrance structures, more so to the door, of which
the special features began to be appreciated only with Withycombe
and Merl.

The general features of the trap have already been described.
Broadly speaking two regions are to be considered alone and in re-
lation with each other, the walls and the entrance mechanism. The
appendages (antennae etc.) are of less importance and will be de-
scribed in a comparative study of the various types of traps.

The walls. — ■ In the species before us the walls are composed of
two courses of cells, the outer and inner, both clothed with a thin
cuticle on their exposed surfaces. In general the outer course cells
are smaller in surface extent than the inner, in the ratio of about
three to two, linear dimensions. The relative thickness of the two
courses varies. Along the profile of the trap, the inner cells are the
deeper, but this relation is reversed on the sides of the trap, where
the outer cells are deeper. This is connected with the movement
of the walls from convex to concave, the outer cells suffering increas-
ing compression during the excretion of water. Nor is the total

Francis E. Lloyd — 244 — Carnivorous Plants

thickness of the wall the same everywhere. Under the entrance the
threshold, a part of the wall, has a thickness at the top of four or
five parenchyma cells plus the epidermis on either side. The threshold
extends upward on both sides to form the '' collar "-Hke thickening
which stands out from the wall in shelf-like fashion. Beneath this
shelf the side wall is attached to the threshold, and is here quite
thin, so that the wall swings here as on a hinge, thus not bringing
any torsion on the threshold and door. This structure excludes the
theory advanced by Brocher, that the walls help to cramp the door in
position {26 — 3).

Chlorophyll bodies occur in both courses, perhaps somewhat fewer
in the inner, but not absent, as Nold has said. Anthocyanin often
occurs in the inner course cells, but is absent from young traps and
increases with age after once appearing. Interspersed with the larger
epidermal cells are smaller ones, more numerous in the inner epidermis,
the basal cells. These bear each a short cutinized cell, the "middle"
cell, bearing two to four glandular, non-cutinized cells to form a cap-
ital. In the outer course, the gland is spherical, of two cells. On the
inside each middle cell bears two or four elongated cells. Darwin
called these hairs the bifids and quadrifids. The former are to be
found only on the inner face of the threshold; the quadrifids else-
where all over the inner surface. In U. vulgaris two of the arms are
reflexed, and the whole is tilted towards the entrance to induce inward
movement of prey, it may be argued. In U. gibba and allies all four
arms extend radially, but two are shorter (those toward the entrance),
and with more spread. These quadrifids are also tilted toward the
entrance. The bifid hairs, forming a chevatix de frise on the inner face
of the threshold, appear to be there to discourage prey from working its
way toward the door. In these hairs each arm is a cell terminating
proximally in a slender stalk. The two or four stalks are united to
form a single short round stalk basing on the middle cell {26 — 9-13).
The arms are not cuticularized, and absorb dyes very readily. They
are generally regarded as the organs of absorption which take up
digested food materials, and at the same time secrete ferments and acid
to accomplish digestion. The function of the spherical glands of the
outer surface is more in question. These hairs may belong to the
category of hydropotes (proposed by Mayr 191 5), the function of
which is to absorb water in submersed plants, the general epidermis
being cuticularized. In the case of the Utricularia trap the function of
water excretion seems likely a reversal of function which may be
determined by the greater activity of the quadrifids in absorption,
these presenting much more surface to the surrounding medium.
That the function of hydropotes may be the excretion of water has had
some support, cited by Meyer (1935). In one form or another both
these kinds of hairs are common to all species of Utricularia. In 1931
Kruck questioned Czaja's contention that the water, when being
excreted by the trap, escapes through the cuticle and therefore the
whole significance of his results from examining the permeability
relations of this membrane. On her part she held that the internal
water is absorbed by the quadrifids, and excreted by the spherical
glands of the outer surface. In proof of this, which she contends is

Chapter XIV — 245 — The Utricularia Trap

convincing, she claims to have followed the path of dyes from the
inside of the trap, which she saw to enter through the quadrifids and
escape from the outer surface glands, and made the observation, in
agreement with this view, that the quadrifids take up the dyes more
readily than the outer glands. Her method of experimentation was
(a) to lay the traps in the solution, and (b) to fill the traps with the
solution. As however she does not tell us in detail how the latter was
managed, one hesitates to accept her observation without reserve.

That the quadrifids are active during digestion was observed by
Darwin and by Goebel. Darwin's experiments showed that sub-
stances in solution (urea, ammonium carbonate, infusions of raw meat)
are absorbed by the quadrifids, but not by these alone as he found the
spherical hairs of the outside surface to do the same, as also the
mucilage glands about the entrance. He realized and admitted that
his experiments were not critical, but they indicated the importance of
the problem. Goebel detected the presence of fat droplets after
feeding, and Schimper noted, in U. cornida, appearances in the ab-
sorbant hairs (here bifid) which had been absorbing food different from
those in traps which had not been fed. The protoplast showed activity
which he compared with Darwin's aggregation, saying that the proto-
plasm swells and the vacuole is broken up more or less, as he observed
also in Drosera and Sarracenia. Less constant in occurrence were
yellow granules or droplets.

Later Kruck, in the paper already cited, presented her results of
study of the cytological changes which are to be seen in the glands in
various conditions of rest and feeding. Her observations he in the
field of cytology and are open to various interpretations. At any
rate they need not concern us here.

The trap wall is traversed by vascular bundles which branch from a
single strand which enters by the stalk. On reaching the trap it
divides into two branches, one of which goes forward around the
longer edge to the entrance where it ends abruptly. The other branch
moves toward the threshold, on reaching which it branches, each arm
following beneath the threshold and ending at one extremity. Xylem
is present but is very meagre.

Far more compHcated in structure is the trap about the entrance.
The opening arises in the very young trap as a sUt caused by the in-
vagination of the rounded primordium. The lips of the slit turn
inwards, the upper becoming the door and the lower the threshold
(Meierhofer). Two conditions are found. In one (as in U. vulgaris)
the wall of the trap bends abruptly in to continue as the door {26 — 6).
In the other {e.g. U. gibba, minor) the wall extends forward to form an
overhang, the door springing away from its inner under surface (26 —
2). In any case, from the edge of the fold arise the antennae, stout
branching emergencies springing from the upper hmb of the opening,
right and left {2/ — 8). The arrangement of these together with their
curvature produces a pair of drift fences funnelling toward the en-
trance, thus serving to guide prey to their doom. This condition is
found in U. gibba and a good many other species, in all of which the
antennae are curved forward and downward in front of the entrance
and are strong prominent appendages. The branches of the main

Francis E. Lloyd — 246 — Carnivorous Plants

trunk of the appendage are long uniseriate hairs. In other species
{U. vulgaris americana, U. oligosperma) the antennae are much smaller
and curve upward, away from the entrance. There are often no
antennae in these species, and there are others from which they are
always absent {U. nana) {28 — 5). In still other species, as will be
seen later, quite other arrangements are met with. In the water-
dwelling species, while they can evidently be regarded as elaboration
of the trapping mechanism, it must be said that their absence does not
seem to make any practical difference in the number of prey cap-
tured. In the mud dwelling species, they may serve to keep the
entrance free from detritus, and so help in preserving the effectiveness
of the trap. In the wet sandy soils and in wet moss, where the water-
is not continuous, such arrangements may be important in keeping
capillary water, in which prey may move, in contact with the entrance,
so that when the trap acts it does not draw in air. The capillary
action in such cases is helped by the mucilage secreted by glandular
hairs in large numbers attached to the door itself, and to the sides of
the entrance.

After the two Hps are laid down during the development of the
trap, the sides of the entrance extend, moving the lips apart so as to
produce a funnel-shaped approach. These sides, called by Cohn the
cheeks, are continuous with the overhang, when this is present, to
form a sort of hood or "vestibule" around the opening {2j — 9).

From each cheek, and from the edge of the overhang, springs an
oblique row of long uniseriate hairs, about four on each cheek, and two
or three from the overhang (27 — 7).

It is only in the front of the opening that the hps are drawn apart.
At their free edges they remain close, and in the final stages of develop-
ment are in mutual contact. At their lateral extremities they are
continuous, though their anatomical character changes. Another
important feature is alteration in the form of the lower lip. Though
transverse at first, it becomes finally semicircular in shape and thick-
ened by the growth of additional layers of wall cells beneath it to form
a massive thickening and strengthening of the wall in this zone. This
structure so produced was called the collar by Darwin, and the abut-
ment by GoEBEL. In this account it is called the threshold. By its
form and strength it preserves the shape of the opening, and resists any
cramping effect (said to occur by Brocher and Czaja) of the dis-
tortion of the walls when the maximum of internal water has been
withdrawn. Measurements made by myself did not reveal any differ-
ence in form in the set and the extremely relaxed condition of the trap
after puncture. In fact, the walls where they articulate with the
threshold are thinner than elsewhere, so that they can bend without
exerting distortion on it {26 — 3), besides which is the fact that the
inner part of the threshold is supported free of the wall, so that this
cannot press upon it.

The structure of the threshold in detail is best understood first by
an examination of a transverse section through its middle point (25 —
I, 2, 5; 2g — 4) and then by viewing it from a point of view which
embraces the whole inner surface, flattened out for convenience of
study {25 — 9) . In the transverse section the threshold is roughly

Chapter XIV — 247 — The Utricularia Trap

triangular, the base forming the free surface, the apex continuous with
the wall. The free surface is sHghtly convex, with broken curves
indicating three regions, an outer, continuous with the cheeks, carrying
scattered stalked glandular hairs; a middle, clothed with a layer of
densely crowded glands, called the pavement epithelium by Goebel,
and an inner of epidermal cells, forming a shelf projecting into the
interior of the trap.

The outer region is part of the vestibule, and we may think of it as
a doorstep. The inner region is merely a part of the inner wall sur-
face, but re-entrant. The middle region is of critical importance. We
shall use Goebel's name for it, recognizing however that the surface is
not epithelial but consists of closely set glandular cells which arise from
the epithelium below. It is a pavement of packed tiles, each tile being
the capital of a glandular hair. We pause here to recall the structure
of the glandular hairs in Utricularia. Arising from a basal epidermal
cell, each consists of a middle cell (Goebel), strongly cuticularized,
short and discoid in shape, supporting a glandular capital of one, but
more usually two cells, sometimes four (quadrifids), uncuticularized
(BiJSGEN 1888). The middle cells may be supported on a shorter or
longer tubular extension of the epidermis cell wall, as is the case of the
hairs surrounding the entrance. Those of the pavement are similar
to the glandular hairs of the general outer plant surface, but differ in
having capitals elongated, at right angles to the axis of the entrance,
so that, on looking down on the threshold, the pavement appears to be
made of closely packed sausages. The capitals may be one or two
celled. Each gland arises from a laterally compressed epidermal cell,
so narrow that the middle cell lies tightly against the neighboring
ones. The terminal cells are similarly tightly packed, forming the
visible pavement (Gislen 191 7) (C/. various figures on 25 to zg).

Like the glandular hairs in general, the pavement glands loosen
and shed their cuticles, but most curiously in a single piece, except in
the inner zone {2g — 4). To describe this behavior we have to recog-
nize three zones of the pavement epithelium, outer, middle and inner.
In the outer zone, the cuticles of its glands enlarge into balloons, but
remain attached mutually and to the glands which bear them. In the
middle zone, broadest at the ends, the cuticles remain mutually at-
tached, but are freed from the capitals which produced them, and
from the inner zone glands, but remain attached to the ballooned
cuticles of the outer zone. The glands of the inner zone behave
individually, their cuticles enlarging and bursting. There is formed
in this way a membrane, which I call the velum, consisting of two
parts, a cushion of cuticular balloons running from one end of the
pavement to the other on the outer zone, and attached to it a thin
membrane, bearing the markings of the capitals which produced it, the
two together forming a valve which, stretching from one side of the
pavement to the other, overlies the door edge (25 — 4-8). The inner
zone is lenticular, broadest at the middle, and scarcely reaching the
outer ends of the pavement. Its glands are larger and not very
tightly packed. The middle zone, entirely free of cuticles, presents a
soft yielding surface into which the door edge can sink somewhat
under pressure.

Francis E. Lloyd — 248 — Carnivorous Plants

We consider now the door or valve. This is a flap, two cell courses
in thickness, forming the upper free edge of the entrance opening, and,
in nature, bulges outwardly (2j — 2, 5). If it is removed by cutting
along its line of attachment to the trap wall and is allowed to lie in
water, it retains the shape it has in situ as Bijsgen observed (1888).
It is, if we disregard minor curvatures, roughly semicircular, the
shorter side being the free edge. For the sake of description we may
flatten the door and then map out certain regions, shown in the dia-
gram {2Q — 13). A wide zone around the edge of attachment is the
hinge region, where strong reverse flexures occur when the door is
opened. The middle region of this zone is the upper hinge, the two
lateral the lateral hinges. The upper hinge is characterised by marked
flexures when the door is at rest. The hinge area surrounds a lentic-
ular middle area, which may be called simply the middle area. At
the lower part of this a small circular patch of the door is quite thin,
and this is the central hinge. Out from just below this project four to
six stout, curved, tapering bristles. That part of the door below the
central hinge is thick and strong. This is the middle piece. Towards
the flanks, the door selvage becomes thinner. With this terminology
{26 — i) we can more easily describe the histology.

As above said, the door consists of two cell layers {2j — 2, 5; 2g —
I, 2), an outer, and an inner. The two are very different in structure,
the general relation between them being that existing in a bimetallic
strip, one of the metals having a greater index of expansion than the
other; the former under changes of temperature is active, the latter
relatively passive. The cells of both layers are equally turgid, but the
inner is capable of ready expansion and contraction of its inner surface,
the outer not. This is ascribable to the differences of structure. The
door has been described as highly flexible and elastic, as for example
by Darwin. Highly flexible it is, but if by elasticity we mean ex-
tensibihty, this adjective does not apply. The tissue has a sort of
cartilaginous quality, bending without breaking in any direction. If
the door is freed in part by cutting a median strip, releasing this from
the pull of the sides, it will spring outwardly and only on plasmolysis
can it be brought back. This shows that the door as a whole is always
normally insistent in pushing outward, and is held in its proper position
only by virtue of its semicircular attachment to the trap wall. As
Benjamin showed, it can be pushed in- but not outwardly. When
fully inwardly inflexed, it is folded along its middle line, becomes some-
what concave, and the tripping bristles then lie in the groove of the
fold {26 — 7), as Ekambaram described it.

The inner reaches of this attachment, that is, the lateral hinges,
coincide with the inner ends of the threshold, the extreme end of the
free door edge coinciding with the inner angle of the threshold. The
outer surface of the lateral hinge therefore lies against the outer reach
of the threshold. But the free edge of the door, starting from the
inner angle of the threshold, passes obliquely across it, the angle be-
tween the face of the door gradually changing till, in the middle reach,
it stands obhquely on edge {26 — 2). Only the middle reach of the
door selvage is thickened and stiff; the outer reaches are thin.

The rest of the attachment extends along the wall of the trap, curving

Chapter XIV —249— The Utricularia Trap

around from one end of the threshold to the other in a semicircular
sweep. Here the door curves inwardly at first, to form the upper hinge,
below this outwardly to form the bulging middle area. The curves of
the upper hinge are most pronounced in front, and are reduced at the
sides. The outward spring of the door depends on the physical prop-
erties of the upper hinge chiefly. The lateral hinges resist this pull,
but can bend passively.

If now we examine the histology of the door we find (29—1, 2),
on inspecting a section in any direction, that in general the outer
course of cells is thin, the inner thick, in the ratio of about one to
three, differing from place to place. In the central hinge they are of
nearly equal thickness, as also in the middle piece. The cells of the
outer course are all flat, their anticlinal walls zigzag (Bijsgen) and
these walls are strongly supported by buttresses at their angles. These
prevent their collapse under bending. This layer must put up with as
much bending as the outer, but passively. Many of these cells bear
glandular hairs {24—1; 26 — 2), some with pyriform, some with
spherical capitals. Those with pyriform capitals are scattered over the
upper part of the door surface, and are shorter stalked as one ap-
proaches the middle point of the door. Near the door edge, arranged
in a crescent parallel with it, is a row of glands with short stalks and
globose capitals, a quite large one at the middle point. Just above the
level at which this stands there are hairs of different structure. These
are stiff, tapering, sharply pointed bristles, four in number, standing at
the angles of a trapezoid, in an oblique posture, extending upward,
then curving delicately (25 — 3, 5). Each is composed of three to
five cells, the basal the shortest, the terminal the longest. Ekam-
BARAM described short hinge cells, like those in the sensitive hairs of
Aldrovanda, but this is a mistake. They anchor in the outer cell layer,
by a broad base, as correctly shown by Meierhofer, without any
bulbous insertion as Merl showed {26 — 4). These four (or in
U. flexuosa six, Ekambaram) hairs constitute the tripping mecha-
nism of the trap. A touch of these in any direction but, according
to CzAjA, best from above down, causes some distortion of the middle
piece of the door selvage. This distorts the door edge from its equal
seat, upsetting the dehcate equilibrium, and permitting the water
pressing against the door to push it in, assuming the trap to be
properly set. As we shall see, the tripping mechanism shows a wide
variety of form in the genus. That just described is found only in the
vulgaris type and in Biovularia, though in the latter the door has a
different plan of structure.

When facing the door, the outer course of cells presents a plan as
follows. In the region of the upper hinge, the cells are isodiametric,
very wavy walled, with strong buttresses (Cohn 1875). Their walls
lie athwart those of the inner course cells. The same is true of the
middle area. In the central hinge they are very small, corresponding
in size to the cells of the inner course. The same is true of the middle
piece, where the cells are very small and their walls are strongly
fortified with broad and thick buttresses. These cells, however, while
small when seen en face, are deep, and equally deep with those of the
inner course. Along the selvage to the outer hmits of the lateral

Francis E. Lloyd — 250 — Carnivorous Plants

hinge, the cells are elongate, and have numerous very small but-
tresses, difficultly seen. That is to say, the cells in the regions of
maximum bending have zigzag walls with many buttresses. Where
the door is stiffest, i.e., in the middle piece, the buttresses are at a
maximum in numbers and size. In these cells also the walls are thick,
especially the outer.

Looking at the inner face of the door, we note a different pattern.
With the central hinge as a center, the inner course consists of elon-
gated cells radiating from this center to this circumference {24 — 9).
The closer to the center the shorter the ceUs become, so that at the
center they are isodiametric and thickly studded with buttresses.
Below the center, the cells of the inner course of the middle piece
are also isodiametric and match the outer course of cells in the degree
of buttressing. From here, tending toward each side of the door, the
cells become longer and run along the selvage parallel to it. This
seems at first glance simple enough, and it seems surprising that the
cells of the whole inner course should have been more than once
described as isodiametric. The mistake is easily explained, for when
the inner surface of the door is examined without flattening it out,
to do which it must be fully plasmolysed, a series of concentric lines
can be seen {24 — 9; 2g — 3). Darwin saw them. They were cor-
rectly understood first by Meeerhofer. They are nothing more than
an optical effect arising from the fact that the inner cells are constricted
at regular intervals (29 — 1-3)- Wherever the constrictions meet the
side walls of the ceUs, these are here buttressed by props. In sections
the spaces between the constrictions are usually taken for single cells,
a mistake which I made myself at first. Within the central area
these lines, indicating the constrictions, run with great regularity from
cell to cell. In the region of the outer and lateral hinges they are
equally present, but are less regular. In the central hinge and in the
middle piece they are also present, but are here quite irregular and
numerous and are only with difficulty observable.

The effect of these constrictions is to render the outer wall of the
inner course of cells readily compressible, like a bellows, without injury
to the cells. Without them it is hard to see how so great flexibility of
the door tissues combined with firmness and quick reaction could be
attained. It is indeed, as Meierhofer exclaimed, a "most wonder-
ful" arrangement. In the upper hinge the constrictions are not so
deep as in the middle area but are more numerous, which may be a
better arrangement for the maximum bending which this has to en-
dure.

It may be pointed out that these cells have been represented (by
Ekambaram and Meierhofer) as having their anticlinal walls con-
stricted like the periclinal. This is not the case. It is true that if a
door is torn from its moorings and laid in water for examination the
injured cells along the torn edge will collapse and their uninjured
neighbors will swell and present the picture recorded by these authors
{24 — 9). But this cannot occur when the door is in situ and un-
injured.

We now consider (a) the way in which the door edge lies in con-
tact with the threshold, so that it can maintain its posture in spite of

Chapter XIV — 251 — The Utricularia Trap

the water pressure it must sustain when the trap is set, and (b) how the
water is prevented from leaking under the door edge. Recalhng the
structure of the threshold and especially that of the pavement, it is
necessary to point out that the latter along its middle reach is curved
in such a manner that it slopes somewhat, so as to face the interior of
the trap. At the bottom of the slope, where the inner zone begins,
there is an abrupt change in the direction of the slope so that a slight
depression is produced (29 — 4). Here the pavement is most closely
packed. The middle reach of the door edge is, as we have seen,
strengthened so as to make a firm edge, which rests against the pave-
ment just in or beyond the depression, its outer selvage surface resting
more or less against the pavement, according to the amount of strain
produced by water pressure. When the water pressure is greatest,
that is, when the trap is fully set, the position of the door edge is
more nearly normal to the pavement than when the trap has just been
released. This can be inferred from the measurements of photographs
of traps before and just after "firing" (Lloyd 1932&) {29 — 11). It is
indicated also by the position of the bristles, which are more erect
when the trap is in the set condition. It may be added here that the
whole shape of the trap is altered a little by the change in postures of
the side walls. Since the ends of the door edge coincide with the inner
angles of the threshold, it follows that its lateral reaches cannot follow
the pavement parallel to its midline. It is only its middle reach (the
middle piece) which impinges edgewise on the pavement. The lateral
reaches merely lie with the outer surface of the selvage flat against the
pavement, thus forming a re-entrant sHt through which the water must
leak under pressure unless this contingency were provided against,
which is the case. The cuticular membrane, the velum, attached to
the outer zone of the pavement, is slung completely across from end
to end of the threshold (25 — 5-8). When the door swings outwardly
after springing the trap, it pushes against the velum which folds
against the door {2j — 4), covering the re-entrant shts on the sides and
blocking the door edge in the middle. When the door is in position,
the velum reaches in front up to the short spherical hairs which stand
in a curved row just below the level of the tripping hairs (24 — i ;
2j — 3-5)- Experimental proof that the velum thus blocks the
entrance by its valvular action, consisted in cutting the side reach.
This was accomplished with a very small knife several times and the
parts carefully examined afterwards for assurance that no other damage
had been done (Lloyd 1932&). In no case after the velum was cut
did the trap reset itself. In Czaja's experiment in which he thrust a
hair beneath the door edge, this not only held up the door edge but
depressed the velum also, but this escaped his attention.

As CzAjA found, the distortion of the door edge (and that of the
velum at the same time) when it rests on a hair, prevents the traps
from working. This does not seem to be the case if the entrance is
filled with the soft body of a large capture, sufficient to plug it {20 —
11). Matheson (1930) states that such prey may eventually be
ingested, indicating that the trap, plugged by the animal's body, still
evacuates its water. In the meantime the prey may be softened and
respond to the sucking action when re-established, and thus eventually

Francis E. Lloyd — 252 — Carnivorous Plants

be drawn in. Mr. J. H. Buzacott writes me that this has been ob-
served at Meringa, Queensland, {20 — 11) where tadpoles of Bnfo nia-
ritius, imported to control insect pests in sugar cane, have been
destroyed in numbers.

On the sucking-in of prey: — The fact that large prey (young
tadpoles and fish fry, worms etc.) can be caught by some part of the
body, usually the tail, has long been known. After a tadpolette has
been trapped but not completely engulfed, it has been stated that the
body is later sucked in. This would obviously be limited by the
volume of the trap. The question has been raised by me, does this
sucking-in actually occur and if so what is the mechanical procedure?
I have recently taken pains to get evidence on this point. I employed
Utriciilaria aff. gibba, sent me from Pasadena, Calif., by Dr. F. Went.

In a series of experiments young mosquito larvae about 2 mm. long
were used. By manipulation it was possible to get one caught by the
tail, the head being too big to enter the trap. When this occurred,
one half of the body was instantaneously engulfed, leaving four or five
joints behind the thorax projecting beyond the mouth of the trap.
The joints served as clear-cut measures. Several cases were observed,
and all followed the same pattern. One example will suffice. The
larva was caught by the tail, the door clamping down between the
sixth and seventh joints, while six remained protruding (11:30 hrs.).
At 18 hrs. only the thorax and head remained protruding. Next
morning, the thorax had also been engulfed, the head only, too big to
enter, being left outside. (Text fig. 8, A, B). Since the body of the
larva prevents the door from assuming its normal set posture, and
though the trap walls did not become concave, as observed from time to
time, it must be inferred that nevertheless the entrance was sufficiently
occluded by the door and larva so that the exhaustion of the water
from the inside of the trap could proceed, creating a suction on the
prey from time to time, and drawing it gradually in. As I did not see
this happening during a prolonged period of observation, I cast about for
more suitable experimental material. This I found in fine shreds of
albumin, made by stirring egg-white in boiHng water. These were soft
and of fairly even caliber. Of a goodly number of experiments I choose
the following. Case i (Text fig. 8 C, D). A shred about two milli-
meters long was presented to the trap by touching the tripping bristles
with its end. One half of the shred was swallowed; the rest remained
protruding. On this some bits of rust, detached from the needle point,
adhered, serving as marks. When examined 18 minutes later, the entire
shred had been taken in. In the meantime the experiment was re-
peated (Case 2) (Text fig. 8 E-H) and kept under close and continuous
observation. Immediately after the door had clamped down on the
partially engulfed shred, the latter was seen to sHde slowly in for two
minutes, when it stopped (F). This movement was the result of
residual wall action. Nothing further happened for about ten minutes
(during which time a partial reduction of pressure within the trap took
place) when the door opened and closed rather slowly (the movement
was quite visible to the eye) and another portion of the shred entered
(11 43 hrs.). By 16:00 hrs. the shred had been entiredly swallowed (H).
The walls were now concave, and the trap, completely reset, reacted

Chapter XIV

— 253 —

The Utricularia Trap

to touch on the tripping bristles. In still another case, a very slender
bit of a swallowed shred remained protruding, and the walls of the
trap had become concave. The delicacy of this protruding shred had
permitted the door to take the set posture, allowing full exhaustion of
the internal water. On stimulating the tripping hairs, this shred was
swallowed.

It is evident from these experiments that: i. When the prey is
soft and yielding, but, caught part way in, is large enough to prevent
the door from taking the set posture, this still may clamp down
enough to enable the trap walls to bring about a sufficiently low
pressure to exert suction and thus draw in the body of the prey and, by
repetition, finally engulf it, if small enough.

E F G H

YiG. 8. — The sucking in of prey by the trap of Utriciduria gibba or a related species;
A, an injured mosquito larva was presented and suddenly but only partly swallowed,
May 21, 1941; B, the same three days later; C, a shred of albumen presented and partly
swallowed at 11:20; D, at 11:38, the shred entirely swallowed (the insert figure indicates
the edge of the door looped over the soft and yielding prey, the surface of the threshold
indicated by the broken line); E, a shred of albumen presented at 11:35, and the trap
observed continually; F, at 11:37, part of the shred indicated by a double arrow had been
slowly sucked in; G, at 11:43, the door quickly (but not very suddenly) opened and
another portion of the shred (/) swallowed; later (H) the remainder was engulfed.

2. If the prey is slender and yielding enough, the door may assume
a sufficiently exact set posture to insure the full setting of the trap,
when it will react normally but, of course, only in response to move-
ment of the tripping bristles. When the prey, still not engulfed,
dies, it may not be swallowed unless the tripping bristles are touched
by some other agent. If a stiff, unyielding object such as a hair is
used (as did Czaja), this cannot happen because of inleakage of water
at the side of the hair.

Traps which have captured mosquito larvae (the head remaining
protruding) do not survive, dying in 10 days or so, evidently from
overfeeding.

We now resume our discussion of the structure of the trap. In
addition to the front view of the velum, which can be seen in the living
trap, a side view can be had under favorable conditions, when it is
seen that the velum forms a bolster in front of the door edge (25 — 3,4).

Francis E. Lloyd — 254 — Carnivorous Plants

It is of interest and mere justice to record that the velum had been
seen previously by two observers, both in 1891. Giesenhagen, I be-
lieve it was, made a drawing for Goebel's paper of 1891 of a trans-
verse section of the threshold of U. flexuosa. In this drawing the velum
was shown in the clearest manner, but no mention of it was made in
the text. And in 1891, at Cambridge (England) R. E. Fry, a student
who was later to become an eminent art critic and Professor of Art
in that university, and who is better and more widely known as
"Roger Fry," prepared a Ms. which was never pubhshed, but to
which I fortunately had my attention drawn when attending the
International Botanical Congress in 1930. It was lying on the shelves
of a bookshop. Roger Fry was evidently a close observer, for in one
of his drawings, meticulously executed in fine pencil and color, he
showed the velum, and in his description he described the pavement
epithelium (he used Goebel's 1891 term), saying that "the whole of
this secretes mucilage, the cuticles of the hairs being raised in a mass;"
but he did not examine further into the matter. One cannot help
wondering why others, who saw other minute details, failed to see the
velum. Roger Fry's Ms. has now been deposited by me in the
Library of the School of Botany, Cambridge University.

Two mechanical types of trap. — Having described in some detail
the structure of the trap of Utricularia vulgaris, it must now be pointed
out that, though working according to the same mechanical principles
and being of the same morphological type, there are two distinct kinds
of traps (Lloyd 1936c). They can be distinguished readily by the
posture of the door in its relation to the threshold (Text fig. 9). If we
consider the entrance as tubular, in one kind the tube is short, in the
other long. U. vulgaris has a trap with a short tube entrance. In it
the door stands approximately at right angles to the axis of the tube,
or at any rate forming a wide angle with it. In the other kind, of
which U. capensis is a good example, the entrance is tubular (Ste-
phens 1923), and the door stands obhquely, forming a narrow angle
with the axis. Considered as a valve, this is the less efficient, ceteris
paribus, but its inferiority is compensated for in various ways, to be
noted. Of the latter kind there are two variants represented by such
species as U. monanthos, and U. dichotoma, on the one hand and Poly-
pompholyx on the other, all purely Australasian types, with differences
demanding separate description.

The description of U. vulgaris above given will serve as a standard
of comparison. Correlated with its position in the short tubular
entrance, the shape of the door is such that its sagittal measurement
is less than its transverse. The top of the threshold is narrow. In
U. capensis, with a long tubular entrance, the door has reversed
measurements: it is longer than broad, and the threshold is broad
(29 — 12, 13). The door stands obhquely. A glance at the diagram
(Text fig. 9) will reveal these differences. It is seen that, considered
as a check valve, the long door, presenting a re-entrant angle with the
threshold, and with no opposing seat, is, with respect to the direction
of the water pressure, at a disadvantage. In our blood vessels the
valves, which are also obliquely set flaps, are in the reverse position.
But from the point of view of the efficacy of the trap, the door would

Chapter XIV

265

The Utricularia Trap

be useless if it were set in this way. In the trap it must be able to
resist the water pressure to which it is normally subjected, until a
trigger action is applied, when it must then be weak enough to fold
up, allowing the entrance of a water column. And it must of course
not allow the leaking in of water when the trap is set. How these de-
mands are met may be understood by examining the structures in-
volved. For models of the U. vulgaris trap, see p. 266.

The door is divided into a relatively thin anterior half, and a
thick posterior half {31 — 3 etc.). The former includes the areas
of the upper hinge and the middle area. The two cell courses of
the upper hinge have the structure seen before, with deeply constricted
walls in the inner course {31 — 9). Those of the middle area, not a
region of sharp bending, have about equal thickness, with no con-
strictions. There is no central hinge and below the mid-point of the

Fig. 9. — Diagrams of the entrances of Utricularia vulgaris (i, 2) and U. capensis (3 riglit)
or caerulea (3 left ) embodying the different mechanical conditions in these two types of
trap; pd, general direction of thrust of the door, and, pt, of the threshold; r, relaxed position
of the door; a, point of impact of prey; PH20, pressure of water against the door; iz, slope
of inner zone of the pavement epithelium; mz, slope of middle zone of same; d, thrust of
the door; Ih, thrust of the lateral hinge; c, composition of these thrusts.

door lie the middle piece and the lateral hinges. Their total thickness
is usually greater than elsewhere, and in the lateral hinges the outer
course cells are thin and the inner thick and are constricted with many
constrictions {30 — 6-8). As these merge into the middle piece the
two cell courses become nearly equal in thickness and the walls are
thick. The threshold is broad and semicylindrical in form (jo — 5-8).
The outer third as seen in sagittal section (the "doorstep") bears
glandular hairs, the middle third of pavement cells supplying a volu-
minous velum of balloon cells, and the inner third being dense pave-
ment {2j — 2). The whole is surrounded by the massive trap walls,
giving firm support. The lower part of the door, when closed, rests
cramped into the relatively narrow arc of the threshold, exerting a
firm pressure by its middle piece {30 — 8). In the set posture the
upper part of the door assumes a convex form, thereby increasing the
pressure of the door selvage on the pavement, widening the angle
between the two {30 — 3). Just after release, the door, now in the
relaxed posture, has its upper part convex. It is watertight in this
condition. As water is withdrawn the upper half of the door becomes

Francis E. Lloyd — 256 — Carnivorous Plants

more and more concave till the set posture is reached. This account
has been substantiated by a photographic record of silhouette of living
traps in the set and relaxed condition of two related species. Seeds of
U. Wehvitschii collected by Young in Angola were grown for me at
the Edinburgh Royal Botanical Garden in 1934 and the traps studied
there. U. capensis was studied alive at Capetown later, and the results
were pubHshed in 1936 {24 — 5) (Lloyd 1936&).

Another, an Australian species, U. lateriflora, typical of a small
group of species distributed in S. E. Asia, and Australia, having very
small traps less than i mm. long, yielded to experimental methods
(1936c) and the results are shown in 33 — 9, demonstrating that the
behavior is quite like that in U. capensis and U. Welwitschii. The
living material was available at Sydney, N. S. W.

The same behavior is displayed by U. caerulea (Asia) {24 — 2) and
by U. cornuta (N. Amer.) (jo — 3) in both of which the living trap
was studied.

In all these, when the trap is in the set condition the outer selvage
of the door rests on the pavement, held there firmly by the thrust of
the lateral hinges. The wide angle between door and threshold is
filled by the massive velum, preventing inleakage of water. A thrust
on the tripping mechanism, the kriss hair (p. 259) in U. capensis and U.
Welwitschii, a group of sessile glands in U. caerulea and U . cornuta,
disturbs a dehcate balance of forces in unstable equilibrium, and the
trap is ''fired."

Both U. monanthos {24 — 3) and Polypompholyx {24 — 8) act in
the same way, and they also have been studied in the living condition.
U. monanthos was grown for me in Edinburgh in 1934 (1936a) and
Polypompholyx could be examined in 1936 at the University of Western
Australia at Perth near which it grows. The structures involved are,
however, to be considered separately.

U. monanthos {j4 — 1-5)-- — In this and allied species, the thresh-
old is very broad, front to back, and near its inner limit is bent,
curving downwards. Beyond the bend lies the dense pavement
which receives the middle piece, which is therefore applied on the
inside of the bend. This looks like a pretty poor arrangement, yet
it works. The major zone in front of the bend is occupied by an
ample velum which arises also from the walls projecting in front of
the door. Here is formed a complete massive ring resting against the
bulge of the upper part of the door when in the relaxed posture. When
in the set posture, the inner portion of the velum arising from the
pavement alone continues to block the entrance of water. The door is
still longer than in U. capensis etc., but the middle piece is relatively
smaller, and the middle area is correspondingly large, occupying about
four-fifths of the door length. When in the set posture, the whole of
this large area is concave, so that the sagittal curve is now continuous
with that of the middle piece, which by virtue of the thrust of the
lateral hinges is impressed against the dense pavement just inside the
bend of the threshold. The trigger consists of a group of sessile hairs
just above the bend of the door. The action when the trap is fired is
like that in U. capensis. It must be confessed at this point that my
earlier account of door action (1932a) based on preserved material of

Chapter XIV — 257 — The Utricularia Trap

U. Hookeri was wrong. This species conforms in every way to U.
monanthos. Living material was examined in Sydney, N. S. W.

In Polypojnpholyx the case is again quite special, for here the door
is as broad as long, but works as in U. monanthos. The whole trap
to be described (p. 262) is extremely curious. Because of the thickness
of the walls and other parts and the masses of glandular hairs on the
door and on the floor of the antechamber, it was difficult to study
the trap in action, and especially to photograph it. Nevertheless the
attempt succeeded {24 — 8). When the trap is set, the door shows a
simple curve, along the sagittal line from the upper hinge, which is
very thick and does little bending, to the edge which lies just within
the ridge of the pavement. When relaxed, just after discharge, the
lower two-thirds of the door is convex, the upper hinge showing little
movement — a slight bending in its distal zone only. It is evident
that the very deep cells of the outer course of this tissue exert a
strong tangential pressure on the lower parts of the door, ensuring a
tight apphcation of the selvage to the pavement when the door is
relaxed and a still tighter application when the trap is set.

The variety of traps. — The following account, necessarily brief,
will give some idea of the diversity of structure and form displayed by
the traps of Utricularia, Biovularia and Polypompholyx. We may con-
veniently follow the grouping into those having short and long tubular
entrances.

Traps with a short tubular entrance. — These are found in the U.
vulgaris type, in U. Lloydii Merl and U. nana St. Hil., in a group of
few species represented by U. globulariaefolia, in Biovularia and in U.
purpurea and associates.

The trap of U. vulgaris has been sufficiently described already.
Those of such species as U. gibba, and of the terrestrial U. subulata,
U. biloba [27 — i) and a number of others, all small plants, show
only slight differences. In U. neottioides {zy — -9), growing in running
water, the traps present a streamhne contour and a deep overhang.
In those species, such as U. reniformis, which live more or less epi-
phytically in wet moss, etc., the antennae are broad at the base, un-
branched, and appear to be adapted to holding water in the entrance
by offering support for surface films. Sometimes the entrance is
tilted forward (U. longifolia) involving the threshold {26 — 4), so that
the pavement also faces forward. There are two apparently unique
S. American species, both small and terrestrial, U. Lloydii Merl and
U. nana St. Hil. The former, U. Lloydii, has two forms of trap, one on
the leaves, the other on the stolons {28 — 1-4). They difTer in the
character of the hairs, and notably in the presence on the door of a
single tripping hair, with a saddle shaped cell next its base apparently
to facihtate hinge movement, on the leaf trap, which has also slender
backwardly curved antennae, while the stolon trap has short forwardly
directed antennae with long hairs, but no tripping bristle on the door.
Such differences are difficult to explain. In U. nana the trap is quite
devoid of appendages, but is otherwise much like that of If. Lloydii
except that the tripping mechanism consists of two bristles set trans-
versely {28 — 5, 6) (Lloyd 1932a).

U. globulariaefolia and U. amethystina represent a group of Central

Francis E. Lloyd — 258 — Carnivorous Plants

and South American species which are terrestrial. Their traps {28 —
7-9) are superficially much different from the vulgaris type, yet con-
form in having a short tube entrance, though this has a long funnel-
shaped approach, lined with numerous long-stalked glandular hairs.
The door, while lacking in well demarked mechanical areas, is ex-
tremely flexible because of very numerous constrictions in the inner
course cells (Lloyd 193 i).

Biovularia has a door in which the middle piece is half its depth
(27 — 5, 6). At the upper edge of this there are always six tripping
bristles arranged transversely and radiating outwardly (Lloyd 1935a).
In U. purpurea and its allies (27 — 2-4) the tripping hairs arise in a
radiating manner from a tubercle centrally placed on the door which
is naked of other glands. Either the entrance is quite simple and
unadorned {U. purpurea) or the lower lip may be extended into a long
upturned rostrum carrying a few unicellular hairs, with a tuft of these
on each side of, but somewhat above, the middle of the entrance
{U. elephas Luetz.). The tripping hair consists of a long stalk {zg —
7), an elongation of an epidermal cell which is part of the tubercle,
expanded at the top, bearing a short basal cell, and a large mucilage
cell with expanded cuticle. The edge of the door is thickened by a
beading which rests in a slight depression of the narrow pavement
(Lloyd 1933a, 1935a; Luetzelburg) . The outer surface hairs are
sickle-shaped mucilage cells and sessile, oil-bearing ones (29 — 6).

The mechanical response following a contact of prey against the
tripping hairs cannot of course be seen, but may be fairly guessed at.
Movement of the hairs causes slight rotation of the knob to which they
are attached. This results in slight displacement of the door middle
piece, disturbing the even contact of its edge on the threshold, thus
allowing the pressure of water to push it in. In a diagram {26 — 8)
I have shown the action (much exaggerated) as in the up and down
plane. The thinness of the door about the knob allows its rotation
(Lloyd 1933a).

Traps with a long tubular entrance. — ■ The species belonging to this
group present a by no means homogeneous picture, and in some cases
are obviously less closely related to each other than those in the
short-tube entrance group. With regard to the mechanism of the
entrance they fall into two sub-groups: (7) That in which the door
when in relaxed posture presents along the sagittal axis a single con-
tinuous curve; and (2) that in which the door shows two curves, a
strong one in the upper hinge region, and a lesser one in the middle
piece. To the former belong U. cornuta (N. America) and caerulea
(Old World), the latter representing a large number of allied species.

U. cornuta will serve as an example (jo — 1-8) . The trap is
wholly devoid of appendages. Just below the entrance there is a
rounded group of sessile glands (Schimper) which may be regarded
as a lure for prey. The tripping mechanism consists of a scattered
group of sausage-shaped glands on the lower half of the upper hinge.
They can be seen when one looks straight into the entrance. The pos-
ture of the door in the living trap in set and relaxed condition was
studied, and recorded photographically. In the set posture, the outer
selvage of the middle piece rests on the middle zone of the pavement.

Chapter XIV — 259 — The Utricularia Trap

From that point as seen along the sagittal line, the door is gently
concave throughout its whole length. The whole extent of the middle
piece is covered by the velum {24 — 6), leaving the upper region with
the tripping hairs exposed. A touch on this surface discharges the
mechanism, and the door immediately returns to the closed but now
relaxed posture in which the whole door is convex outwardly. In the
set posture, while concave along its middle axial line, it is slightly
convex transversely, that is, it is saddle-shaped. It was possible to
make transverse sections of the hving trap, and these disclosed the
door posture in the middle piece region, from which it was clearly seen
that the close application of its selvage to the pavement is procured by
the thrust of the thick lateral hinges. The release from this posi-
tion results only from the longitudinal extension of the shallow fold
already present in the set posture of the door.

ScHiMPER (1882) was the only previous student of this plant.
Since he accepted the Cohn and Darwin view, he was not aware of
any special significance to be attached to the structures of the entrance
mechanism.

U. caerulea {31 — i), U. ogtnospenna, U. equiseticaulis, U. bifida
(Asia), U. cyanea (Australia) and a lot more species, with the general
features of the trap very similar, conform to U. cornuta, except in
relatively unimportant details. They are usually provided with two
simple antennae and a small overhang, and the tripping mechanism
consists of a group of short-stalked glandular hairs, the longer nearer
the top of the door, and the shorter as the middle piece is approached.
Goebel's very brief account of the trap of caertdea shows the general
position of the door correctly though sketchily. U. bifida is evidently
of this group (Goebel) as I have myself determined, confirming
Goebel's drawing as correct. Only bifid hairs are present in the
interior of the trap, as Darwin observed. In such species, however, the
glands below and at the edge of the threshold have a single capital.

As a type of the second sub-group, we choose U. capensis, and
U. Welwitschii. A number of other species pecuhar to S. America,
Central and South Africa, all small plants, fully conform to this type.
A description of the form of the door and of its manner of operating
has already been given above (p. 255). The tripping mechanism (jz —
8, 9) consists of a curiously formed large trichome, the capital cell
of which resembles in shape a Malay kriss, called therefore the kriss
trichome (Lloyd 1931), supplemented by a group of curved glandular
hairs on the upper part of the door. In some species the kriss trichome
is not present, its place taken by large globular sessile glandular cells
{U. pellata, U. Deightonii Ms.) {31 — 6, 7). A conspicuous feature of
this group of plants is the development of a funnel shaped approach
to the entrance by spreading of the cheeks, and the lining of this
funnel with about ten rows of stout glandular hairs radiating towards
the entrance. In U. Welwitschii these are reduced to mere sessile
glands but a rostrum bears a radiating row of longer glandular hairs
{31 — 4). The S. American U. peltata, so like U. capensis except for
the globular tripping hairs, has in common with some African species
(£/. Deightonii in Ms.) minute peltate leaves very thickly covered with
stiff mucilage (29 — 9, 10), a significant fact of geographical distri-

Francis E. Lloyd — 260 — Carnivorous Plants

bution. Some Asian species {U. rosea, U. Warhurgii) (ji — lo, ii)
studied by Goebel have an extension of the funnel to form a long pro-
jecting beak of the shape of a knife blade, armed with gland hairs.
Apparently some degree of trap dimorphism occurs in U. rosea, affect-
ing the size of the trap and the form of the beak (Lloyd 1932a). Species
showing these pecuharities are found also in Australia and New Guinea.
U. Kirkii, occurring in central Africa, is apparently unique (jj —
I, 2). Of the same general form and appearance of U. capensis, the
threshold retreats into the interior, and has no step leading to the
pavement. The tripping mechanism consists of two long upwardly
curved bristles based at the juncture of the hinge and the blunt edged
middle piece. The latter is fortified by two large tubercles developed
from inner course cells, each semi-pyriform, with a thin line of tissue
between, along which the middle piece can fold during opening.

U. orhiculata (32 — 1-4). — This, representing a group of species in
Asia and Africa, was examined by Goebel, who did not observe more
than the stubby, branched, glandular antennae. The entrance mech-
anism is very peculiar. The velum is supplemented by membranes
arising from the stalked glands of the step. The tripping mechanism
of the door consists of three glandular hairs set in a triangle on the
upper half of the door. One is mallet-shaped, placed at the inwardly
directed apex of the triangle. The other two are at first large,
globular, nearly sessile glands (32 — 6). In maturity, the capital, con-
taining a large mass of stiff mucilage, bursts in a regular fashion, re-
leasing a long sausage-like mass of jelly which remains attached to the
hair. Two of these hang down in front of the entrance, and with the
mallet-shaped hair receive the impact of prey which trips the door
(Lloyd 1932a). A Thibetan species (Brit. Mus.: L.S. and T. 802),
similar to U. muUicaulis, has a broad fan-shaped rostrum armed with
radiating glands, which extend forward and in front of the entrance.
U. hrachiata is also like this (Compton) (32 — 5, 7-9).

U. longiciliata {jj — 3, 4) is a unique terrestrial species of S.
America, and has been described by Merl (191 5). The traps are
very small (0.3 mm.). The lower lip projects as a strong bifurcated
rostrum, the arms extending laterally. The upper lip forms a short,
slightly upturned beak. The middle piece of the door is exceptionally
thick and bears a single tripping hair (Merl). This consists of a thin
stalk and a disc shaped basal cell and a capital of spindle shape. It pro-
jects straight forward (Lloyd 1932a). The internal glands are few in
number but large.

A small group of species from India, East Indies and Australia in-
cludes U. lateriflora, U. simplex (Australian), U. calliphysa (Borneo)
and two probably unnamed species from India and Ceylon, all terres-
trial and small in size (jj — 5-9). The minute (0.3-0.5 mm.) traps
have a pronounced upper rostrum and a row of short glandular hairs
on each side of the trap leading up to the lower angles of the entrance.
Sometimes there is a frieze of low tubercles on each side above the
mouth (Stapf). The tripping hairs (jj — 8) stand in a prominent
group, marked by sessile, transversely long capitals on the upper part
of the door. The internal glands are few but large.

We shall now speak of that variety of traps represented in the fore-

Chapter XIV — 261 — The Utricularia Trap

going discussion by U. monanthos (p. 256) or alternatively by U.
dichotoma (they are much alike). The structure of the entrance
mechanism has been described. Beyond this is the form of the trap,
with its appendages and glands.

The appendages, when a full complement is present, consist of a
rostrum, upper (dorsal) wings and lower (ventral) wings {34 — i). The
rostrum projects forward from the overhang; the upper wings arise
one from each side of the trap above the door, and the lower wings
extend each from a point near the insertion of the stalk up to the
lower angles of the entrance. In the various species one pair or the
other of the wings may be either suppressed or greatly enlarged or
extended, with great differences in the character of their margins
{34> 35)- The rostrum is always present, but may be short or ex-
tended, to a maximum in U. tuhidata, and sometimes once branched
{U. volubilis) (34 — 6). It will suffice to refer the reader to the figures,
made with a minimum of detail, for some grasp of the great variety
within this restricted group of purely Australian plants.

In U. volubilis, which grows anchored in rather deep water, there
are three forms of traps {34 — 6). Its runner stolons bear traps in
groups of three, and these have wings crenately margined, and the
rostrum is short. The shoots bear numerous hgulate leaves and
among these, exposed directly to the water, are two kinds of trap,
large, reaching a length of five mm. (this species has the largest trap)
with rather short cylindrical stalks, so that they stand near the surface
of the substrate; and smaller ones (2 mm. or less) on very long leafy
stalks, and these stand 4-5 cm. above the substrate. The distinction
between the water traps is not a sharp one for there are gradations of
size and form, but they are on the whole quite recognizable. The traps
in the substrate are similar to those of U. monanthos in that the wings,
of which there are two pairs, are not laciniate; the rostrum is short
and unbranched. The large water traps have a long, sometimes
branched rostrum, and the edge of the shallow overhang bears addi-
tional fimbriae more or less branched, while the wings are deeply
laciniate. The abundance of fimbriae seems to clutter up the front
end of the trap. In the long stalked kinds, the ventral wings are
much reduced in size and may or may not have a single thread-like
lobe. The dorsal wings are single slender processes, as is also the
rostrum, which may be once branched. This trap resembles much that
of U. tuhulata, which is a submersed floater. Similar behavior is seen
in U. Hookeri, also an anchored submersed plant in which the traps
are long stalked {ca. i cm.). It bears traps of various sizes, the largest
4 mm. long, to smaller ones 1.5 mm. long. In the large traps the
wings are slender and fimbriate, but prominent, and the rostrum single,
rather long and straight or curved downwards. In the small traps the
ventral wings are represented by low ridges, the slender dorsal wings,
and the single rostrum, all very long (34 — 7).

In the little known floating species U. tuhulata, (j5 — 8) the traps
have a very long rostrum and filiform dorsal wings, without branches.
The ventral wings are absent. In U. Menziesii, different in habit,
being totally buried in wet quartz clay with only the leaf blades and
flowers showing, there is complete uniformity of trap structure (55 —

Francis E. Lloyd — 282 — Carnivorous Plants

lo, ii), but physiologically the traps behave differently, there being
three sets, one growing downward, one growing laterally and one up-
ward {20 — 9).

There are still other varieties of traps displaying various permu-
tations of size and shape of the appendages. So far as known these are
represented in plate 35.

Lastly the genus Polypompholyx {36 — ■ 1-9), the trap of which was
described with respect to the entrance mechanism on p. 257. It has a
very special form in this genus (there is little variety), in which the
stalk plays a special part of the approach to the door.

The form of the trap and a number of anatomical details were
described by F. X. Lang in 1901 from material in the Goebel collec-
tion, which I examined later.

The traps are of various sizes, the largest measuring 4 mm. in
length. For the most part they are smaller, about 1.5-2 mm. In one
species (possibly P. latifolia, though Bentham did not admit this
species) the traps are dimorphic both in size and structure {36 — 8, 9).
In all the species (probably four) they present the following characters.
Viewed from above, the body of the trap is seen to be roundly trian-
gular with a forked rostrum in front and a broad wing on either side.
The margins are entire but carry stiff hairs. The fork of the rostrum
is seen to clasp the stalk, over which the whole forward part of the
trap is incHned. The top of the trap body is almost fiat — this is the
upper side of the three sided body. Seen from below the trap body evi-
dently has two. lateral faces, from the upper angles of which the
wings extend. The stalk, which now hides the rostrum, gradually
swells on its approach to the trap, and is molded into two low
ridges, one on each side, just before the insertion is reached. These
ridges are strongly ciliated, forming guiding fences directing prey to
the entrance of the trap, which is approached only laterally because of
the contact of the rostrum on the stalk. The wings complete two
funnel effects, one on each side. Viewed now from the side the stalk
is seen to be increasingly massive as it approaches the trap, and this is
due to a large intercellular space which inflates the lower moiety
below the rostrum. The upper half is expanded into a ridge which
becomes deeper under the rostrum, then to be reduced. The loss of
height is, however, compensated for by a comb of stiff hairs with long,
tapering capitals, and their ends curiously distorted (Lang) as if bent
during development by impinging against the rostrum. This ridge
being tightly pressed against the rostrum divides the approach to the
door into two lateral vestibules, so that the prey must advance under
the wings from behind, to be diverted by the combs of bristles on the
sides of the stalk toward a space beneath the rostrum. This space has
the wall of the trap for its floor and the rostrum and door for its
roof, and is an antechamber leading to the entrance proper. Its floor
is clothed with mucilaginous hairs with long whip-lash capitals, lying
pointed toward the entrance. The roof, which is chiefly the door,
bears similar hairs, longer toward the door insertion, shorter toward
its free edge. The entrance is a small semi-circular hole in the trap
wall, which stands at a steep angle with the floor of the antechamber.
The semicircular edge of the entrance is clothed with pavement epi-

Chapter XIV — 263 — The Utricularia Trap

thelium, the middle zone of which lies just within this edge. The
outer zone, which carries the velum, faces outwardly {36 — 8, 9). The
inner region bears glandular hairs of various forms, at first with
conical capitals, then with bifids. Quadrifids of large size occupy the
interior wall surface. The door lies almost at right angles to the plane
of the threshold, result of the forward bending of the rostrum. The
action of the door has been already described (p. 257). Histologically
the door presents a unique feature in the very great depth of the inner
course cells in the upper hinge region, the door gradually tapering in
thickness toward the edge. Of this we may say that these thick cells
can exert a strong tangential thrust so as to press the door selvage
firmly against the pavement, the outer zone of which bears the velum,
seen in living material at Perth, W. Australia. The door selvage is not
thickened. Its cells are of equal thickness in both courses, and there is
no obvious middle piece. This means that the door selvage must bend
over the pavement, not impinge edgewise on it. The tripping mechan-
ism consists of short, bent, glandular hairs, 30-40 in number, scattered
on the surface of the door below the middle point {36 — 5).

The dimorphism in the traps of P. latifolia has been indicated.
There are two sizes of traps. In the larger, the threshold behind the
pavement bristles with a dense fringe of conical glands of graduated
sizes, described by Lang. Inside this pale stand some bifid glands.
In the small sized trap there are no conical glands. In their place
there are glands with single-celled capitals of the form of the bifids and
quadrifids. Inside the traps are bifids {36 — 8, 9).

The walls consist of four courses of cells, the two epiderms and
two courses of parenchyma. The epiderms vary in thickness. The
outer is thickest in the middle of the sides, and the inner thickest at
the angles, here forming a hinge structure.

The total thickness of the three walls, which have four courses of
cells throughout, is always greatest at the middle of their faces, pro-
ducing a hinge effect at the angles. Further, the outer epiderniis is
always thin at the angles and progressively thicker toward the middle
of the faces, while the inner is thick at the angles and th n elsewhere,
the more readily allowing compression on the inside of the angles and
on the outside of the faces. It is evident from mere inspection that
these massive walls must exert a big pull when the trap is exhausted

of water (3^ — 7)-

In closing this account one cannot but wonder at the astonishing
variety of trap structure. It is not less astonishing that there is no
evidence that one form of trap is superior to another in action. The
fact of variety is one with the same phenomenon observed when we
survey attentively some other unit of structure. It seems as though
nature, or to deify her fruitfulness, Nature, is not nor ever has been
content to make some one thing, however satisfactory, and to let it go at
that. She must show that she is not bound to the details of a pattern
that, in this case, she can make a whole shelf full of different kinds of
traps, as if to puzzle you to pick the best.

Digestion. — Goebel remarked the great difficulty, because of their
small size, of studying the traps ot Utricularia to determine the pres-
ence or absence of digestive activity. It had of course been quite

Francis E. Lloyd — 264 — Carnivorous Plants

apparent to Darwin, Cohn, Mrs. Treat and others that animals
caught in the traps disintegrate, but the natural inference, that diges-
tion was effected by the plant, was not sustained by evidence, for a
few experiments done by Darwin in which he introduced minute
fragments of meat, albumen and cartilage into the traps, gave only
negative results, and he concluded that Utricularia cannot digest its
prey. Bijsgen fared no better — he worked with an acid medium
with which Luetzelburg also got meagre results.

GoEBEL regarded Utricularia as capable of digestion because of its
close relationship to Pinguicula, but confessed that no evidence had
been forthcoming. Luetzelburg (1910), one of his students, obtained
evidence with sap expressed from large numbers of traps removed
individually, ground up with clean sand and glycerin, and perco-
lated. The extract thus obtained showed a sHght activity, visible
after 3 days, in an acid medium. It was, however, much more active
in an alkaline medium, and the conclusion that a trypsin was present
was arrived at.

During prolonged observation of the experiments it was noticed
that there was never any odor of putrefaction, and culture tests
showed that bacteria did not grow in the presence of the expressed
juices, yet these could liquify gelatin in four days. The presence of an
agent inhibiting the growth of bacteria was inferred and this inference
was strengthened by experimental evidence that bacteria are only
feebly produced in trap fluid put on a gelatin surface. This led to the
discovery of benzoic acid in the trap fluid, this substance having been
found also in the leaves and glands of Pinguicula by Loew and Aso,
and in the pitchers of Cephalotus by Goebel.

Adowa (1924) attacked the same problem. He first made sahne
and acid (HCl) extracts of the whole plant, and tested their efficacy
in digesting gelatin, fibrin, milk casein and egg-albumin. The tissues
of the whole plant contain, he found, two proteoclastic ferments,
alpha- and beta-protease, the latter active in an acid medium. The
former is rendered a httle more active with the addition of CaCl2 to
it in a neutral medium. He then made extracts of three lots of ma-
terial (fl) stems, {h) green (young) traps and (c) red and blue traps,
and tested these separately. In neutral gelatin, the effects of these
three extracts were in the ratios of 18.5 for green traps, 6.5 for colored
traps, and 3.5 for stems; in alkaline gelatin the ratios were 22, 23.5
and 6. In acid gelatin the effects were rapid at first but stopped quickly,
while in the alkahne and neutral media the action was continuous. The
conclusions were drawn that (/) the extract of the traps contained more
alpha-protease than that of the stems; (2) that of green traps affects
alkaline gelatine over a long period (24 days) to the same extent as
that of colored traps; (3) the extract of green traps acts more ener-
getically on neutral gelatine than that of the colored traps; {4) the
protease content of branches is very insignificant; (5) alkahne gela-
tin is the best medium for digestion by undiluted extracts, neutral
gelatin for diluted extracts (50% and less); {6) beta-protease both
from the branches and from the traps shows a weaker activity than
alpha-protease, and (7) extracts diluted 8-16 times act in neutral but
not in alkahne medium.

Chapter XIV — 265 — The Utricularia Trap

It seemed evident from all the foregoing that digestive ferments
are present, but principally in the traps.

KiESEL (1924), however, took the opposite view. He found that
fragments of fibrin were digested in the traps, but if acidified with
0.2% HCl were not. The trap fluid, obtained by means of a fine
pipette directly from the traps and preserved under toluol showed
no power to digest fibrin, gelatin or albumin. He concluded that the
digestion in the traps of Utricularia is the work of microorganisms.
What these might be was investigated by Stutzer (1926). Traps
washed in sterile physiological solution were minced and the contents
thus obtained were sown on agar plates. He found bacteria of the
Bacterium coli group to be dominant and suggested that they play the
important role of digestion. Other bacteria play a secondary role.
Those of the kind found in the digestive tracts of insects etc., are also
to be met with, but these he thought play the same role in the traps
as there, namely, to conserve the nutrient mass during digestion and
hinder the development of putrefactive bacteria. It is possible, in ad-
dition, that Bacillus aquatilis communis, one of the soil bacteria, takes
some part in digestion, since it can digest albumin.

And there the matter stands at the moment. On the one hand,
it is held that the presence of benzoic acid inhibits bacterial action,
and that any digestive action is the work of the ferments secreted by
the trap itself; on the other the digestion is referred to the activity of
bacteria. Hada (1930) takes a middle ground, holding that "the

animals captured are decomposed not only by the enzyme secreted

by the plant, but also by the bacteria which increase rapidly after the
death of the animals." Since his paper is in Japanese, I do not know
what evidence he puts forward.

Prey and their fate. — The presence of sugar as well as mucilage
in the glandular hairs at the entrance of U. vulgaris, shown by Luetzel-
BURG, was thought by him to indicate that these hairs form a lure to
attract animals. The presence of special groups of glands near the
entrance in some species {e.g. U. cornuta) seems to support this view.

While it is true enough that animal prey captured by the traps of
Utricularia sooner or later succumb and are digested, there are ex-
ceptions in organisms which are able to live and multiply in the re-
stricted space of the interior of the trap, notably Euglena, Heteronema,
Phacus (Hegner) and probably others, including diatoms and desmids,
often seen. There is at present no evidence of obligate relations;
these forms seem to be caught probably accidentally, and can live
inside the trap indefinitely, though Hegner states that when plants
are kept a long time in an aquarium, the Euglena runs out. Protozoa
when captured generally succumb, but some remain alive for a long
time, a fact noted long ago. The presence of decaying Paramaecia in
the trap does not affect the Euglenae.

That Paramaecium is sometimes quickly killed and at others re-
mains alive for a long time (75 min. to 17 days) (Hegner) seems to
indicate that the physiological conditions in the traps are not always
uniform. Luetzglburg thought he detected a paralysing effect of
the extract which he used for digestion experiments on small crusta-
ceans, but that it must be weak. Hada advanced the idea that

Francis E. Lloyd

— 266 —

Carnivorous Plants

animals are killed by being compressed by the walls in becoming
convex. This can hardly be the case as animals have been seen to Hve,
meanwhile freely moving, for days.

Appendix: —

Here in a position of obscurity I ask leave to present two models,
in the form of mouse traps, designed ad hoc, to illustrate the way in
which the trap of Ultricularia has been and at present is thought to
work.

Two models are offered. One, Fig. id, represents the mechanism of
the trap as conceived by Cohn, Darwin and others. In this the door
is a passive check valve, easily pushed inwards, but not outwards.
In the model a small hole in the bottom of the door allows the mouse
to see the bait thus enhancing the effect of the lure by adding sight
to smell. This model is an improvement on the Utricularia trap in
having the bait on the inside. Its extreme simplicity is in contrast to
that of the second model, Fig. ii, which affords an analog in which the

Fig. io. — A mouse trap designed to embody the idea held by Cohn and by Darwin
and others for fifteen years after them.

complexity of the Utricularia trap as now understood is suggested with-
out exaggeration.

A description of this model is presently given. A box is provided
with a door having two hinges (hi, h2). Below h2 the part d2 swings
independently from that above, di. Pressure applied at the arrow pr
cannot push in the door; but rotation of d2 on h2, so that its edge
clears the stop {sp), allowing inward swing. Outward swing is pre-
vented by a backstay st4. A handle tr on d2, actuated by a mouse,
accomplishes inward opening by pulling on the string sts, whereby the
doodad {d) is pulled away from the top of the plunger pi, allowing play
to the spring ^2. This spring then pulls on the string sh actuating the
double pulley p 'X 2, one element smaller than the other in the ratio M.
The outer pulley pulls on sh, swiftly opening the door. To this is
attached a device called a booster, B, the purpose of which, hke the
sudden inward gush of water in nature, ensures the entrance of the
mouse into the trap. This is now momentarily open, and of course
would remain so unless power were available to close it again. This is
supplied by an electric motor m which starts to rotate when an electri-
cal circuit is closed by a contact point on the plunger coming into
contact with e. The motor continues to rotate till the plunger, push-
ing the spring ^2 into its set posture, the door being pushed back into
position by its spring Si. When this is completed, the contact point

Chapter XIV

267 —

The Utricularia Trap

on the plunger comes into contact with the contact point e, below,
and the relay r then stops the motor. The power from the motor
is applied to the plunger through the gear p2 etc., ending in a cani c,
the whole being adjusted so that the cam comes into a position which
allows the lever / to swing downwards when the door is actuated
again by, it is confidently hoped, a second mouse. In the meantime,
the mouse first caught can employ his time admiring the interior
effect, and possibly suggest improvements. A digestion chamber could
of course be provided.

Fig. II. — a mouse trap intended as a model embodying present ideas of the Utric-
ularia trap as a mechanism (with apologies to Heath Robinson).

A captious reader may find difficulty in accepting the analogy as
complete. I can say only that he would be right; but at least a pur-
pose is served, to indicate that the Utricularia is a pretty complex bit
of mechanism.

Literature Cited:
Adowa, a. N., Zur Frage nach den Fermenten von Utricularia vulgaris L., I. Bioch. Z.

150:101-107, 1924; II. 153:506-509, 1924. XT ^^ TJ . /- ^ A

Barnhart, J. H., Segregation of genera m Lenhbulariaceae. Mem. N. Y. Bot. Lrard. 6:39-

Bath, \v., liber Kaulquappen in den Fangblasen von U. vulgaris. Sitzungsber. d. Ges.

Naturforsch. Freunde Berlin 1905:153-155- „ . .

Benjamin, L., Uber den Bau und die Physiologie der Utriculanen. Botan. Zeitung 6:1-5;

17-23; 45-50; 57-61; 81-86, 1848.
Brocher, Frank, Le probleme de I'Utriculaire. Ann. de Biol, lacustre 5:33-46, 19"-
Brocher, F., a propos de la capture d'anopheles par les Utriculaires. Ann. Parasitol. 5:

Brumpt, E., Capture des larves de Culicidees par les plantes du genre Utricularia. Ann. de

Parasit. humaine et compar. 3:403-411, 1925-
Buchenau, Franz, Morphologische Studien an deutschen Lentibularieen. Botan. Zeitung

23:61-66;. .69-71; 77-80; 85-91; 93-99. 1865. ^ . rr. ■ , ■ 1 • T 13

BtJSGEN, M., Uber die Art und Bedeutung des Tierfangs bei Utricularia vulgaris L. Ber.

d. deutsch. bot. Gesellsch. 6:55-63, 1888.
Candolle, a. p. DE, Physiologie vegetale 11:528, 1832.

Chandler, Bertha, Utricularia emarginata Benj. Ann. Bot. 24:549-555, 19 10.
Clarke, W. G. and R. Gurney, Notes on the genus Utricularia and its distribution in

Norfolk. Trans. Norfolk and Norwich Nat. Soc. 11:128-161, 1920-1921.
Cohn, Ferd., tJber die Funktion der Blasen von Aldrovanda und Utricularia. Cohns Bei-

trage zur Biologie der Pllanzen i(3):7i-92, 1875.
CoMi'TON, R. H., The morphology and anatomy of Utricularia brachiata Oliver. New

Phytologist 8:117-130, 1909. .

Crouan Freres, Observations sur un mode particulier de propagation des Utricularia.

Bull, de la Soc. bot. de France 5:27-29, 1858.
Curry, Dalferes P., Breeding of Anopheles mosquitoes among aquatic vegetation of

Gatun Lake, accompanied by periodic long flights of A. albimanus Wied. Southern

Med. Journ. 27:644-651, 1934. _ . ,, ^ r, j j . u

Czaja, a. Th., Ein allseitig geschlossenes, selektivpermeables System. Ber. d. deutscn.

bot. Gesellsch. 40:381-385, 1923.

Francis E. Lloyd — 268— Carnivorous Plants

CzAjA, A. Th., Die Fangvorrichtung der Utriculanah\a.se. Zeitschr. f. Bot. 14:705-729,

IQ22.

CzAjA, A. Th., Physikalisch-chemische Eigenschaften der Membran der Utriculariah\a.se.

Pfiugers Arch. f. d. Ges. Physiol. 206:554-613, 1924.
Darwin, Charles, Insectivorous Plants. New York 1875.
Dean, B., Report on the supposed fish-eating plant. Commissioners of Fisheries of the

State' of New York, Report 18:183-197, 1890.
Drude, O., Die insektenfressenden Pflanzen. Schenk's Handbuch der Botanik 1:113-146

{Utricidaria, pp. 133-135), Breslau 1881.
Ekambaram, T., Irritability of the bladders in Utrictilarm. Agric. Journ. India 11:72-79,

Ekambaram, T., Uiriadaria flexuosa Vahl. Bot. Bull, of the Presidency College Madras,
Sept., 1918:1-21.

Ekambaram, T., A note on the mechanism of the bladders of Utricularta. Journ. Indian
Bot. Soc. 4:73-74, 1926.

Fermi, C. & Buscaglione, Die proteolytischen Enz3Tne im Pfianzenreiche. Lentralbl. t.
Bakt., Parasit. u. Pfianzenkr. II, 5:24-33; 63-66; 91-9S; 125-134; 145-158, 1899.

Fernald, M. L., Expedition to Nova Scotia. Rhodora 23:89-111, 1921.

Fernald, M. L., Specific segregations and identities in some floras of eastern North Amer-
ica. Rhodora 33:25-63, i93i- . xt 77. • 7 ■ 7 • u 1 c

FRANfA, C, Recherches sur les plantes carnivores, II. Utricidana vulgaris. Bol. boo.

Brot. I, ser. 2:11-37, 1922.
Gardner, G., Travels in the interior of Brazil 1836-1841. London 1846.
Gates, F. C, Heat and the flowering of Utricularta resupinata. Ecology io(3):353-354,

1929.
Geddes, Patrick, Chapters in Modern Botany. New York 1893.
GiBBS, R. D., The trap of Utricidaria. Torreya 29:85-94, 1929.
Gislen, T., Beitrage zur Anatomic der Gattung Utricularia. Arkiv for Bot. 15:1-17,

GLtJCK, H., Biologische und morphologische Untersuchungen uber Wasser- und Sumpfge-
wachse. Jena 1906. , tt j

Goebel, K., Vergleichende Entwicklungsgeschichte der Pflanzenorgane. Schenk s Hand-
buch der Botanik 111:99-431, 1884. (re Utricularia etc., pp. 236-241.)

Goebel, K., IJber die Jugendzustande der Pflanzen. Flora 72:1-45, 1889.

Goebel, K., Der Aufbau von Utricularia. Flora 72:291-297, 1889.

Goebel, K., Morphologische und biologische Studien, V. Utricularia. Ann. Jard. Bot.
Buit. 9:41-119, 1891.

Goebel, K., Pflanzenbiologische Schilderungen, II. Marburg 1891.

GuRNEY, Robert, Utricularia in Norfolk: the effects of drought and temperature. Trans.
Norfolk and Norwich Nat. Soc. 11:260-266, 1921/22.

Hada, Y., The feeding habits of Utricularia (with English abstract). Trans. Sapporo Nat.

Hist. Soc. 11:175-183, 1930. r TT, ■ 1 ■ Tj- 1

Hegner, R. W., The interrelations of protozoa and the utricles of UtriaUaria. tJiol.

Bull. 50:239-270, 1926. .

HoEHNE, F. C, & KuHLMANN, J. G., Utricularias do Rio de Janeiro e seus arredores. Mem.

Inst. Butantan 1:1-26, 1918. , -r.- ■ . ou-

HovELACQUE, MAURICE, Rechcrches sur I'appareil vegetative des Bignomacees, Rmnan-
thacees, Orobanchees et Utriculariees. 765 pp.. Lib. Acad. Med. Pans, 1888 {Utricu-
laria, pp. 635-745). T ^ QQ

Im Thurn, E. F., Among the Indians of Guyana. London 1883. _ , 00 t-

Im Thurn, E. F. & D. Oliver, The botany of the Roraima Expedition of 1884. irans.
Linn. Soc. Lond. ser. II, 2 (bot.), 1881-1888; Part 13:249-300, 1887.

Irmisch, Thilo, Botanische Mitteilung, I. tjber Utricularia minor. Flora 41:33-37, 1858.

Jane, F. W. & Wells, B. R., Observations on the seeds and seedhngs of Utricidaria vul-
garis L. Trans. Norfolk and Norwich Naturalists Soc. 14:31-54, 1935.

Kamienski, Fr., Vergleichende Untersuchungen uber die Entwicklungsgeschichte der Utri-
cularien. Botan. Zeitung 35:761-775, 1877. ^, , o o

Kamienski, Fr., Lentihulariaceae. Naturliche Pflanzenfamilien 4, Abt. 36:108-123, i»95.

KiESEL, A., Etudes sur la nutrition de VUtricidaria vulgaris. Ann. Inst. Pasteur 38:879-

Kruck,'m., p'hysiologische und zytologische Studien uber die Utriculariah\a.se. Bot. Arch.

33:257-309, 1931-

Lang, F. X., Untersuchungen liber Morphologic, Anatomic und Samenentwicklung von
Pohpompholvx and Byblis. Flora 88:149-206, 1901.

Lloyd, F. E., The mechanism of the watertight door of the Utricularia trap. Plant Phys-
iol. 4:87-102, 1929. .... r TT, ■ J •

Lloyd, F. E., The range of structural and functional variation in the traps of Utricularia.

Flora 125:260-276, 1931. . , c TT, ■ 1 ■

Lloyd, F. E., The range of structural and functional variety in the traps of Utricularta
and Polypompholyx. Flora 126:303-328, 1932a.

Chapter XIV

269 — The Utricularia Trap

Lloyd, F. E., Is the door of Utriadaria an irritable mechanism? Canadian Journ. Res.

7:386-425, 19326. /-. J- T TJ

Lloyd, F. E., The structure and behaviour of Utricularia purpurea. Canadian Journ. Kes.

8:234-252, 1933a. _, ., • I » jj

Lloyd, F. E., Carnivorous plants — a review with contributions. Presidential Address,

Trans. Roy. Soc. of Canada, Ser. Ill, 27:35-101, 1933^- ....,•/•, j-

Lloyd F. E., The types of entrance mechanisms of the traps of Utricularia (including

Poly pom pholyx). Presidential Address, Section K-Botany, B. A. A. S., Leicester,

Sept. 1933c, pp. 183-218. /- J- T 1?

Lloyd, F. E., Additional observations on some Utricidariaceae. Canadian Journ. Kes.

10:557-562, 1934- ^. , ^ .

Lloyd, F. E., Utricularia. Biol. Reviews 10:72-110, 1935a- . ^, . , . „ .,

Lloyd, F. E., Struktur und Funktion des Eintrittsmechanismus bei UlnciUaria. Beih. z.

Bot. Centralbl. A, 54:292-320, 19356. c f .

Lloyd, F. E., The traps of Utricidaria. Proc. Sixth Intern. Botan. Congress, Sept. 1935,

Lloyd, F. e'., The trap of Utricularia capensis, how it works. Journ. S. Afr. Bot., 2:75-94,

Lloyd,^F. E., Notes on Utricularia, with special reference to Australia, mth descriptions of

four new species. Victorian Naturalist 53:91-112, 1936c. _ _

Lloyd, F. E., Further notes on Australian Utricularia with a correction. Victonan JNat-

uralist 53: 163-166, 1937a- ^ , , ^ c at u^f ...r-

Lloyd, F. E., Utricularia: its development from the seed. Journ. b. Air. Hot. 3.155-

164, 19376. . „,

Lxjetzelburg, p. v., Beitrage zur Kenntnis der Utricularia. flora 100:145-212 1910.
Matheson, Robert, The utihzation of aquatic plants as aids in mosquito control. Amer.

Nat. 64:56-86, 1930. .,

Mayr, F. X., Hydropoten an Wasser- und Sumpfpflanzen. Diss. Erlangen, 1914- Bein.

Bot. Centralbl. I, 32:278-371, 1915- . , ^ . , , u- u. j

Meierhofer, Hans, Beitrage zur Kenntnis der Anatomic und Entwickelungsgeschichte der

Utricularia-BlsLsen. Flora 90:84-113, 1902.
Merl, E. M., Beitrage zur Kenntnis der Utricularien und Genliseen. flora 108:127-200,

1915.
Merl, E. M., Biologische Studien iiber die Utricidariahl^ise. Flora 115:59-74. 1922.
Merl, E. M., Beitrage zur Kenntnis der brasilianischen Utricularien. Flora 118-119:

Merl, E. M., A new Brazilian species of the genus Utricularia. Bull. Torr. Bot. Club

Merz, M., Untersuchungen iiber die Samenentwickelung der Utricularien. Flora 84:69-87,

1897.
Meyen, F. J. F., Neues System der Pflanzenphysiologie. Berlin 1837.
Meyer, F. J., Zur Frage der Funktion der Hydropoten. Ber. d. D. Bot. Gesellsch. 53:542-

Morren, Ed., La theorie des plantes carnivores et irritables. Bull, de TAcad. Roy. Belg.

II, 60:1-60 (repaged?), 1875.
MosELY, H. N., Bull. U. S. Fish Commission 4:259, 1884/5.
NoLD, R. H., Die Funktion der Blase von Utricularia vulgaris (Ein Beitrag zur Llektro-

p'hysiologie der Drusenfunktion). Beihefte Bot. Centralb. 52:415-448, i934-
Oliver, Daniel, The Indian species of Utricularia. Journ. Linnean Soc, Bot., 3:169-176,

Oppenheimer, C, Die Fermente und ihre Wirkungen. 5. Aufl., 2:1106-1108, 1925-
PORSILD, M. P., Stray contributions to the flora of Greenland, VI-XII. Medd. om Gr0n-
land, Komm. f. Videnskab. Unders^g. i Gr^nland 93:i-94, i93S {Ulriadana, pp. 25-

34)-
Prat, S., Plasmolyse und Permeabilitat. Ber. d. D. Bot. Gesellsch. 41:225-227, 1923. _
Pringsheim, N., Uber die Bildungsvorgange am Vegetationskegel von Utricularia vulgaris.

Monatsbericht k. Akad. d. Wiss. 1869:92-116. ^

Ridley, H. N., On the foliar organs of a new species of Utricularia from St. Thomas,

West Africa. Ann. Bot. 2:305-308, 1888.
RossBACH, G. B., Aquatic Utricularias. Rhodora 41:113-128, 1939.
St. Hilaire, A. de. Voyages dans les provinces du Rio de Janeiro et du Minas Geraes.

Paris 1830. , . r • u

SCHENCK, H., Beitrage zur Kenntnis der Utricularien. Pnngsheim s Jahrb. f. wissensch.

Bot. 18:218-235, 1887.
ScHiMPER, A. F. W., Notizen uber insectenfressenden Pflanzen. Botan. Zeitung 40:225-

234; 241-248, 1882.
Schwartz, O., Plantae novae vel minus cognitae Australiae tropicae, Rep. spec. nov. reg.

veg. 24:80-109, 1927. ^^;

SIMMS, G. E. Bull. U. S. Fish Commission 4:257, 1884/5. , , ,< \

SIMMS, G. E. & H. N. MosELY, Naturforscher 17:276, 1884. Ref. in Centralbl. f. Agn-

culturchemie 14:69, 1885.

Francis E. Lloyd — 270 — Carnivorous Plants !

Skutch, a. F., The capture of prey by the bladderwort, A review of the physiology of the )

bladders. New Phytologist 27:261-297, 1928. _ i
Stapf, O., Lentibulariaceae, in Flora of Tropical Africa. London, 1906.

Staff, O., Lentibulariaceae (of Borneo). Journ. Linn. Soc, Bot., 42:115, 1914. '

Stephens, Edith L., Carnivorous plants of the Cape Peninsula. Journ. Bot. Soc. South j

'Africa, Part IX:2o-24, 1923. I

Stutzer, M. J., Zur Biologic der Utricularia vulgaris. Arch. Hydrobiol. i7 730~735> 1926. j

Thompson, G. M., On the fertilization of flowering plants (Lentibulariaceae) . Trans, and :

Proc. N. Zealand Inst. 13:278-281, 1881. _ 1

Tieghem, Ph. van, Anatomic dc I'utriculaire commune. Ann. Sci. Nat. Bot. V, 10:54-58, j

1869; C. R. Acad. Sci. Paris 67:1063-1066, 1868. I

Topp, C. A., Notes on the genus Utricularia. Victorian Naturalist 1:71-74, 1884. ;

Topp, C. A., Note on Utricularia dichotoma. Victorian Naturalist 3:74-75. 1886. ;

Treat, Mary, Plants that eat animals. N. Y. Daily Tribune, i Feb. 1875. Reprinted '

without illustrations in Gardeners' Chronicle, March 6, 1875: 303-304. ,
Treat, Mary, Is the valve of Utricularia sensitive? Harper's New Monthly Mag. 52:382-

387, 1876.
Treviranus, C. L., Noch etwas iiber die Schlauche der Utricularien. Bot. Zeit. 6:444-448,

1848. ;

Ule, E., liber Standortsanpassungen emiger Utricularien in BraziUen. Ber. deutsch. bot. '

Ges. 16:308-314, 1898.

Uphof, J. C. Th., Einiges zur Biologic der terrestrischen Utricularien. Oest. bot. Zeit. 82: j

207-212, 1933. I

Warming, E., Bidrag til Kundskaben om Lentibulariaceae. Vidensk. Medd. nat. For. |

Copenhagen, Nos. 3-7:33-58, 1874. , ^ . j , ;
Wehrle, Emil, Studien iiber Wasserstoflionenkonzentrationsverhaltmsse und Besiedelung

an Algenstandorten in der Umgebung von Freiburg im Breisgau. Zeitschr. f. Bot.

19: 209-287, 1927. I
Withycombe, C. L., Observations on the bladderwort. Knowledge 39:238-241, 1916. _

Withycombe, C. L., On the function of the bladders in Utricularia vulgaris. Journ. Lin |

nean Soc, Bot., 46:401-413, 1922-1924. Paper published in 1923. ;

Wylie, R. B. & YocoM, A. E., The endosperm of Utricularia. Univ. of_Iowa_Studies Nat. ',

Hist. 10:3-181, My., 1923.

PLAT E S

— Plate I. —

Fig. I. — Heliamphora nutans (grown at the Edinburgh Botanical Gar-
den).

Fig. 2. — H. Macdonaldae (Photograph by Dr. G. H. H. Tate taken on
Mt. Duida, Venezuela).

Fig. 3. — Sarracenia purpurea (Photograph by Charles Macnamara,
Ontario).

Fig. 4. — S. purpurea: a leaf cut lengthwise.

Fig. 5. — S. purpurea: view facing the opening.

Fig. 6. — 5. Drummondii.

Fig. 7. — 5. flava.

Fig. 8. — S. psittacina.

Fig. 9. — S. minor (Photograph by Professor J. C. Th. Uphof).

— Plate 2. -—

Fig. I. — Heliamphora nutans. Front view, spoon enlarged and longitu-
dinal view of pitcher cut near the sagittal plane, in which the sections
i""2, 3-4 and 5-6 are indicated by corresponding numbers.

Fig. 2. — Juvenile leaf (after Krafft 1896).

Fig. 3. — Assimilating leaf (after Krafft 1896).

Fig. 4. — Abnormal juvenile leaf (after Goebel 1891).

Fig. 5. — Hair from zone 2, fig. i, above. Below, detentive hair from zone
A-

Fig. 6. — V-shaped twin hairs, from outer surface of pitcher.

Fig. 7. — Nectar gland from spoon, in longitudinal section.

Fig. 8. — Nectar gland from outer surface of pitcher.

Fig. 9. — Surface view of same.

Fig. 10. — Nectar gland from inner surface of pitcher.

Fig. II. — Transverse section of spoon indicating size and distribution of
nectar glands.

Fig. 12. — Sarracenia purpurea. Upper two-thirds of pitcher, seen whole;
view of interior of an entire pitcher: the zones are indicated by num-
bers; front view of bell; petiole in longitudinal section of a pitcher in
which this was abnormally long.

Fig. 13. — Juvenile leaf, shown in the same way as the pitcher in fig. 12.

Fig. 14. — Surface and longitudinal views of the tesselated cells of zone 2.

Fig. 15. — Surface view of a nectar gland from outer surface of pitcher.

Fig. 16. — Longitudinal section of same.

— Plate J. —

Figs, i and 2. — Sarracenia purpurea.

Figs. 3-6. -^ S. psiltacina.

Figs. 7-12. — S. minor.

Figs. 13-18. — S. Jonesii.

Fig. I. — Nectar gland from surface of the nectar roll.

Fig. 2. — Hair from outer surface of pitcher.

Fig. 3. — Nectar gland from interior surface of the pitcher.

Fig. 4. — Pitcher, cut lengthwise to show interior surface; a-h, transverse

section of nectar ridge, its position indicated in the larger figure.
Fig. 5. — Juvenile leaf (after Goebel 1891).

Fig. 6. — View looking into the entrance of leaf, such as in fig. 4.
Fig. 7. — Pitcher of S. minor.
Fig. 8. — Front view of same. Nectar droplets are indicated, as also

in next figure.
jTjg g _ The lateral view of same, with section of the nectar roll and of

the wing where indicated.
Fig. 10. — Surface view of nectar gland with a few adjacent imbricated

cells.
Fig. II. — Detentive hair, from zone 4.
Fig. 12. — Nectar gland surface of nectar roll, S. minor.
Figs. 13-18. — Developmental series, pitcher of S. Jonesii.
Fig. 13. — The mouth begins to be seen.
Fig. 14. — Lips of mouth are seen. The lateral and basal views of this

stage are seen in fig. 15.
jTjc i5 _ The lower edge of the mouth begins to extend, as the apex

elongates.
Figs. 17 and 18. — The lower edge of the mouth has been extended as a
ridge, the ala ventralis.

— Plate 4. —

Fig. I. — Darlingtonia calif ornica, as seen growing 25 miles east of Cres-
cent City, Calif.

Fig. 2. — The same, in flower in early spring (Photograph by Dr. Frank
Morton Jones, taken in Plumas Co., Calif., 1920).

Fig. 3. — The same, flowers (near Florence, Oregon).

Fig. 4. — The same. View looking up into the dome of the leaf.

Fig. 5. — The same. A leaf split lengthwise.

Fig. 6. — Nepenthes Mastersiana. A pitcher split lengthwise, showing the
waxy zone above and the glandular zone below.

Fig. 7. — N. ventricosa.

Fig. 8. — N. Balfouriana (N. mixta x Mastersiana).

Fig. 9. — • N. ampidlaria.

Fig. 10. — N. bicalcarata.

Fig. II. — N. ampullar ia, looking into a pitcher.

— Plate 5. —

Figs. i-io. — Darlingtonia calif jniica.

Figs. 1-3. — Various juvenile leaves.

Fig. 4. — A series of sections through a juvenile leaf at levels indicated in
the drawings (in mm., the 7 mm. section being proximal).

Fig. 5. — Sections through the wings of a mature pitcher at various levels
(indicated in mim.); at 10 n m. the edges of the basal wings are evi-
dent, at 65 n m. the edges of the mouth.

Fig. 6. — Juvenile leaf which had failed to become ascidiate, except
slightly in the petiolar region.

Fig. 7. — Interior view of the lower half of a large mature pitcher dome;
ii, T2, veins of the tube which fuse to enter the fishtail.

Fig. 8. — Lateral view of the interior of the dome. The hairs and fenes-
trations are only partially indicated.

Fig. g. — Interior of dome looking forward (distally), showing the articu-
lation of the nectar roll with the edges of the fishtail.

Fig. 10. — Interior view of dome looking backward, showing relation of
the ala vzntralis with the nectar roll.

Fig. II. — Sarracenia psittacina. — A young leaf showing the early stage
of the infold to form the entrance tube.

(

J'

— Plate 6. —

Figs. 1-23. — Darlingtonia calijornica.

Figs. i-io. — Various stages of the development of the leaf showing espe-
cially the development of the da ventralis as part of the lateral lips of
the mouth and quite apart from the stipular wings.

Figs. 4-6 and 10 show the development of the fishtail.

Fig. 9. — Longitudinal interior view of a leaf of the dimensions of fig. 8;
nr, nectar roll; tnr, marginal roll of the fishtail; /, upward fold of the
wall of the dome.

Fig. II. — Diagrams idealizing the development of the nectar roll and its
continuation to form the marginal roll of the fishtail; a, a theoretical
condition corresponding to the condition actually recorded in plate
5-y, b, a theoretical condition further advanced than a, in which the
'fold' {see fig. 9) has not fused and the edge of the nectar roll is still
evidently confluent with the edge of the marginal roll of the fishtail.

Figs. 12 and 13. — Supposed appearance of pitchers on the conditions in-
dicated theoretically in figs, iia and b.

Fig. 14. — A mature pitcher on a small plant (the figure is larger than
natural size) showing the posture taken in relation to the soil surface,
in which the fishtail serves as a ramp leading to the entrance of the
dome.

Fig. 15. — Appearance of mature pitchers of upright posture, front and
side view.

Fig. 16. — External view of the dome of a young pitcher somewhat ad-
vanced beyond that of fig. 10. (Goebel published a similar figure).

Fig. 17. — Seedling, with juvenile leaves only and, as it happens, with
three cotyledons.

Fig. 18. — Surface view of nectar gland in a juvenile leaf.

Fig. 19. — Surface view of a nectar gland such as seen in fig. 23.

Fig. 20. — Nectar gland from outer surface of pitcher.

Fig. 21. — Nectar gland from inner surface.

Fig. 22. — Surface view of same.

Fig. 23. — Lateral sectional view of nectar gland with at least two tiers of
cells.

— Plate 7. —

Figs. 1-25. — Nepenthes.

Figs. 1-6. — Developmental series of the pitcher leaf.

Fig. I. — Very young stage (after Hooker 1859).

Fig. 2. — Somewhat later stage, the lid having appeared (after Hooker).

Fig. 3. — Longitudinal section of a still later stage showing lid appressed
to the rim (after Hooker).

Fig. 4. — The origin of the lip as a two lobed structure (after Bower).

Fig. 5. — Definitive form of the first seedling leaf (after Stern). The
transverse stipular juncture is represented by emergencies.

Fig. 6. — A similar stage of development (after Macfarlane), in which
the transverse stipular membrane is evident.

Fig. 7. — • Front and side view of a pitcher produced on a forced sprout
showing venation as broken lines. The small number and the dis-
tribution of digestive glands are indicated.

Figs. 8 and 9. — Lateral and front view of a pitcher on a forced shoot,
showing the spur is a flat structure, and the distribution of veins.

Figs. 10-13. — Other pitchers produced on short shoots in which various
relations of the petiolar ' blade ' to the wings of the pitcher are noted.

Fig. 14. — The digestive gland in N. Pervillei, covered by a deep pocket.

Figs. 15-20. — Transverse section of the rim of several species. The arm
of the inner ridge bears a nectar gland.

Fig. 15. — N. sp. aff. Balfouriana.

Fig. 16. — N. venlricosa.

Fig. 17. — N. gracilis.

Fig. 18. — iV. Lowii.

Fig. 19. — N. ampuUaria.

Fig. 20. — -N. inermis Danser, "peristome almost none". A, transverse
section through peristome; B, looking into a nectar gland pit.

Fig. 21. — Front and side view of the nectar gland in its pit, N. ampul-
lar ia.

Fig. 22. — A^. intermedia, showing the two strongly developed groups of
peristome teeth at the base of the lid.

Fig. 23. — N. bicalcarata, divided lengthwise, showing therefore only one
of the well developed, claw-like projections at the base of the lid.

Fig. 24. — Transverse section through the lid and mouth of a very young
pitcher of N. anipidlaria. The glandular inner limb of the peristome is
seen to be an emergence from the inner surface of the pitcher wall.

Fig. 25. — N. Tiveyi. Under surface of the lid, bearing a prominent me-
dian ridge with prominent processes.

— Plate 8. —

Figs. 1-20. — Nepenthes. 1

Figs. 1-3. — Stellate hairs, sometimes emergent, sometimes in pits.

Fig. 4. — Tufted hair, producing the rusty pubescence of Nepenthes.
There are several varieties of form.

Figs. 5 and 6. — Front and lateral views of the peculiar stomata on the '

waxy zone of the interior of the pitcher. '

Fig. 7. — Front and lateral views of the epidermis clothing the peristome; ,

a-b and c-d correspond in position. N. ampuUaria {of. figs. 17 and
18).

Fig. 8. — Nectar gland from inner surface of the lid.

Figs. 9 and 10. — Digestive glands from two species, in section; cut.
suberized course of cells.

Fig. II. — Front view of a digestive gland, standing in its pocket. ,

Fig. 12. — Epidermal cells of the peristome of N. Lowii.

Fig. 13. — Nectar gland of the peristome edge, N. ampuUaria. The
outermost course of cells is suberized.

Fig. 14. — Digestive gland in young, thin-walled condition of N. ampul-
lar ia.

Fig. 15. — External alluring (nectar) gland from the midrib of the blade.
S, suberized layer.

Fig. 16. — Nectar gland from the ridge beneath the lid, of N. Tiveyi. S,
suberized course of cells.

Fig. 17. — Transverse section of epidermis of the peristome of A'. am-
puUaria. The cells appear ridged.

Fig. 18. ^ The ridges are seen to be due to the overlying of next cells
(see also fig. 7).

Fig. 19. — Quite young stage of development of a pitcher and lid in sec-
tion, showing folding of the lid and origin of the inner peristome ridge
as an outgrowth of the wall.

Fig. 20. — Transverse section of the petiole showing it to be bifacial.

— Plate g. —

Figs. i-6. — Cephalotus follicular is.

Fig. I. — ^ A clump of plants photographed at Albany, Western Australia.

Fig. 2. — A plant, grown at the Edinburgh Botanical Garden.

Fig. 3. — Pitcher split lengthwise.

Fig. 4. — A shadow picture of a glandular patch.

Fig. 5. — An entire plant in flower. Inset: flower enlarged. From a
large clump of plants sent bj- me from Austraha to the Botanical Insti-
tute, Munich (Photograph by Dr. Th. P. Haas).

Fig. 6. — View looking down on the same culture, to show the two kinds
of leaves, pitcher and photosynthetic, the latter in full development;
an aberrant one may be seen at the left of the picture (Photograph
by Dr. Haas).

Fig. 7. — Genlisea re pens (from Sao Paulo, Brazil).

Fig. 8. — The same. Tubular portion of the trap showing the lobster-
pot structure.

Fig. 9. — Byhlis gigantea (Western Australia). Various aspects of the
leaf to show the distribution of glands. The sessile glands are dis-
cernible as white dots on the leaf surface.

'f ■■:■-■ -u*^

i-mm

— Plate 10. —

Figs. 1-24. — Cephalotus Jollicularis.

Figs, i and 2. — Lateral and front views of a mature, full sized pitcher,

4.5 cm. long.
Fig. 3. — View of interior of pitcher. Distribution of veins shown: Vi, Vi

and Vi, one of each of the median and following lateral veins of the ven-
tral group, v; di, d-i, veins of the dorsal group {d), the former a single

median vein.
Fig. 4. — Sections through the base (lowermost), middle and top of a

pitcher stalk, showing distribution of veins: v, ventral; i, 2 and 3,

the three pairs of ventral veins.
Fig. 5. — Lateral view of a young pitcher 3 mm. long. For section

through a-b, see fig. 7.
Fig. 6. — Interior view of a similar pitcher, showing the distribution of

veins at this age. These can be identified with those seen in fig. 3.
Fig. 7. — Section through young pitcher through a-b of fig. 5.
Fig. 8. — Section through base (lowermost), middle and through base of

blade of a mature foliage leaf; v, ventral.
Fig. 9. — Section through rim of a juvenile pitcher (figs. 10-12) showing

tooth and distal end of the median dorsal vein.
Figs. 10, 11. — Frontal and lateral views of a juvenile pitcher 8 mm. long.
Fig. 12. — Interior view of same, showing deep collar and distribution of

veins. Compare with fig. 3. See also fig. 24.
Figs. 13-18. — Various intergradient forms of leaves between the foliage

leaf and pitcher. For normal foliage leaf, see figs. 2 and 6, plate 9.
Fig. 19. — Diagram of section of a medium sized digestive gland from

glandular patch.
Fig. 20. — Diagram of section of a small digestive gland (above) and sur-
face view of same (below).
Fig. 21. — Diagram of section of a gland from external surface of pitcher

and elsewhere.
Fig. 22. — Surface view of same.

Fig. 23. — Two stomata from surface of a glandular patch.
Fig. 24. — Longitudinal section of a trap 1.25 mm. long; v, ventral; d,

dorsal.

— Plate II. —

Figs. i-ii. — Genlisea.

Figs. 1-4, 7, 8 and 10. — Series showing the development of the trap.

Fig. 5. — A single plant (from Brazil) X 2; inf., inflorescence stalk; /.,
trap.

Fig. 6. — Portion of an arm laid open of G. ornata. In this species the
number of rows of detentive hairs is small. The position of the
glands and the shape of the inner epidermal cells (on a larger scale)
are indicated.

Fig. 7. ^ — ^ A trap in which the arms are beginning to develop; vent., ventral
view; lat., lateral view.

Fig. 8. — Distal end of a young trap showing the arms in a stage of
growth later than that seen in fig. 4. Dotted hnes indicate veins, as
in fig. 7.

Fig. 9. — Mature trap, one arm having been laid open and shown in its
posture thereafter. A portion of the laid-open arm at a is shown
above and to the right; below this, the sections along a-b and c-d
are shown.

Fig. 10. — A mature trap. The numbered lines indicate the positions of
the sections shown in plate 12.

Fig. II. — A portion of the oral termination of the tube, with the ad-
jacent portion of an arm, diagrammed in perspective. Note prey
caught, seen through the window cut in the wall, on the right.

— Plate 12. —

Figs. i-i8. — Genlisea.

Fig. 1. — Transverse section through the mouth of a twisted arm, being I

an amplified figure of part of fig. 4 {cf. 11 — 10, at j). i

Fig. 2. — Section through the mouth at the top of the neck {11 — 10 at 5)

showing the rows of detentive hairs.
Fig. 3. — Section through the mouth at the top of the neck (// — 10 at

5) showing particularly the giant articulating hair ("distance piece").
Fig. 4. — Section across arm (// — 10 at j). The obliquity of the rows

of detentive hairs is to be noted. 1

Fig. 5. — Transverse section through neck {11 — 10 at i), at the level j

of the row of glands, indicated by is, fig. 9, which is a longitudinal ^

section in the same region. !

Fig. 6. — Transverse section through neck at 2 (11 — 10), through the |

basal cells of the detentive hairs. '

Fig. 7. — Transverse section through a ridge in the belly of the flask

(7/ — 10 at 4).
Fig. 8. — Longitudinal section with perspective at 2 (// — 10). Inter-
cellular spaces are hatched. I
Fig. 9. — Longitudinal section at / (// — 10) and approximately in the

plane in which one of the vascular strands lies. ^ 1

Fig. 10. — Longitudinal dorsi-ventral section showing also the interior of

the belly of the flask.
Fig. II. — Mucilage gland from external surface lies.
Fig. 12. — Short detentive hair from edge of arm.
Fig. 13. — Detentive hair from interior of arm or neck.
Fig. 14. — Glandular hairs (presumably digestive) from belly of flask.
Fig. 15. — Articulating hair ("distance piece") from edge of arm.
Fig. 16. — Glandular hair from interior of upper part of neck and of the

arms.
Fig. 17. — Digestive hairs with four celled capital. Otherwise as in fig.

14-
Fig. 18. — View of the depression into which the large giant cell of a dis-
tance piece fits.

I

i

— Plate I J. —

Fig. I. — Byblis gigantea (W. Australia). Inset: commensal (true) bug,

as yet undescribed.
Fig. 2. — Drosophyllum lusitanicum, in culture (Munich). Right, a piece

of a leaf with captured prey.
Fig. 3. — The same (Photograph by Dr. A. Quint.^nilha).
Fig. 4. — Pinguicida vulgaris (Alberta, Canada). X '2.
Fig. 5. — At left of numeral, a piece of a leaf of Byblis; at right of nu- j

meral a part of a leaf of Drosera capensis. !

Fig. 6. — Pinguicida vulgaris. Two views of the same plant taken 24

hours apart to show leaf movements. Collected 20 mi. east of j

Crescent City, Calif. J

Fig. 7. — Drosera capensis. X V^. \

Fig. 8. — Time-lapse motion pictures of D. capensis, showing leaf move- ,1

ments. Total period about one and one half hours.
Fig. 9. — D. rotundifolia, a leaf with captured prey.
Fig. 10. — Pinguicula vulgaris, a small area of leaf surface with the muci- 1

lage glands in the focal plane of the camera lens.
Fig II. — The same, with the focal plane of the lens at the level of the

digestive glands. A captured insect is seen.

— Plate 14. —

Figs. 1-6. — Drosophyllum lusitanicum.

Fig. I. — A bit of a leaf showing the distribution of stalked and sessile

glands on the under face.
Fig. 2. — A stalked gland in longitudinal section.
Fig. 3. — Surface view of epidermal cells of the capital of a stalked gland,

to show the buttresses.
Fig. 4. — A bud showing three leaves in outward circination.
Fig. 5. — Transverse section of leaf showing the three vascular bundles

and distribution of the glands.
Fig. 6. — Sessile gland in section.
Figs. 7-15. — Byblis.

Fig. 7. — Bud of Byblis linifolia showing leaves with outward circination.
Fig. 8. — Bud of Byblis gigantea showing absence of circination.
Fig. 9. — Section across a longitudinal depression of the leaf showing

glandular hair, intermediate in form between a mucilage hair (fig. 12)

and a true digestive hair (fig. 10).
Fig. 10. — Digestive hair, one of a row as seen in fig. 13.
Fig. II. — Transverse section through a leaf, showing distribution of ad-
hesive glands, vascular bundles and the outer extensive palisade

zone.
Fig. 12. — An adhesive gland, the right hand figure showing the pores.
Fig. 13. — Surface view of leaf showing row of digestive glands.
Fig. 14. — Adhesive gland showing the curled state of the stalk when dry.
Fig. 15. — Other adhesive glands from flower stalk.

— Plate ij. —

Figs. 1-5 — Pinguicula.

Fig. I. — Series showing a leaf on different days after capture of prey,
Oct. 2, 3, 4, 6, and 7., 1939. The maximum amount of inrolling of the
margin observed is indicated. P. vulgaris (of California).

Fig. 2. — Digestive glands from the upper leaf surface.

Fig. 3. — Adhesive gland.

Fig. 4. — Glands from scape.

Fig. 5. — Section through margin of leaf, showing the large terminal
tracheid.

Figs. 6-24. — Drosera.

Fig. 6. — Longitudinal section through the glandular capital of a tentacle.

Fig. 7. — Transverse section through a tentacle stalk at the narrow region.

Fig. 8. — Transverse section through the gland of tentacle.

Fig. 9. — Transverse section through the gland of a marginal tentacle of
D. rotiindifolia {see fig. ii).

Fig. 10. — Apex of a tentacle gland showing exfoliated scale-like par-
ticles, seen lying in the mucilage secreted.

Fig. II. — Longitudinal section through the gland of a marginal tentacle
of D. rotiindifolia.

Fig. 12. — Gland of tentacle after thorough treatment with sulfuric acid.
Only cuticular membranes left.

Fig. 13. — Surface view of the outer epidermal cells of a tentacle gland
to show the buttresses.

Fig. 14. — Living tentacle gland. The mucilage was entirely removed,
the gland was then placed in mineral oil. Droplets of mucilage ap-
peared.

Fig. 15 ,— Section of leaf: absence of palisade.

Fig. 16. — Glandular trichome from petiole {D. gigantea).

Fig. 17. — Apical view of gland shown in fig. 22.

Fig. 18. — Small gland from tentacle stalk.

Fig. 19. — Longitudinal section of the apex of a leaf rhizoid of D. ery-
throrhiza.

Fig. 20. — A young leaf of D. pygmaea, showing the immense stipules.

Figs. 21 and 22. — Sessile glands of D. Whittakeri.

Fig. 23. — Sessile gland of D. capensis.

Fig. 24. — Scarious trichomes from petiole of D. rotuniifolia.

— Plate i6.—

Figs. i-i8. — Drosera.

Figs, i, 2, 3. — D. pygmaea. i, ventral, 2, lateral views of a marginal
tentacle gland; 3, sectional view of leaf blade gland.

Figs. 4-6. — Stages in the development of the gland in D. rotundifolia
(after Homes).

Fig. 7, earlier, and Fig. 8, later stage in the development of a leaf bud at
the base of a tentacle in D. rotundifolia (after Behre).

Figs. 9, 10. — Still later stages of same in D. peltata. In fig. 10 the ten-
tacle is completely displaced (after Vickery). The broken line in-
dicates one face of the thick section.

Fig. II. — End of a dropper while extending.

Fig. 12. — Same, on beginning to round up to form a bulb in D. peltata
(after Vickery).

Fig. 13. — Young and older seedling. A dropper is developing on the
older seedling in D. peltata (after Vickery).

Fig. 14. — Ventral, lateral and basal views of a gemma of D. pygmaea.

Fig. 15. — Same after Goebel. The right hand one is germinating.

Fig. 16. — D. pygmaea (or a related sp.), showing mature gemmae (after
Goebel).

Pic 17. — Diagram of mid-sectional view of the true D. pygmaea (col-
lected by Dr. Pat Brough near Sydney, N. S. W.).

Fig. 18. — Young leaf from same, dorsal view.

— Plate 17. —

Fig. I. — Aggregation in the stalk cells of Drosera rotitndifolia. These
photographs were provided by Dr. A. Akerman. A and B, successive
photographs of the same field taken 2 minutes apart; B and C, taken
15 minutes apart.

Fig. 2. — Dionaea muscipula (Photograph by Dr. Cornelia M. Smith).

Fig. 3. — Dionaea leaf which has captured a Harvest-man or Harvest-
spider {Phalangiiim sp).

Fig. 4. — Drosera gigantea. This species grows to a height of four feet
(Western Australia).

Fig. 5. — Aldrovanda vesiculosa (Silesia). On the right of the numeral 5,
half a whorl of eight leaves, axial view.

Fig. 6. — Dionaea. Moving pictures taken at one-sixteenth of a second
intervals. The last one was taken half an hour later.

E

J

— Plate iS.—

Figs. i-i8. — Dionaea muscipida.

Fig. I. — Inner surface of a trap lobe, showing the distribution of veins,
of the digestive glands and, along the scalloped margin, the alluring
glands. The three trigger hairs are seen.

Fig. 2. — A trigger hair from a fully developed large leaf.

Fig. 3. — A trigger hair from a minute seedling leaf.

Fig. 4. — a, diagram to indicate postures of the trap lobes in (/) open,
(2) closed and (j) "narrowed" states (after Ashida); b, trap nar-
rowed to the extreme.

Fig. 5. — Section of the upper epidermis and adjacent parenchyma of a
mature large trap.

Fig. 6. — Lower epidermis of same.

Fig. 7. — Young leaf: growth movement involved in the unfolding of the
trap.

Fig. 8. — A trap about 2 mm. long showing the twisted stalk and the
symmetry of the cilia.

Fig. 9. — Upper surface of the left lobe of a trap 2 mm. long, taken from
a seedling.

Fig. 10. — Sections of a mature leaf, drawn to scale, to show the relative
sizes of the epidermal and parenchyma cells.

Fig. II. — A stellate hair from exterior surface of trap (and elsewhere).

Fig. 12. — A digestive gland. The section on the left is cut trans-
versely to the leaf lobe. At right, apical view of gland.

Fig. 13. — Base of digestive gland, the section cut at right angles to
that of FIG. 12.

Fig. 14. — xAlluring gland in section.

Fig. 15. — Longitudinal section through base of trigger hair to show the
sensitive cells (arrow point).

Fig. 16. — Transverse section through the sensitive cells of trigger hair.

Fig. 17. — Transverse section of lobe of a minute (2 mm. long) trap.

Fig. 18. — Longitudinal section of same.

— Plate 19. —

Figs. 1-23. — Aldrovanda vesiculosa.
Figs, i, 2 and 3. — Stages of germination.
Fig. 4. — Apex of root cap showing its obsolescence.

Pjc 5. — a series of juvenile leaves such as at first formed on the plu-
mule; the perfect, mature form on the right.
Fjg. 6. — The development of the leaf showing the early occurrence of
twisting which brings the trap into its definitive posture. E, F and H,
dorsal views; G, ventral view.
jTjg 7 _ Immature trap, as in riG. 5, showing the curvatures which, be-
coming fixed, determine the behavior of the trap during narrowing.
Fig. 8. — Interior (upper) face of a trap lobe. Distribution of cruciform
and digestive glands shown, as also the sensitive hairs and reflexed
lobe margin, with its teeth.
Fig. 9. — Transverse section through a footstalk (petiole), showing vast

intercellular spaces.
Fig. 10. — Section through the midrib of a trap.
Fig. II. — Transverse section through the wall of a trap in the region

marked 2, fig. 23.
Fig. 12. — Three sections through the reflexed margin of the trap. Note

the irregularity of structure.
Fig. 13. — The reflexed margin of trap with tooth.
Pjq j4 — Digestive gland viewed from above and laterally; a and b are

are normal to each other.
Fig. 15. — A cruciform gland.

Pjg. 16. — A mucilage gland from the outer surface of the trap and else-
where.
Pic 17. — Surface view of the reflexed lobe margin, showing the cells of
one surface intruding between those of the other. The arrowpoints
indicate inward or outward intrusion.
Fig. 18. — Crossing of marginal spines to close the trap against the escape
of prey. The posture of the teeth in this figure is exaggerated in rela-
tion to the borders, but represents what does occur during early
approach of the edges.
Fig. 19. — Transverse section of the wall of the trap in the region be-
tween 3 and 4, fig. 23 of this plate. The component cells are cut
longitudinally.
Fig. 20. — A trigger hair (above and to the right of the numeral).
Fig. 21.- Diagrams showing (a) open, (b) closed, (c) early narrowing
and {d) completely narrowed stages of action. After Ashida, checked
by my own observations.
Fig. 22. — Diagram to show zone of major movement during closing and

narrowing. ^ , • 1 j

Pjc. 23. — Diagram of transverse section of a trap to show the thick and
thin regions, the distribution of digestive and cruciform glands, and
of the trigger hairs.

— Plate 20. —

Fig. I. — Utricular ia vulgaris. Leaf with traps.

Fig. 2. — Utricularia aff. emarginata, showing water stolons, and those en-
tering the muddy substrate.

Fig. 3. — U. purpurea.

Fig. 4. — • U. intermedia.

Fig. 5. — U. lateriflora (x J^).

Fig. 6. — U. capensis. The leaves (x 2).

Fig. 7. — U. volubilis.

Fig. 8. — U. Menziesii.

Fig. 9. — The same, a young plant, showing better the distribution in
space of the traps. One tuber is seen.

Fig. 10. — Polypompholyx tenella, an entire plant, the scape in bud. En-
larged about 6 times. Two forms of traps.

Fig. II. — Photograph (x 8) of a tadpole (Biifo marinus), 4 days old,
captured by the head by a trap of U. flexuosa at Gordonvale, Queens-
land (transmitted by Mr. J. H. Buzacott, Feb. 10, 1936)-

Fig. 12. — Floats at the base of a young inflorescence of U. stellaris (Sin-
gapore).

I

— Plate 21. —

Sh, chief stolon; 5/v, secondary stolon; 5/3, stolon of the third order; /, leaf; rhi, rhizoid;
as, air shoot; ic, scape; tr, trap.

Figs. 1-6. — Jjlricidaria vi.lgan's.

Fig. I. — Early germination of seeds.

Fig. 2. — Emergence of primary leaves.

Fig. 3. — Aberrant seedling.

Figs. 4 and 5. — Seedlings in advanced stages.

Fig. 6. — Longitudinal section of seed (after van Tieghem).

Fig. 7. — Germiination of U. emarginata. A, embryo, above, its apex en-
larged; D, embryo with one cotyledonoid a shoot; B, early germina-
tion; C, one primary shoot; E, two primary shoots.

Fig. 8. — Growing point, U. vulgaris.

Figs. 9 and 10. — Branching, U. vulgaris. Note dwarf branch in leaf axil.

Fig. II. — U. oligosperma.

Fig. 12. — U. gilha.

Fig. 13. ■ — U. intermedia.

Fig. 14. — The same, with scape.

Fig. 15. — U . reniformis.

Figs. 16 and 17. — U . n.inor.

Figs. 18 and 19. — U. exoleta aS.

Fig. 20. — U. gibba.

Fig. 21. — U. bitoba, a leaf-branch.

Fig. 22. — U. paradoxa (Ms. name); s, level of mud, with water above.

Fig. 23. — The same. Stiff hair clothing the plant, bifid and quadrifid
hairs and a trap (material: Young 1421, in Brit. Mus.).

— Plate 22. —

1, /2, primary and secondary leaves; st, with sub-numerals, stolons; Ir, trap; pod, podium;
e, embryo.

Figs, i-ii and 28. — U. monanthos.

Figs. 1-3. — Early germination, during which the primary leaf develops
precociously.

Fig. 4. — A primary trap develops in the place of a primary stolon.

Figs. 5 and 6. — Normal germination.

Fig. 7. — The upper moiety of the embryo has developed into a very long
podium.

Figs. 8 and 9. — Two aspects of the same embryo. Partial concrescence
of the primary trap with its axillary bud, and of the whole with the
primary leaf.

Fig. 10. — Seedling with a well-developed podium.

Fig. II. — The bud which produces the scape is far advanced.

Fig. 12. — U. orbiculata.

Figs. 13 and 14. — U . capensis. Fig. 13, primary leaf bearing adventive
traps each with an axillary bud in adaxial position.

Figs. 15 and 16. — U. rosea.

Figs. 17-22. — U . bifida. Fig. 18, precocious development of primary leaf;
later emergence of primary stolon from podium (3d, 4th, 5th and 6th
drawings of the series to left of numeral 18); fig. 19, lateral origin,
near base of embryo, of primary leaf; fig. 20, concrescence of pri-
mary leaf and stolon; fig. 21, the primary vegetation point has pro-
duced only a leaf, while an adventitious growing point has emerged
laterally, bearing otherwise normal leaf and stolon; fig. 22, primary
stolon apparently from primary leaf.

Figs. 23-26. — Polypompholyx tenella.

Fig. 27. — U. Barnesii (Ms. name).

Fig. 28. — U . nionanthos. A stage of development following that shown
in FIG. 3.

— Plate 2 J. —

Lettering as in plate 21.
Fig. I . — Utricidaria clandestina.
Figs. 2-6. — U. oligosperma.
Fig. 7. — U. simplex.
Figs. 8-11. — U. resupinata.

Figs. 12 and 13. — U. Stephensae (Ms. name) aff. U. cymbantha Oliver.
Figs. 14-18. — Utricularia {Biovidaria) olivacea.
Figs. 19 and 20. — U. monanthos.
Figs. 21 and 22. — - U. cymbantha.
Figs. 23 and 24. — f/. Menziesii.
Fig. 25. — f/. volubilis.

— Plate 24. —

Fig. I. — Utricularia gibba. View looking into the entrance showing the I

velum. X 50. ^

Fig. 2. — U. caeridea. Trap before {left) and after actuation. Note the /

changed door profile. |

Fig. 3. — U. monanthos, trap before {left) and after actuation. 1

Fig. 4. — View looking into the entrance of U. dichotoma, showing the \

circular outer velum. Ca. 40 x.

Fig. 5. — U. capensis. Trap before {left) and after actuation. \

Fig. 6. — Transverse section of the threshold of U. cornuta, showing the 'i

velum. '
Fig. 7. — U. Welwitschii. The tripping hair.
Fig. 8. — Polypompholyx miiUifida. Trap entrance before {above) and

after actuation. j

Fig. 9. — Inner surface of the door of U. gibba, showing the concentric 1

sulci. Note that in the upper part of this illustration the anticlinal i|

walls appear constricted, this being artifact. t

)
I

I

J
I

I

i

— Plate 25. —

Fig. I. — Utricularia Deightonii (Ms. name) aff. peltata. Sagittal section )

of trap. ■• j

Fig. 2. — Sagittal section of entrance to trap of U. peltata. |

Fig. 3. — The same, of trap of U. gibha showing very approximately the j

normal (but not set) posture of the door and velum. I

Fig. 4. — Profile of the door and velum in the set posture, in an entirely
whole trap of U. emarginata. The lens has, of necessity, to penetrate
a considerable thickness of tissues, and hence a sharp picture is unob-
tainable. The velum is seen as a bulbous mass just above the thresh-
old and in front of the lower door edge.

Fig. 5. — Sagittal section of door and velum, U. gibha — the same as in j

fig. 3, but at higher magnification to show details of structure. <

Fig. 6. — View as from the inside of the trap of the middle reach of the ;

velum in U. vulgaris.

Fig. 7. — U. gibba. Transverse section of the threshold with the velum, j

from which its origin can be discerned. .

Fig. 8. — The same. View of the velum as one looks into the entrance.

Fig. 9. — U. intermedia (or U. vulgaris). View looking down on the pave- ,

ment, showing, however, only a narrow middle fore and aft strip.
The velum is at the lower edge of the picture. The outer, middle and
inner zones are discernible.

— Plate 26. —

Fig. I. — Diagram of a sagittal section of entrance of the trap of Utricu-
lar ia gibba, with terminology.
Fig. 2. — Entrance of U. gibba, exoleta etc. in sagittal section. The
arrows (3, 4) indicate the general direction of the movements of prey
in approach. The structure of the pavement is seen through the door
as if transparent.

Fig. 3. — U. gibba. View of threshold and door edge, as seen from the
viewpoint indicated by the arrow i in fig. 2. Arrow i indicates the
direction of thrust of the side of the door on the lateral reaches of
the threshold; arrow 2, thrust of the door lateral hinges on the middle
piece, and of this against the pavement in its middle reach.

Fig. 4. — Sagittal section of entrance of U. longifolia.

Fig. 5. — U. purpurea. The knob at the middle of the door. See fig.
2-4, PLATE 27.

Fig. 6. — U. vulgaris.

Fig. 7. — The same, the door swung widely open.

Fig. 8. — Diagram of the sagittal section of the door of U. purpurea to
show that movement of the knob up or down lifts the door edge; c,
center of rotation; d, b, displacements of point a when the trigger
hairs are touched downwardly or upwardly, respectively.

Fig. 9. — Quadrifid hair of U. gibba, quite young.

Fig. 10. — Reflexed quadrifid hair of U. neottioides.

Fig. II. — Quadrifid hair of Poly pom pholyx.

Fig. 12. — The same, side view.

Fig. 13. — The same, diagram of the shape of a single one of the four cells
of the capital.

H

— Plate 2y. —

Fig. I. — Trap of Utricitlaria biloba. Only a few of the bifid and quadri-

fid hairs are shown.
Fig. 2. — Entrance of U . purpurea, sagittal section. The door is regarded

as transparent.
Fig. 3. — The same, the door in open and closed postures.
Fig. 4. — Trap of U. elephas Luetz.
Fig. 5. — Trap of Utricularia (Biovularia) olivacea.
Fig. 6. — Front view of door of same to show the six tripping bristles. [

Fig. 7. — Traps of U. resupinata displaying dimorphism.
Fig. 8. — Trap of U. exoleta (Queenlad).
Fig. 9. — Trap of U. neoUioides.

J
1

Plate 28. —

Figs. 1-4. — Utricularia Lloydii Merl, an example of dimorphism.

Fig. I. — The kind of trap found only on the leaves.

Fig. 2. — Trap found only on the stolons.

Fig. 3. — Door with single tripping bristle, from leaf trap.

Fig. 4. — Door of stolon trap, devoid of tripping bristle.

Fig. 5. — U. nana.

Fig. 6. — Entrance of same, showing the two tripping bristles.

Figs. 7-9. — U. globulariaefolia.

Fig. 7. — Entrance.

Figs. 8 and 9. — Two forms of trap (dimorphism).

— Plate 2g. —

Figs. 1-13. — Utricidaria.

Fig. I. — Diagrammatic representation of the histology of the door.
Above, view en face of the door seen from the inside of the trap;
below, section through a-b.
Fig. 2. — Sagittal section of the door: 7, upper hinge; 2, middle area;

5, central hinge; 4, middle piece; cut, cuticle; eel, cellulose.
Fig. 3. — Cells of the inner course of a portion of the door, the central
hinge approximately in the middle; the bases of the tripping bristles
shown in dotted lines. Numerous props.
Fig. 4. — Transverse section of the threshold at the middle point, showing

the velum. The- posture of the door edge in broken lines.
Fig. 5. — Section through the lateral reach, near its outer end, of the
threshold in U. purpurea. The posture of the door thrusting against
the pavement indicated by the oblique arrows. The direction of
water pressure against the door and velum shown by the arrows
below.
Fig. 6. — Surface hairs of U. purpurea; a, sickle-shaped, glandular (muci-
lage) hairs; 6, c, young stage of oil-bearing hairs; d-g, mature stage
of same, showing oil deposit held by the raised cuticle.
Fig. 7. — Glandular hairs of the central boss of the door of U. purpurea,

shown in developmental stages, numbered serially.
Fig. 8. — Diagram of wall structure, showing the inner quadrifid {left)
and the outer spherical glands. Vascular tissue between the two
cell courses.
Fig. 9. — Peltate leaves of U. sp. aff. peltata (from Angola). They bear a

deep coating of mucilage on their upper surfaces.
Fig. 10. — Mucilage glands of the same with their heavy loads of stiff mu-
cilage {b)\ c, scheme of branching: a, leaf, with three branches
(stolons of second order) emerging from the primary stolon seen
axially.
Fig. II. — U. gibba, schematized to show relation of door and threshold
in the wide-angle type of entrance (cf. 30 — 3). The angular rela-
tion of the door posture {p-d) to the general level of the threshold {p-
t) is ca. 90 ; r, relaxed posture of door; s, set posture; pr, various
possible directions of impact of prey on the tripping bristles; uh,
upper hinge; ch, central hinge; Ih, lateral hinge; <j, middle piece.
Fig. 12. — Narrow angle door: its shape and areas {cj. jo — 4) and,
below, the areas of the threshold, of U. capensis; uh, upper, and Ih,
lateral hinge; ma, middle area; mp. middle piece; oz, outer zone of
pavement; tnz, middle zone; and iz, inner zone, of same.
Fig. 13. — Wide angle door {U. gibba or vulgaris etc.); ch, central hinge.

I

I

— Plate JO. —

Figs. i-8. — Utricularia cornuta.

Fig. 1. — The trap in lateral and frontal views.

Fig. 2. — Entrance, with alluring glands below, from in front.

Fig. 3. — Sagittal section of entrance. The door {di) is shown in the set
posture; d-i its open posture; dz normal relaxed posture and di, totally
relaxed posture, as when the trap is punctured; dt, direction of
thrust of lateral hinge; 0/, longitudinal thrust of door; c, compo-
nent of the two thrusts; dh, inner angle of threshold, as also in fig. 5.

Fig. 4. — The door seen from in front. Inner course cells in double out-
line. Lettering as above.

PiQ ^ — View of pavement from above as from point d, fig. 3; oz, mz,
iz, outer, middle and inner zones of pavement; de, position of door

edge.
Fjg 6 _ Section through door and threshold taken as between c and d,

FIG. 3; I, set, and 2 relaxed posture of door, approximately.
fjG 7. _ Section of same through d, fig. 3. Broken lines indicate opening

flexures.
Fig. 8. — Section of same through c, fig. 3; mp, middle piece; /, rear
part of middle zone of pavement indicated in outline; d, dotted line
indicating the thrust of the door edge into the soft pavement cells
(arrows indicate the thrust of the lateral hinges).

— Plate 31.— ^ ^

Fig. I. — Entrance of Utricular ia caeriilea.

Fig. 2. — Trap of U. Gibbseae. ^

Fig. 3. — Trap of U. capensis.

Fig. 4. — Entrance of U . Welwitschii.

Fig. 5. — View from below of trap of U. rosea aff. (N. Queensland).

Fig. 6. — Entrance of U. Deightonii (Ms. name).

Fig. 7. — Front view of entrance of same.

Fig. 8. — Door of U. Welmtschii, showing the kriss tripping hair {see also

24 — 7).
Fig. 9. — Entrance of U. capensis. (

Fig. 10. — U. rosea, entrance, small trap. ^

Fig. II. ^ — The two forms of the trap of U. rosea. ^

— Plate J2. —

Fig. I. — Utricular ia orhiculata. Entrance.

Fig. 2. — Front view of the door, U. striatida.

Fig. 3. — U. striatida. Entrance.

Fig. 4. — Transverse section through the entrance, showing the front

view of the tripping hairs with their mucilaginous masses.
Fig. 5. — U. midticaulis, trap.
Fig. 6. — Development of the peculiar tripping glands of U. orhicidata,

numbered in series, 1-8. In 8 the mucilage mass has fallen away.
Fig. 7. — Utricidaria sp. (Thibet, Ludlow, Sherreff and Taylor, No.

5264, Brit. Mus.) to display the form of the forked rostrum.
Fig. S. — Utricidaria sp. (Thibet, L., S. and T. 5802), door and single

rostrum with long radiating fingers. i

Fig. 9. — Rostrum of U. midticaulis (Fig. 5) from below. |

1

— Plate 33. —

Fig. I. — Utricularia Kirkii. Entrance.

Fig. 2. - — Transverse sections of entrance of U. Kirkii through a, b, c,
and d, fig. i.

Fig. 3. — U. longiciliata, trap, side and front views.

Fig. 4. — • The same, entrance, showing single glandular tripping bristle.

Fig. 5. — U. lateriflora.

Fig. 6. — Utricularia sp. (Ceylon, Simpson No. 9482) related, if not iden-
tical, with U . calUphysa Stapf.

Fig. 7. — The same, in sagittal section.

Fig. 8. — U. Barnesii (Ms. name). Trap and views, from beneath, of
the lower and upper lateral combes.

Fig. 9. — U. lateriflora: set and relaxed postures of the door, traced from
photographs.

— Plate J4. —

Fig. I.- Ulricularia monanlhos, trap, studded with glands.

Fig. 2. — Entrance of same, sagittal section.

Fig. 3. — Same, in diagram to show set {di), relaxed (^2) and quite open
{dz) posture of the door. Arrows indicate water pressure on the door;
nip, middle piece, the arrow 4 indicating the direction of the thrust of
the lateral hinges on it; pmp, arrow indicates the direct thrust of
middle piece itself; ov, iv, outer and inner velum; pe, pavement; uh,
upper hinge; 2, inner angle of the threshold, th; j, line of attach-
ment of door to wall.

Fig. 4. — Section of trap so made as to allow one to look down through
the door (regarded as transparent) so as to see the opening guarded
by the outer velum. The two groups of glands on the door's outer
surface are shown. The abrupt bend in the threshold is indicated by
the transverse, curved dotted line. The other two dotted lines indicate
the limits of the lateral hinge. Lettering as in previous figure.

Fig. 5. — Section across the entrance showing the deeply curved concave
threshold (th) and the door fitting into it, held by the thrust of the
lateral hinges (long arrow) on the middle piece {mp, short arrow).
The dotted lines indicate flexures of the door edge on opening.

Fig. 6. — The various kinds of traps of U. volubilis. A, short stalked, basal
trap; B, long stalked trap, with foliaceous stalk; C, rostrum of A;
D, E, dorsal and lateral views of stolon trap.

Fig. 7. — Variants of the trap of U. Hooker i. A, large trap with deeply
laciniate wings; B, its rostrum and wings seen from above; C,
medium sized trap with broader wings; D, small trap with ventral
wings suppressed, the dorsal wings and rostrum very slender.

— Plate 55. —
Traps of Utricularia species (all of the U. monanthos type) :

Fig. I. — Utricularia Singeriana F. Muell.

Fig. 2. — U. Wallichiana Wight.

Fig. 3. — ■ U. Hamittonii F. E. Lloyd.

Fig. 4. — U. Dimstani F. E. Lloyd.

Fig. 5. — U. lasiocaidis F. Muell.

Fig. 6. — Looking into the entrance of any of these species, the outer,

circular velum, through the opening of which are seen the glands on

the door surface.
Fig. 7. — U. Moorei F. E. Lloyd.
Fig. 8. — U. tubulata F. Muell.

Fig. 9. — U. Holzii F. Muell. and U. albiflora R. Br.
Fig. 10. — U. Menziesii R. Br., seen from beneath.
Fig. II. — U. Menziesii, lateral view.

I
f

I

1

— Plate 36. —

Figs. 1-9, Polypompholyx; Figs, io-ii, U. tiihulata.

Fig. I. — Trap (side view) of P. mnUifida F. Muell.

Fig. 2. ^ — Same, sagittal section: c, antechamber; </, door; /c, large inter-
cellular space in the stalk; r', ridge along ventral surface of the |
stalk; s, a zone of hairs seen in fig. i. '

Fig. 3. — Transverse section, embracing part indicated by the parallel

broken hnes in fig. i, looking inwardly; s, space above door; d, ;

door; c, back wall of antechamber; r, n, ridge along the stalk; ic,
intercellular space in the stalk.

Fig. 4. — Trap, view from above. Arrows indicate directions of approach j

of prey, as also in fig. i. |

Fig. 5. — Entrance, showing the door in the relaxed posture. The broken * j

lines indicate the set posture (c/. 24 — 8). I

Fig. 6. — Trap of P. tenella, from below. '

Fig. 7. — Transverse section through body of trap of P. niuUifida, show-
ing the occurrence of deep compression cells.

Fig. 8. — Transverse section of the threshold of the large traps of P.
latifolia.

Fig. 9. — Same, of the small traps. i

Fig. 10. — Growing point of Ulricularia tubulata.

Fig. II. — A whorl of very young leaves and traps alternating at the same
level. Lateral view with growth apex above the numeral, axial view
below. '

I

— Plate 37. — ^

Moving pictures showing the action of the trap of Utrkularia, and of the sensitive hairs

of Dionaea (all frames are 1/16 sec. apart, except in fig. 2, in which they are 1/160 sec. ;

apart, and in fig. lA, which is time lapse); —

Fig. ik. — Utrkularia purpurea. Trap showing the exhaustion of the

contained water and the consequent collapse of the walls, indicated ^

by the distortion of a contained bubble of air. Moving picture: ,

time lapse spread over about 2 hours. J

Fig. iB. — The same, viewed edgewise, before and after action. i

Fig. 2. — U. vulgaris. View looking into the entrance showing the open- ,

ing of the door (in 1/160 sec.) and the subsequent closing in 4/160 \

sec. The open door is seen in the third frame from top. The round 1

object in front of the door is the knob of a glass probe. ;

Fig. 3. — t^. vulgaris. The capture of a copepod. The trap was set in a

shallow glass tank with walls to guide the copepod to the mouth of 1

the trap. '
Fig. 4. — The same. The lateral profile of a trap before and after {below)

actuation.
Fig. 5. — Dionaea muscipula. Bending and straightening of a sensitive

hair.
Fig 6 _The "Darwin experiment," referred to in the text: the sudden

disappearance of colored particles (here particles of carbon) resting on I

the door, on its actuation by a needle point slowly moved across the '■

entrance.
Fig. 7. — U. gibba, capturing a larva.

Fig. 8. — U. purpurea. A trap swallowing a bubble. _ =

Fig. 9. — U. vulgaris. Trap swallowing a glass bead and, in doing so,

jumping at the probe. The trap had been removed from the plant.

— Plate jS:—

Animation of the trap of U. gibba, by Mr. Harold Peberdy, illustrating
its action in the capture of prey (Courtesy of the Associated_Screen
News, Montreal).

— »*.

"V^

-1

^

Iffi

f^Shfi

'''-»

^^

3Uf -

'.ft...

i

i

W'

■ ^ Ni-^

•

w

.^♦-.-^

m&^ --

W^'

. 'r

' J\ •-

— .."sr-^

..

0

4

IP

t

5

1

^'

,J

Bpr:'

-»»■''.

■*^.^'-w

-•*..

■H

\««»<».-«M

P

^^^"'^

(

4

r^

P *

UttAXH)

StT PQSTURE

V

INDEX of PLANT and ANIMAL NAMES

ACACLA, 30, 56

Acaulopage, 173

Aegopodium podagraria, 119

Aldrovanda, i, 2, 5, 6, 7, 177, 181,
182, 186, i8g, 190, ig2, 193, 194,
195, 200, 203, 204, 205, 206, 208,
2og, 210, 211, 237, 239, 249, 267

• verticillata, 195

vesiculosa, 177, 194, 195, 210,

pi. 17, 19

var. australis, 195

Aldrovandia, 195. 210, 211

vesiculosa, 21c, 211

Amoeba, 170*, 173, 174, 175
proteus, 17s

sphaeronucleus, 174

terricola, 173

verrucosa, 17s

Anabaena azoUae, 3
Andromeda polifolia, 115
Anguillulidae, 169
Anopheles, 267

albimanus, 267

Anuria, 172
Aphanomyces astaci, 176

Magnusi, 176

ovidestruens, 176

Apogonia spherica, 73
Aquilegia, 3
Aranella fimbriata, 215
Aristolochia, 3, 214
Arthrobotrys, 4, 170, 171

oligospora, 169, 17c*

Ascaris, 36

lumbricoides, 36

Asclepias, 4

Cornuti, 8

Auricularia, 119
AzoUa, 3

Bacillus aquaticus communis, 265

Bacterium coli, 265

fluorescens liquefaciens, 72, 75

prodigiosum, 75

Bdellospora helicoides, 174
Begonia, 132, 134

Biovularia, i, 2, 7, 213, 214, 217,
227, 228, 249, 257, 258

minima, 213, 227, 228

olivacea, 90, 213, 227, 228, pi.

23, 27
Brachionus, 172
Brachydeutera argentata, 214
Brocchinia, 226
Bufo marinus, 252, pi. 20
Bulbophyllum, 65
Byblidaceae, i, 95, 98
Byblis, I, 2, 5, 35, 95, 97, 98, 99,

100, 268, pi. 13, 14

gigantea, 95, 96, 98, 99, 118,

pi. Q, 13, 14

linifolia, 95, g6, 99, 117, pi. 14

Caltha dioneaefolia, 5, 6
Cecropia, 56
Cephalotaceae, i

Cephalotus, i, 2, 15, 16, 38, 49, 62,
79, 82, 86, 89, 264

follicularis, 40, 81, 89, pi. g, 10

Chironomidae, 78
Chlorion Harrisi, 38
Chrysamphora (see Darlingtonia), i,

Cladonia uncinalis, 115

Cladophora, 171

Cochlonema dolichosporum, 174

verrucosum, 174

Codiaeum, 59, 60
Colura, 3

Colurus, 171, 172

Coprinus, 4

atramentarius, 92

Cordyceps, 2, 169, 174
Cosmiza, 215
Cotyledon, 67
Crinum, 65
Culicidae, 78, 267
Cyclops, 234

Dactylaria, 170

Candida, 171

Dactylella, 2, 170, 175

bembicoides, 170*, 175

• tylopaga, 170*, 173

Daphneae, 234

Daphnia, 234

Darlingtonia = Chrysamphora, i,
2, 10, 23, 24, 25, 26, 28, i^, 35,
38, 40, 42, 43, 44, 45, 46, 47. 48,
49, 69, 83

californica, 36, 39, 40, 49, 50,

pi. 4, 5, 6

Delphinium, 3

Diaptomus, 176

Dionaea, i, 2, 5, 6, 7, 105, 109, 135,
165, 177, 178, 181, 182, 186, 187,
188, 191, 192, 193, 194, 195, 196,

197, 199, 200, 2CI, 202, 203, 204,

205, 209, 210, 211, 239, pi. 17, 37

muscipula, 138. 177, 178, 179,

210, 211, pi. 17, 18, 37
Dipsacus sylvestris, 3, 8
Dischidia, 3, 65

pectinoides, 3

Distomum hepaticum, 113

Distyla, 172

Dorniphora venusta, 38, 39

Drosera, 1, 2, 4, 5, 7, 35, 39, 7°, 89,
98, 99, 100, 102, 104, 105, 108,
109, no, 113, 115, 116, 117, 118,
119, 120, 124, 129, 134, 135, 136,

138, 139, 141, 150, 151. 158, 159,
160, 161, 162, 163, 164, 165, 166,
167, 168, 178, 182, 194, 2CI, 210,

211, 245, pi. IS, 16
anglica, 115, 157

auriculata, 118, 130, 168

binata, 116, 117, 118, 131, 132,

135, 139, 143, 144, 145, 157, 167

capensis, 116, 118, 119, 128,

132, 134, 135, 139, 143, 145, 161,
165, 168, pi. 13, IS

■ dichotoma, 70, 116, 117, 118,

139, 159, 166

erythrorhiza, 117, pi. IS

filiformis, 115, 117, 118, 134,

13s, 163, 165

gigantea, 115, 116, 118, 120,

126, 139, pi. IS, 17 t

intermedia, 115, 116, 131, 132,

139, 143, 144, 146, 165, 166, 167

longifoha, 131, 138, 166, 168

lunata, 118

paleacea, 120

peltata, 118, 120, 130, 134, 168,

pi. 16

■ pygmaea, 117, 118, 120, 122,

130, 131, 166, pi. IS, 16

regia, 115, 116, 117, 118, 168

rotundifolia, 115, 116, 117, 118,

119, 120, 130, 131, 132, 133, 134,
13s, J38, 139, 142*, 143, 144, 157,
160, 163, 165, 166, 167, 168, pi.
13, IS, 16, 17

spathulata, 133, I43

subhirtella, 118

VVhittakeri, 115, 118, 120, 123,

125, 165, 167, 168, pi. /i

Droseraceae, i, 95, 166, 177, 210

Drosophila, 112, 139

Drosophyllum, i, 2, 5, 96, 99, 100,
loi, 102, 104, 105, 117, 118, 122,
124, 137, 152, 155, 166, 168

lusitanicum, 99, 105, pi. 13, 14

Empusa, 2

Endocochlus asteroides, 174
Erica tetraU.x, 5, 109
Eriophorum vaginatum, 115
Euglena, 172, 265
Exyra, 36, 37

■ Ridingsii, 36, 37

■ Rolandiana, 36, 37, 38

semicrocea, 36, 37

Flagellarla, 59
Frullania cornigera, 3

Galium verum, 112, 161
GenUsea, i, 2, 4, 7, go, 94, 106, 213,
216, 217, 230, pi. II, 12

filiformis, 90

ornata, 90, 94, pi. IJ

repens, 92, pi. 9

Gloriosa, 59
Gossypium, 60

Heliamphora, I, 2, g, 10, 11, 14,
IS. 16, 25, 30, 39, 43, 69, 89, 98,
211

• Macdonaldae, 9, 12, 13*, pi. I

minor, 9, 11, 12, 13*, 16

nutans, 9, 11, 12, 13*, 16, pi.

I, 2

Tatei, 9, 12, 13*

■ Tyleri, 11, 12, 13, 16

Heteronema, 265
Hydrolea, 5

Impatiens, igi, 192

Iris, 22, 2g, 30, 31, 32, 39, 61, 67

pallida, 67

Lathraea squamaria, 3

Lejeunea, 3
Lemna, 119

Lentibulariaceae, i, 7, 94, 95. 213,
228, 267, 268, 270

Marcgravl-^, 3
Marchantia, 131
Martynia, 5

Megarhinus acaudatus, 77
Metopidia, 171, 172
Metriocnemus Edwardsii, 36

Knabi, 36

Mimosa, 142, 179, 189, igi
Mirabilis longifolia, 5
Misumenops nepenthicola, 77

Thienemannii, 77

Molinia coerulea, 6, 8
Monacrosporium, 170
Monostyla, 171, 172

Nelumbo, 30

Neosciara Macfarlanei, 38

Nepenthaceae, i. 79, 80

Nepenthes, i, 2, 10, 16, 30, 31, 35,
38, 39, SI, 52, 59, 60, 63* 64, 65,
66, 67, 69, 70, 71. 73, 74, 75. 76,
77, 78, 79, 80, 81, 82, 86, 88, 8g,
98, 114, 160, 161, 163, 164, 166,
168, 211, 212, pi. 7, S

ampullaria, 52, 53, 54. 55. 56,

57, 58, 64, 65. 66, 77, 78, 79, pl-
4,7,8

Balfouriana, pl. 4, 7

bicalcarata, 52, 55, 56, 57, 58,

59, 62, 63*, 66, pl. 4, 7

borneensis, 59

Boschiana, 55

Burbidgei, 56

clipeata, 55, 60, 61

destillatoria, 51, 76, 79

dubia, 55, 57

Edwardsiana, 56

formosa, 62

Francis E. Lloyd

— 349 —

Carnivorous Plants

gracilis, 51, 55, pL 7

gracillima, 65, 67

Hibberdii, 75

• inermis, 55. 57. 64. 66, pi. 7

intermedia, 55, S6. 58, 59. 62,

63*, pi. 7

Ladenburgii, 57

Lowii, 55, 57. 58. 64, pi. 7. *

Mastersiana, 68, 73, pi. 4

maxima, 55, 57

melamphora, 80

mixta, 75, pi. 4

Morganiana, 75

paradii^iaca, 72

Pervillei, 68, pi. 7

phyllamphora, S5, 57. 80

Rafflesiana, 52, 5g

■ var. nivea, 56

Rajah, 51, 56, 57

stenophylla, 65

• Tiveyi, 57, 64, pl- 7, ^

Veitchii, 51, 52. 57, 80

ventricosa, 55. 56. 57. 66, 68,

78, pl. 4, 7

Vieillardii, 55

villosa, 51, 64

Nicotiana tabacum, s
Noteus, 172
Nymphaea, 32

Papaipema appassionata, 36

Paramaecium, 172, 265

Paranema, 172

Pecten, 204

Pedilospora dactylopaga, 173. i75

Pelargonium zonale, 5

Phacus, 265

Phalangium, pl. 17

Phaseolus multiflorus, 212

Philodina, 172

Phoridae, 78

Phragmites, 209

Physiotum, 3

majus, 3

Pinguicula, i, 2, 5, 7, 95. 106, io7.

108, 109, no. III, 112, 113, 114.

119, 161, 162, 163, 164, 165, 166,

168, 213, 264, pl. IS

alpina, 114

cuneata, 106

gypsicola, 105. 106

hirtiflora, no

vulgaris, 106, 109, 112, 114,

119, pl. 13, IS

Pittosporaceae, 95

Pleurothallus, 65

Podostemonaceae, 226

Polypompholyx, i, 2, 7, 60, 95, 98,
213, 215, 216, 217, 218, 229, 230,
254, 256, 257, 262, 268, 269, pl.
26, 36

latifolia, 230, 262, 263, pl. 36

multifida, 95, pl. 24, 36

tenella, 95, 230, pl. 20, 22, 36

Pothos, 60

Primula sinensis, s

purpurea, 17, 18, 19, 22, 23, 24,

25, 26, 27, 28, 31, 33. 34. 35. 36.
37. 38, 39. 43, 45. 46. 5°. 146, pl-
I, 2, 3

gibbosa, 17, 18

var. stolonifera. 39

venosa, 17, i?

RhYNCHOSPORA alba, IIS

Ricinus communis, 74
Roridula, i, 5, 8, 78, 98
Rotatoria, 171, 172, 175, 176
Rotifer, 172

Salicornia, 65, 79

Salpina, 171, 172

Sarcophaga, 38, 77

Sarracena, 18

Sarracenia, i, 2, 9, 10, 11, 15, 16,
17, 18, 19, 20, 22, 24, 30, 32, 33,
34, 35. 38, 40, 43, 45, 46, 47, 49,
S9, 62, 69, 81, 82, 83, 86, 89, 146,
163, 245

adunca, 18, 49, 66, 80

Courtii, 17, 26

Drummondii, 17, 27, 29, 30,

33, 35, 37, 43. pl- I
flava, 17, 21, 29, 30, 31, 33, 35,

37. 38. 39, 46. 77, pl- I

■ Jonesii, 17, 30, pl. 3

■ minor, 17, 26, 27, 28, 29, 31,

33, 34, 37, 43, pl- ', 3

oreophila, 17

psittacina, 17, 19, 22, 23, 24,

25. 26, 27, is, 34, 37, 43, pl. I, 3, S

rubra, 17, 33, 34. 37

Sledgei, 17, 33, 35, 37

variolaris, 39, 49

Sarraceniaceae, i, 39, 40. 45. 48
Saxifraga rotundifolla, 5

umbrosa, 5

Scheuchzeria palustris, 115

Schrankia, 178

Scirpus caespitosus, 115

Sequoia gigantea, 115

Sicyos, 192

Silene, s

SommerstorfBa spinosa, 173, i75

Sparmannia, s

Sphagnum, 19, 115, 116, 133

Spirillum, 206

Spirogyra, 207, 211

Squalella, 172

Stentor, 172

Stylonychia, 172

Stylopage, 175

Syngonium podophyllum, 60

Tarsius spectrum, 58
Telenchus scandens, 169
Tetralobus, 215
Thomisus callidus, 77

nepenthephilus, 77

Tillandsia, 213, 226
T-orenia, 132
Tradescantia, 18
Trichothecium. 170
Tussilago farfara, 119
Typha, 209

Utricularia, I, 2, 3, 6, 7, 8, 60, 89,
go, gi, 92, 94, 95, 106, 127, 152,
160, 162, 167, 168, 194, 199, 204,
207, 2og, 210, 211, 213, 214, 216,
2ig, 222, 226, 227, 228, 231, 236,
237, 23g, 242, 244. 247. 257, 263,
264, 265, 266, 267, 268, 269, 270,
pl. 2Q, 32, 33, 35, 37

albiflora, 230, pl. 33

amethystina, 21^. 257

Barnesii, 232, pl. 22, 33

Berendii, 7

bifida, 217, 218, 259, pl. 22

biloba, 225, 257, pl. 21, 27

brachiata, 216, 232, 260, 267

Bremii, 225

caerulea, 231, 255*, 256, 258,

259, pl- 24, 31

calliphysa, 232, 260, pl. 33

Campbelliana, 226

capensis, 217, 218, 231, 254,

255*. 256, 259, 260, 269, pl. 20,

22, 24, 2Q, 31

capilliflora, 214, 215*, 218, 230

clandestina 225, pl. 23

— ,- cleistogama, 217, 221*, 235

■ cornuta, 146, 227, 231, 245,

256, 258, 259, 265, pl. 24, 30

cyanea, 259

cymbantha, 213, 214, 222, 224,

f'- ^-' , ..

Deightonr, 259, pl. 25, 31

dichotoma, 229, 230, 254, 261,

270, pl. 24

Dunstani, 214, 215*, 230, pl.

35

Dusenii, 213, 216, 226, 227

elephas, 219, 228, 258, pl. 27

emarginata, 217, 218, 219, 222,

224, 267, pl. 20. 21, 2S

Endresii, 213, 226

equiseticaulis, 259

exoleta, 216, 217, 218, 219, 222,

223, 224, pl- 21, 26, 27

fimbriata, 215

tlexuosa, 220, 223, 237, 238

249, 254, 268, pl. 20

gibba, 222, 224, 227, 230, 244,

245, 252, 253*. 257, pl. 21, 24, 2S,
26, 2Q, 37, 3S

Gibbseae, pl. 31

globulariaefolia, 213, 232, 257,

pl. 28

■ Hamiltonii, pl. 35

Holzii, pl. 35

■ Hookeri, 95, 218, 229, 230, 257,

261, pl. 34

Humboldtii, 213, 226

intlata, 2:0, 221, 223

intermedia, 225, pl. 20, 21, 25

Kirkii, 232, 260, pl. 33

lasiocaulis, pl. 35

lateriflora, 232, 256, 260, pl. 20,

Lloydii, 232, 257, pl. 28

longiciliata, 231, 232, 260, pl.

33
■ longifolia, 213, 218, 226, 227,

257, pl- 26
Menziesii, 60, 95, 229, 230, 231,

261, pl. 20, 23, 35
minor, 214, 223, 224, 225, 242,

245, 268, pl. 21

mixta, 222, 223

monanthos, 215, 217, 218, 229,

230, 254, 256, 257, 261, pl. 22, 23,

24, 34, 35

montana, 219, 226, 227

Moorei, pl. 35

multi aulis, 232, 260, pl. 32

nana, 232, 246, 257, pl. 28

■ nelumbifolia, 213, 217, 219,

226

neottioides, 217, 226, 257, pl-

26. 27

ochroleuca, 225

ogmosperma, 259

• oligosperma, 216, 218, 220, 222,

223, 233, 246, pl. 21, 23
olivacea. 227

orbiculata, 215, 216, 217, 225,

231, 232, 260, pl. 22, 32
paradoxa, 225, pl. 21

peltata, 232, 259, pl. 25, 29

■ purpurea, 214, 215, 216, 219,

221*, 227, 228, 229,235,238,241,

257, 258, 269, pl. 20,26,27, 29, 37

radiata, 220

Rendlei, 226

reniformis, 213, 216, 217, 219,

226, 227, 257, pl. 21

resupinala, 216, 225, 226, 268,

pl. 23, 27

■ rigida, 216, 226

■ rosea, 231, 260, pl. 22, 31

simnlex, 213, 231, 232, 260,

pl- 23

Singeriana, pl. 35

stellaris, 220, 221, 223, pl. 20

Stephensae, 214, 224, pl. 23

striatula, 232, pl. 32

subulata, 215, 216, 226, 231,

232, 257

Thonningii, 220

tubulata, 215, 2ig, 229, 230,

261, pl. 33, 36

violacea, 218, 229, 230

volubilis, 229, 230, 231, 261,

pl. 20, 23, 34

vulgaris, 94, 213, 215, 216, 217,

218, 2ig, 220, 221, 222, 223, 224,
22s, 226, 227, 233, 235*, 237. 241.
242, 243, 244, 245, 249, 254. 255 .
257, 26s, 267, 268, 269, 270, pl.
20, 21, 23, 26, 2g, 37

americana, 246

Wallichiana, pl. 35

Warburgii, 231, 260

Welwitschii, 60, 256, 259, pl.

24, 31
Utriculariaceae, 269
Utriculariae, 215

VaCCINIUM OXYCOCCUS, IIS
WllHELMINA NEPENTHICOLA, 83

Wyeomyia Smithii, 36

Zizanta, 2og
Zoopagaceae, 175
Zoopage phanera, 174
Zoophagus, 4. 175

■ insidians, 17c*, 171, i73, i75.

176

AUTHOR INDEX

Abderhalden, 71, 7g, 159

Adowa, A. N., 264, 267

Akerman, A., 127, 129, 148, 149,

150, 151, 153, 155, i6i, 165, pi. 17
Aldrovandi, U., 195
Amadei, C, 195
Ames, M. E. P., 49
Ames, O., 133, 165
Anderson, L. E., 30
Arber, A., id, 16, 23, 38, 49, 6o, 62,

79, 82, 83, 85, 89
Arisz, W. H., 129, 165
Armit, W. E., 229
Arnaudow, N., 173, 175
Arnold, W., 119
Asenjo, C. F., 36, 38
Ashida, Joji, 186, 189, 19^1, 192, 193,

194, 195, 197, 201, 202, 203, 204,

205, 206, 207, 208, 209, 210, 211,

212, pi. 18, IQ
Aso, R., HI, 114, 264
Austin, R. M. L., 36, 48, 49

Baas Becking, L. G. M., 69, 75,

78, 161

Baillon, H., 30, 38, 59
Balfour, T. A. G., 194, 210
Barnhart, J. H., 49, 53, 215, 228,

231, 267
Barrett, C, 96
Bartram, Ch., 18
Bartram, J., 178
Bartram, Wm., 18, 32, 38
Batalin, A., 21, 32, 38, 46, 48, 49,

106, log, 114, 142, 159, 165, 188,

189, 19:1, 191, 193, 210
Bath, W., 267
Beck, A. B., 165, 167
Behre, K., 130, 131, 132, 133, 134,

138, 14.3. 145, 164, 165, pl. 16
Beijerinck, M. VV., 131, 165
Benecke, W., 147, 165, 167
Benjamin, L., 90, 94, 234, 243, 248,

267
Bennett, A. W., 165
Bentham, G., 9, 10, 14, 16, 95, 262
Berger, J., 36, 38
Bischoflt, G. W., 53, 79
Bobisut, O., 66, 67, 68, 79
Bok, R., 69
Bokorny, T., 129, 147, 148, 151,

153, 166
Bower, F. O., 53, 54, 57, 59, 60, 62,

79, 80, */. 7
Brackenridge, J. D., 40, 49
Brass, L. J., 229

Braun, A., 49

Brocher, F., 233, 235, 236, 237, 238,

239, 240, 241, 242, 243, 244, 246,

267
Brongniart, A. T., 64, 79
Brough, P., 131, pl. 16
Broussonet, P. M. A., 186, 210
Brown, R., 81, 89
Brown, W. H., 6, 186, 187, 188, 190,

191, 192, 193, 201, 205, 210
Bruce, A. N., 97, 98
Bruecke, 189
Brug, S. L., 80
Brumpt, E., 267
Buchenau, F., 221, 222, 223, 224,

267
Budde, E., 175
Bunning, E., 206, 212
Biisgen, M., 89, 162, 163, 166, 236,

247, 248, 249, 264, 267
Buller, A. H. R., 68, 92, 94, 131
Burbidge, F. W. T., 52, 58, 59, 79
Burdon-Sanderson, J., 186, 187, 188,

189, 19^, 205, 210, 211
Burnett, G. T., 18, 38
Buscaglione, 5, 8, 209, 211, 268
Buzacott, J. H., 252, pl. 20

Canby, W. M., 38, 45, 49, 177, 180,
210, 211

CandoUe, A. P. de, 59, 79, 138, 166,

215, 234, 267
Candolle, C. P. de, 60, 79, 182, 189,

210
Carry, T. H., 4, 8
Caruel, T., 7
Caspary, R., 195, 196, 197, 199,

210
Catalan, M., 179
Catesby, M., 18, 39
Chandler, B., 267
Chenon, L. J., 195
China, W. R., 98
Christy, M., 4, 8
Clairville, J. P. de, 234
Clarke, W. G., 267
Clausen, J., 112, 113
Clautriau, G., 73, 75, 79
Clusius, C, 18
Coelingh, W. M., 127, 134, 149, 150,

151, 156, 166
Cohn, F., 19s, 196, 199, 201, 209,

210, 21-', 234, 235, 236, 239, 243,

246, 249, 259, 264, 266, 267
Coker, W. C, 210, 212
CoUa, S., 112, 114
Collinson, P., 18, 32, 39, 178, 210
Commelin, J., 51
Compton, R. H., 216, 260, 267
Correns, C, 116, 166
Couch, J. N., 170, 175
Couvreur, 71, 72, 79
Cramer, C., 8, 209, 210, 211
Crouan, 234, 267
Curry, D. P., 267
Curtis, M. A., 177, 178, 180, 184,

186, 188, 210
Curtis. W.. 79, 180, 210

Czaja, A. T., 116, 127, 166, 197,
201, 205, 206, 207, 208, 209, 210,
236, 239, 240, 241, 243, 244, 246,
249, 251, 253, 267. 268

Czapek, F., 79, 112, 114

Dakin, W. J., 84, 87, 88, 89

Danser, B. H., 51, 55, 57, 58, 65,
78, 79

Darlington, W., 40

Darwin, C, 5, 7, 8, 32, 39, 49, 69,
70, 90, 92, 94, 97, 99, 100, lOI,
102, 103, 105, 106, 107, 108, 109,
110, III, 112, 113, 114, 115, 118.

120, 121, 122, 125, 126, 127, 128,

129, 136, 139, 140, 141, 142, 143,
145, 146, 147, 148, 149, ISO, ISI.
153, 155. 156, 158, 159, 161, 162.
166, 177, 180, 1S2, 183, 1S4, 186,

187, 188, i8g, 190, 191, 193, 194,
195, 199, 200, 208, 209, 210, 211,
234, 235, 236, 243, 244, 245, 246,
248, 250, 259, 264, 266, 268

Darwin, E., 118, 135. 166, 168, 179,

180, 210
Darwin, F., 4, 8, Sg, 90, 112, 146,

147, 162, 163, 166, 193, 210
Dean, B., 214, 243, 268
Delpino, F., 5, 6
Dernby, K. G., iii, 113, 114, 159,

161, 162, 166
Dewevre, A., 100, loi, 103, 105
Dickson, A., 52, 53, 59, 61, 62, 64,

66, 79, 81, 82, 83, 84, 85, 87, 88,

89
Diderot, D,, 179, 210
Diels, L., 96, 98, 117, n8, 120, 130,

164, 166, 210
Dillon, J. J., 195
Dillwyn, L. W., 178, 210
Dixon, H. H., 133, 166, 224
Dobbs, A., 178
Domin, K., 95, 98
Dover, C, 77, 79
Drechsler, C, 170, 171, 173, 174,

175
Drude, O., 8, 166, 268
Dubois, R., 71, 72, 79, 159, 166

Duchartre, P. E., 30, 39
Dufrenoy, J., 152, 153, 155, 166
Duval-Jouve, J., 65, 79, 195, 200,
21C

Ebstein, 70

Edwards, H., 35, 40, 45, 48, 49

Edwards, S., 180

Eichler, A. W., 82, 89

Ekambaram, T., 237, 238, 240, 241,

243, 248, 249, 250, 268
Ellis, J., 138. 177, 178, 179, 183,

186, 210, 211
Emerson, 7
Engel, E. O., So
Engler, A., 7, 8g
d'Entrecasteau, 81

Faivre, E., 59, 79
Fenner, C. A., 15, 16, 21, 22, 23,
33, 39, 65, 73, 74, 79, 96, 97, 98,

99, loi, 102, 104, 105, 107, 108,
109, 114, 117, 120, 121, 122, 123,
124, 125, 126, 127, 129, 166, 195,
196, 197, 199, 200, 205, 209, 211

Fermi, C, 5, 8, 209, 211, 268

Fernald, M. L., 226, 268

Fernandes. A., 105

Flacourt, E. de, 51

Franca, C, 8, 99, 100, 103, 104, 105,

122, 155, 166, 214, 243, 268
Frankland, 166
Fraustadt, A., 182, 194, 211
Fresenius, 169. 175
Fry, R. E., 254

Gardiner, W., 122, 146, 147, 148,
149, 151, 153. 157, i66. 211

Gardner, C. A., 81, 89

Gardner, G., 213, 226, 268

Gardom, 135, 136

Gates, F. C, 268

Geddes, P., 113, 114, 136, 166, 268

Geitler, L., 175

Gerard, J., 4, 18

Gibbs, R. D., 268

Gicklhorn, J., 170, 172, 173, 175, 176

Giesenhagen, 254

Giessler, A., 119, 166

Gilburt, W. H., 65, 79, 89

Girard, F. de, 94

Gislen, T., 247, 26S

Glauer, 147, 166

Gleason, H. A., 9, 12, 13, 16

Gluck, H., 223, 224, 225, 268

Goebel, K., 3, 6, 7, 8, 9, to, 11, 14,
15, 16, 19, 22, 23, 26, 29, 30, 31,
32, il, 39, 40, 42, 43, 45, 46, 47,
48, 49, 52, 53, 54, 55. 59, 60, 61,
62, 66, 68, 71, 72, 73, 76, 79, 80,
82,83,86, 87,88,89,90,91,92,94,

100, loi, 102, 103, 105, 106, 107,
108, 109, 110, III, 113, 114, 116,
117, 118, 122, 124, 127, 130, 131,
134, 139, 147, 148, 151, 153, 154,
162, 163, 164, 166, 177, 180, 181,
182, 184, 185, 188, 194, 195, 196,
199, 200, 201, 211, 217, 218, 219,
220, 222, 223, 224, 225, 226, 227,
228, 231, 232, 234, 235, 243, 245,
246, 247, 254, 259, 260, 262, 263,
264, 268, pl. 2, J, 6, 16

Gorup-Besanez, E. von, 69, 70, 79,

HO, 114
Graves, J. A., 133, 166
Gray, A., 30, 39, 48, 49, 136, 235
Green, J. R,, 71, 79, 107, 114, 161,

166
Gressner, H., 107, 114
Grimm, H, N., 72, 79
Grisebach, H. R. A., 94, 227
Gronland, J., 120, 121, 122, 166
Grout, A. J., 133, 166
Gruetzner, 70
Guenther, 79
Gurney, R., 267, 268

Francis E. Lloyd

— 351 —

Carnivorous Plants

Gustafsson, 0., 112

Guttenberg, H. von, 182, 188, igi,

192, 211, 212

Haas, Th. P., pi. g
Haberlandt, G., 66, 68, 73, 79, io4,
122, 124, 163, 166, 182, 185, 188,

193, I95i 200, 211
Hada, Y., 241, 265, 268
Haidenhain, 70, 71

Hamilton, A. G., 82, 83, 84, 85, 86,

87. 89, 98
Harshberger, J. W., 99. lOS
Hausleutner, 209, 211
Hayne, F. G., 138, 166
Hegner, R. W., 39, 24°. 241, 265,

268
Heide, F., 58, 62, 64, 79
Heinricher, E., 79, 116, 166, 211
Hepburn, J. S., 23, ^t,, 34, 35, 36,

38, 39, 48, 49, 74. 76, 77, 79
Hermann, P., 51
Higley, 32, 39
Hildebrand, 6
Hoehne, F. C, 90, 227, 268
Holm, T., 181, 211
Holier, H., 160, 167
Homes, M. V. L., 105, 120, 122, 123,

124, 125, 136, 151, 152, 153, 154,

155. 157. 166, p\. 16
Hooker, H. D., Jr., 121, 142, 143.

144, 166
Hooker, J. D., 5, 8, 13, 18, 19, 20,

21, 23, 32, 39, 42, 49, 52, S3, 54,

55. 57, 58, 59. 61, 62, 63, 64, 66,

68, 69, 79, 13s, 136, 166, 179. 211.

212, pi. 7
Hornemann, J. W.. 113, 114
Hovelacque, M., 107, 114, 226, 243,

268
Huie, L. M., 120, 121, 122, 125, 127,

151, 157, 166
Hunt, J. G., 64, 80

Im Thurn, E. F., 9, 16, 226, 268
Irmisch, T., 222, 223, 268

Jane, F. W., 217, 218, 268

Janse, 163, 167

Janson, E., 119, 129, 148, 151, 167

Janti, J. V. de, 8o

Jensen, H., 76, 77, 80

Johnson, Th., 18

Jones, F. M., 8, a, 36, 37, 38, 39,

45, 48, 49, 77, 79, 178, 183, 184,

188, 211, pi. 4
Jost, L., 147, 165, 167

Kamienski, F., 213, 214, 218, 227,

228, 268
Keck, K., 49

Kellermann, C, 162, 163, 167
Kiesel, A., 265, 268
Killip. E. F., 16
Kirschleger, M., 131, 167
Klein. J., 114

Knoll, F., 66, 67, 68, 69, 80
Kny, L., 66, 80
Kok, A. C. A., 127, 128, 167
Konig, C, 180
Konopka, K., 123, 124, 125, 126,

127, 136, 157, 158, 167
KonsuloS, S., 176
Korthals, P. W., 51, 53, 64, 80
Korzschinski, S., 195, 196, 211
Kostytschew, S., 119, 167, 194, 211
Krafft, S., 9, II, 14, 15, 16, 39, pi. 2
Kramer, W. de, 75, 76
Kruck, M., 127, 152, 167, 239, 240,
_ 241, 243, 244, 24s, 268
Kuhlmann, J. G., 268
Kupper, W., 67
Kurtz, F., 40, 42, 43, 49, 182, 211

Labbe, E., 160, 167

La Billardiere, J. J. Houton de, 81

Lambert, G., 33, 39

Lang, F. X., 95, 96, 98, 216, 262,

263, 268
Lassus, A. de, 201, 211
Leavitt, R. G., 130, 131, 132, 133,

167
Lehmann, 215
Lemmon, J. G., 42, 45, 49

Linderstr0m-Lang, 160, 167

Lindley, J., 30, 39, 183, 211

Lindman, C. A. M., 112, 114

Linnaeus, C., 17, 18, 32, 39, 112,
1,58. 178, 179, 19s, 210, 211

Lipman, 180

Lloyd, F. E., s. 8, 9. 10, 13. 16, 78,
80, 87, 89, 95, 98, 211, 214, 216,
218, 229, 233, 238, 240, 241, 243,
251, 254, 256, 257, 258, 259, 260,
268, 269

rObel, M. de, 17

Loew, O., Ill, 114, 119, 148, 264

Loureiro, J., 52

Lubbock, J., 130, 167

Ludlow, pi. 32

Ludwig, F., 6, 8

Luetzelburg, P. von, 226, 228, 234,
235, 243, 258, 264, 265, 269

Macbeth, A. K., Jr., 131, 165, 167

Macbride, J., 18, 19, 32, 39, 49, 66,
80

McClosky, 47

MacDougal, D. T., 19, 21, 22, 39,
163, 167, 206, 211

Macfarlane, J. M., 9, 10, 16, 19, 20,
24, 25, 28, 30, 39, 41, 42. 43. 45,
49, 52, 53, 55. 56. 57, 59, 60, 61
62, 64, 65, 66, 68, 74, 79, 80, 183,
185, 186, 187, 188, 190, 191, 192,
211, pi. 7

McGee, J. M., 168

McLean, R. C, 113, 114

Macnamara, C, pi. i

Mangin, L., 65, 80

Marloth, R., 117, 167

Marshall, W., 108

Massart, J., 163, 167

Matheson, R., 214, 251, 269

Maury, P., 89

Mayr, F. X., 244, 269

Meehan, 210

Meierhofer, H., 220, 235, 243, 245,
249, 250, 269

Mellichamp, J. H., 17, 19, 26, 32,

39, 49
Menzel, R., 80
Menzies, A., 81
Merl, E., 127, 201, 216, 217, 218,

219, 237, 238, 239, 240, 241, 243,

249, 260, 269
Merz, M., 216, 217, 269
Metzner, P., 206, 211
Meyen, F. J. F., 64, 80, 121, 138,

167, 180, 187, 188, 211, 234, 269
Meyer, A., 100, loi, 103, 105
Meyer, F. J., 244, 269
Milde, J., 138, 167
Mirande, R., 172, 176
MirimanoEf, A., iii, 112, 114, 167
Mohnike, 72, 80
Monti, G., 19s, 197, 201, 211
Mori, A., 200, 201, 211
Morison, R., 18
Morren, Ch., 30, 39, 60, 80
Morren, E., 5, no, 114, 159, 167,

179, 210, 269
Morrison, A., 130, 167
Mosely, H. N.. 269
Moulaert, B., 133, 167
von Mueller, 229, 230
Munk, H., 185, 186, 187, 188, 189,

211

Naudin, M., 121, 131, 167

Neubauer, M., 165, 167

Nicolosi-Roncati, F., no, 114

Niel, C. B. van, 120

Nitschke, T., 115, 116, 118, 120, 122,
125, 127, 130, 131, 136, 138, 139,
140, 141, 142, 167, 196, 211

Nold, R. H., 127, 241, 242, 243, 244,
269

Dels. W.. 167

Okahara, K., 159, 160, 161, 162,

167, 168
Oliver, D., 16, 268, 269
Olivet, R., 112, 114
Olsen-Sopp, 113
Gosterhuis, J., 164, 165, 16S
Oppenheimer, C., 112, 114, 269
Osborn, C. S., 8

Oudemans, C. A. J. A., 55, 64, 65,
66, 80, 177, 180, 182, 184, 187,
194, 211

Oudman, J., 127, 128, 129, 165, 168

Oye, P. van, 80

Page, 210

Peberdy, H., pi. 38

Peirce, J. G., 7, 8

Penzig, O., 100, loi, 105, 120

Peyronel. B., 164. 168

Pfeffer, W., 2, 8, 112, 127, 141, 146,

163. 164, 168, 194, 209, 211
Planchon, J.-E., 8, 95, 98
Plukenet, L., 196, 211
Porsild. M. P.. 269
Prantl, H., 7, 89
Prat, S., 240, 269
Price, J. R., 131, 167
Pringsheim, N., 218, 220, 222, 223,

224, 269

QUINTANILHA, A., 99, 102, IO4, 10$,
152, 153, 168, pi. 13

Raumer, E. von, 162, 163, 167
Rees, M., 70, 80, no, 114, 159, 168,

194, 211
Regel, E., 163, 168
Rennerfelt, E., 176
Rennie, E. D., 130, 168
Ridley, H. N., 269
Riley, C. v., 38, 39
Robinson, Heath, 267
Robinson, \V. J., a, 34, 39, 76, 80,

132, 133. 141, 159. 168
Roeper, J., 166
Rosenberg, O., 126, 127, 128, 157,

158, 168
Ross, H., 9=;. gS
Rossbach, G. B., 269
Roth, A. \V., 135, 138, 168
Roxburgh, 195
Ruiter, C. de, 58, 80
Rumphius, G. E., 77, 78
Ruschmann, G., 127, 164, 168

S.\CHS, J., 55

Saint-Hilaire, A. de, 30, 39, 90, 94,

269
St. John, E. Q., 33, 35. 39. 48, 49.

79
Salisbur>', 133
Sarrazin, M. S., 17, 18
Scarth, G. W., 207, 211
Schacht, 121
Schenck, H., 269
Schenk, A., 8, 166, 209, 211, 268
Scherffel, A., 176
Schikora, F., 176
Schimper, A. F. W., 32, 33, 39, 146,

148, 150, 151, 153, 168, 231, 245,

258, 259, 269
Schleiden, M. J., 234
Schhcht, 163, 168
Schmid, G., 117, 118, 119, 126, 130,

141, 142. 162, 164, 168, 186
Schmitz, P. H. S. J., 80
de Schoenfelt, 210
Schomburgk, R., 9, 16
Schwartz, O., 269
Sharp, L. W., 187, 188, 201, 210
Sherreff, pi. 32
Shocotyoff, 162
Silberschmidt, K., 206, 211
Simms, G. E.. 269
Simpson, pi. 33
Sims, J., 52, 80, 178, 180, 211
Skutch, A. F., 270
Small, J., 120, 137, 140, 168
Smith, 18, 39
Smith, C. M., 177. 181, 195, 196,

211, pi. 17
Smith, W. \V., 9
Sorensen, 74
Solander, D., 178
Solereder, H., 99. 105, 118, 168,

182, 212
SommerstorEf, H., 171, 172, 173,

176
Sparrow, F. K., 173, 176
Spoehr, H. A., 119, 168
Starke, A., 80
Stahl, E., 163, 164, 167, 168

Index of

— 352

Personal Names

Stapf, O., 226, 232, 233, 260, 270

Steckbeck, D. W., 19, 39

Stedman, H., 98

Stein, B., 201, 212

Stephens, E. L., 116, 168, 195, 213,

254, 270
Stern, 52, 53, 55, 56, 61, 62, 65, 68,

71, 74. 75, 80, 206, 212, pi. 7
Storer, T. I., 6
Stutzer, M. J., 160, 168, 265, 270

9, 10, II, 12, 13,

Tait, W. C, 99
Tate, G. H. H.,

pi. I

Tate, L,, 70, 8c, 88, 89, 159, 168
Taylor, pi. 32
Temminck, C. J., 80
Teruuchi, 71, 79
Thienemann, A., 77, 78, 80
Thompson, G. M., 270
Thomson, R. B., 132
Thomson, T., 195. 212
Tieghem, Th. van, 234, 270, pi. 21
Tischutkin, N., 71, 72, 80, no, in,

112, 114, 159, 168
Topp, C. A., 270
Torrey, J., 40, 42, 43, 3°
Tournefort, 17, 18
Tradescant, J., 18
Treat, M., 168, 234, 235, 264, 270
Trecul, A., 65, 80, 120, 121, 122,

136, 138, 168
Treviranus, C. L., 64, 80, 138, 16S,

234, 270

Troili-Petersson, 113

Troll, W., II, 16, 19, 20, 30, 31, 32,
38, 39, 41, 43, 46, 47, 50, 51, 52,
53, 56, 59, 60, 61, 62, 63, 79, 80,
81, 83, 89, 168, 196, 212

Turner, W., 4

Tutin. T. G., 94

Ule, E., 213, 270
linger, F., 65

Uphof, J. C. Th., 17, 23, 39, 270,
pi. I

Valkonov, A., 176

Vickery, J. W., 130, 132, 133, 134,

168, pi. 16
Vines, S. H., 70, 71, 72, 73, 80
Voelker, A., 69, 80
Vogl, A., 19, 39, so
Vouk, v., 80
Vries, H. de, 146, 147, 148, 149, 151,

153, 168

Walcott, M. v., 38, 39

Wallace, A. R., 72, 76

Warming, E., 90, 94, 120, 122, 168,

216, 217, 218, 228, 230, 270
Weber, Fr., 137, 165, 168
Wehrle, E., 242. 270
Wells, B. R., 268
Went, F. W., 224, 252
Weyland, H., 164, 168
Whately, 135, 136. 168
Wherry, E. T., 17, 23, 35, 39

White, J., IS9, 168

Wilkes, 40

Will, H., 69, 70, 79, 80, no, 114,

159, 168, 194, 211
von Wilier, 114
Wilson, V\'. P., 26, 39
Wilstatter, 119
Winkler, H., 132, 135, 168
Winzor, F. L., 131, 165, 167, 168
Withering. 136
Withycombe, C. L., 237, 238, 240,

241, 242, 243, 270
Wolf, F. A., 30
WoUny, E., 116. 168
Woolls, W., 83, 89
Woronin, 169
Wright, C, 90, 9+, 227
Wunschmann, E., 51, 59, 66, 79,

80
Wylie, R. B., 216, 270

YocoM, A. E., 216, 270
Young, R. G. N., 256, pi. 21
Young, W., 178, 179, 212

Zacharias, E., 80

Zeeuw, J. de, 66, 68, 75, 76, 80, 161

Ziegenspeck, H., 127, 157, 167, 188,

189, 192, 212
Zimmermann, A., 66, 80
Zipperer, P., 9, 16, 19, 22, 28, 32,

a, 34, 39
Zopf, W., 169, 171, 17s, 176