University of Iowa

Iowa Research Online

Theses and Dissertations

1914

Morphology of cannabis sativa L

Joyce Reed
State University of Iowa

This work has been identified with a Creative Commons Public Domain Mark 1.0. Material in the
public domain. No restrictions on use.

This thesis is available at Iowa Research Online: https://ir.uiowa.edu/etd/3895

Recommended Citation
Reed, Joyce. "Morphology of cannabis sativa L." MS (Master of Science) thesis, State University of Iowa, 1914.
https://doi.org/10.17077/etd.jn36uz6v.

Follow this and additional works at: https://ir.uiowa.edu/etd

 DEPARTMENT O F BOTANY.

THE STATE UNIVERSITY OF IOWA.

MORPHOLOGY OP CANNABIS SATIVA L.

by

JOYCE REED

THESIS

SUBMITTED IN PARTIAL FULFILLMENT OF R E Q U I R E M E N T S

FOR THE DEGREE OF MASTER OF SCIENCE.

JUNE 1914.
 INDEX.

Introduction Page

Object of study 1

Methods 2

Obligations 2

The Archichlamydeae 3

Characters of Urticales 3

Characters of Cannabis sativa L 4

Historical resume 6

The Inflorescences

Flowering season 10

Staminate inflorescence 10

Pistillate inflorescence 11

Floral organs and their Development

Staminate flower 13

Pistillate flower 13

Perianth of pistillate flower 13

Floral bract of pistillate flower 16

Nature of Ovary and O r i g i n of O v u le 18

Megasporangium and Female Gametophyte

Ovule 22

Megasporangium 23

Female gametophyte i 25

Fertilization 26
The Microsporangium

Development of microsporangium 27

Parietal layers 28

Tapetum 28

Development of Microspore

Heterotypic division 32

Homoeotypic division 37

Male gametophyte 39

Dehiscence and Pollination 41

Sterilization of Pollen

Table showing percent of sterilization 47

Embryo and fruit 48

Discussion

Relative rank of species 49

Diclinism 51

Sterilization of pollen 59

Summary 65

Bibliography 68

Explanation of Plates \ 77

Abbreviations used 78

Plates

Appendix
 INTRODUCTION.

Object of Study.

Cannabis sative L.,the common hemp,has long been of interest

because of its c om m e r c ia l , a g ri c u ltural,chemical,and physiological i m ­

portance and because of its dioecious character. The flower,fruit,and

seed of Cannabis have been discussed by many writers. The dubiety r e ­

garding its floral structure,as to the correlation of the floral parts

of the two sexes,and the nature of the organs that compose them,has

been dealt with in an extensive literature. Since the plant is a strict

ly dioecious form the problem of its sexuality is of great importance

and has been considered by numerous experimenters on diclinism. The

nature,origin,ratio and significance of sex are questions which are

being much investigated and discussed. Research on the subject of d i ­

clinism is going more and more into the inner morphology of plants in

an attempt to solve the important question of sex determination and

related problems of heredity. The critical morpho l o g y of the floral

structures of Cannabis sativa L . ,especially that of the staminate

flower,has been but briefly touched upon. Hence at the suggestion of

Dr. R. B. Wylie this study was undertaken in the hope of contributing

somewhat to the knowledge of the morphology of the plant,and perhaps

aid in solving the great question of s'exuality. Since the species is

included in the Urticales,one of the primitive orders of the Archi-

chlaraydeae,a brief discussion of the characters of this great group

will be entered upon.

 2

Methods.

The material for the study was collected at different times

in and around Iowa City,Iowa during the summers of 1912 and 1913*

Most of the material was killed in one per cent chromo-acetic acid

which proved very satisfactory for all stages of development. After

washing,the material was run up to seventy per cent alcohol from which

portions of each collection were embedded in paraffin. Microtome se c ­

tions were cut 5-18 micrones in thickness. The stains used were Dela-

field's haematoxylin,Flemming's triple stain and Haidenhain's iron-

alum-haematoxylin. Delafield's haematoxylin was used to advantage'in

the study of the younger stages. Flemming's triple stain and Ha i d e n ­

hain's iron-alum-hasmatoxylin proved to be better for the older stages

and the latter was especially good in the study of the nucleus.

Obligations.

The work was carried on under the supervision of Dr. R. B.

Wylie to whom I am greatly indebted for kindly encouragement,and

helpful suggestions and criticisms. My thanks are also due to P ro­

fessor B. Shimek for aid in securing reference books,and to Mr. C. H.

Farr for his interest and numerous courtesies.

 3

The Archichlamydeae.

The Archichlamydeae or Choripetalae,the more primitive of the

two great divisions of Dicotyledones are distinguished in the main

from the Sympetalae by apetaly or polypetaly. The group is vast and

widespread including olQOO species and 180 families scattered in all

regions. The most primitive of Dicotyledones belong hers but the c l a s s ­

ification of the group is confusing and indefinite because of the

fluctuating characters and the varying opinions concerning the details

of classification. The general sequence of the divisions of the group

is based upon the development of the perianth and floral axis and the

arrangement of the floral members. However there can be no real sequence

among the divisions of the Archichlamydeae,for they represent,mainly,

parallel or divergent lines of development. The more p rimitive orders

of the group are especially puzzling since they exhibit a rather h e t e r o ­

genous assemblage of forms. This region has recently received much a t ­

tention from morphologists but still greatly needs investigation in

order that the various forms may be recombined into natural groups.

Characters of the Urticales.

The Urticales or Urticineae are among the more primitive of the

Arehichlaymdeae,being the ninth of the orders or alliances arranged by

Sngler (24) and given the same rank by Gray (30)- The order contains

about lp60 species and is largely tropical although also represented in

our native flora. The Urticales include both herbaceous and woody plants

with variously shaped always stipulate leaves and s m a l l ,inconspicuous

flowers closely aggregated into thick inflorescences. The flowers are

monoecious,dioecious or polygamous and are cyclic in arrangement of

 4

parts with a double perianth of similar members,or,seldom naked. The

stamens are opposite the perianth segments and usually of the same

number. The ovary is superior,composed of one or two carpels,usually

unilocular and contains a single ovule with two integuments. The fruit

is a nut or drupe and the seeds usually contain endosperm. The order

Urticales has been variously divided into families and subfamilies by

different authors. Goebel (28) divides the group into the families

Urticaceae,Ulmaceae, Platanacsae and Cannabinsae. Engler (24) and Britton

(10) include in the order the families Ulmaceae,Moraceae and Urticaceae.

According to Gray (30) the order comprises but one family,the Urticaceae

which is subdivided into five tribes the Ulmeae,Celtideaa,Cannabinsae,

Moreae and Urticiae.

Characters of Cannabis sativa L.

Cannabis sativa L.,the species under consideration,is the sole

representative of its genus in our flora. Gray (30) has included C a n ­

nabis and Humulus in the tribe Cannabineae of the family Urticaceae.

Britton (10) has classified it in the family Moraceae separating it

from the Urticaceae chiefly because of the character of its filaments

which are straight in the bud instead of reflexed and of the ovule

which is suspended instead of erect.

Cannabis sativa L. is a stout erect,annual,branching herb v a r y ­

ing in size with the climate and soil in which it grows. In temperate

regions it varies from 3to 10 feet in height while in warmer climates

it may attain the height of 15 to 20 feet. The leaves are digitate 5-

11 palmately divided,alternate above and opposite-petioled below and

beset with stiff hairs. Persistent subulate stipules are present. The
 5

flowers are dioecious,small,greenish and axillary. The staminate f l o w ­

ers composed of a five segmented perianth and five drooping stamens

are panicled. The pistillate flowers consist of a gamophyllous incon­

spicuous perianth closely embracing an ovoid ovary and are s pi k e-clust­

ered. The inner bark of the stem is fibrous furnishing the hemp fiber

of the market. The plant is typically dioecious the pistillate plant

being larger and more robust than the staminate and having a darker

and more luxuriant foliage when mature. The two plants are scarcely

differentiated before the flowering season. However when cultivated

the staminate plants are recognizable at an early stage by certain

minor characters which are useful to special experts in weeding them

out. The staminate plants do not continue growth after flowering but

turn yellow and soon die. The pistillate plants,however,continue their

growth and become larger and stronger remaining green until killed by

the frost.

This species is found widely distributed in all temperate r e ­

gions of North America. It also inhabits cooler parts of India where

it is native and whence it was transplanted to Europe and to America.

 6

Historical Resume.

As before mentioned Cannabis Sativa L. has long been of i n t e r ­

est to botanists and appears frequently in the literature. Most of the

work centers around the structures of the pistillate flower and the

nature of the dioecious character of the species. Since the recent p a ­

pers on the subject are the more critical and also more available for

study these only are taken into consideration. Frequent reference to

the older works is made in the later publications thus extending the

scope of the summary. The details of the literature bearing upon p a r ­

ticular parts are taken up in connection with the several structures

in the following pages and hence only a brief synopsis of the field

covered will be made at this point.

The most ponderous and complete work on the subject of Cannabis

was published by two Italian botanists,Briosi and Tognini (9},in 1894

and 1896. This work which appeared in two parts embraces a detailed

study of the internal anatomy of the plant and is illustrated by n u m ­

erous photographs and plates. Part I,published in 1894,deals with the

flowers,- the inflorescence,organography,organogeny,megasporangium,

female gametophyte,microsporangium and male gametophyte. A discussion

of the diclinism of the species, is also included here. Part II,which

appeared two years later,is a study of the internal anatomy of the

vegetative organs,- co t y l e do n e s , le a v e s,stipules,stem,and root.

In 1898 a contribution to the morphology of the pistillate

flower and inflorescence of the Canna'oineae was made by Zinger (69).

Humulus japonicus and Cannabis sativa were studied comparatively

as regards the nature and development of the inflorescence and floral

 7

organs of the pistillate flower. The author c onstantly refers to the

work of various writers on the subject and is inclined to agree with

the views of the older morphologists rather than those of the more

recent authors.
Montemartini (44) in 1902 published a work upon the morphology

of the ovary and ovule of Cannabis. The vascular supply of these parts

was studied with a view to interpreting the origin of the parts. The

conclusions agree in the main with those of Briosi and Tognini (9)

on this point.
A treatise on the m o r phology,teratology,and diolinism of the

flowers of the Indian-grown Cannabis by Major D. Prain (pO) was p u b ­

lished in 1904. The description of the normal flowers is followed

by accounts of abnormalities observed by the author. A full discussion

of the nature of the ovary and origin of the ovule is taken up,the

conclusions being supported by teratological phenomena observed. The

subject of diclinism is discussed briefly with reference to the a b ­

normalities and also to the work of older writers on the subject.

The work of Modile w s k y (43) which appeared in 1908 is a study

of the embryo development of several genera of the Urtieaceae. Ela-

tostema sessile,Dorstenia drakeana,Urtica d i o i c a ,Dorstenia contray-

erva.Urtica cannabina,Urtica pilulifera,Pilea grandis,Pilea nummal-

ariaefolia,Boehmeria platyphylla,Celtis occidentalis,Cannabis sativa

and Humulus japonieus are the species which were studied and compared.

The female gametophyte,fertilization and development of the embryo

are considered in the different species and' general conclusions are

drawn which apply to ^ e group as a whole.

 8

Since Cannabis sativa L. is dioecious,easily cultivated,and

very wide in its distribution it has been the subject of much inves­

tigation in relation to questions of sex. Problems of sex differences,

sex ratio,sex constancy,sex determination and sex inheritance have led

to numerous experiments to which more detailed reference will be made

in the discussion of dielinism. Some of the more important of the in­

vestigations were carried on by Haberlandt (23 ) ,Saccardo (53),Heyer

(34),Dusing (21),Hoffman (35),Fisch (25),Noll (46),Strasburger (62),

and Sprecher (57)*

A paper published by Noll (46) in 1907 dealing with the d e t e r ­

mination of sex in dioecious plants records many experiments chiefly

upon hemp plants. Isolated pistillate individuals were pollinated in

various ways in an effort to determine the time of sex differentiation

and the cause of sex proportion. Although Noll's interpretations of

his results are questioned by later authors on the subject,the e x ­

periments are carefully recorded and have contributed largely to the

knowledge of the subject.

Strasburger (62) in 1910 made a contribution concerning the

cause of sex determination. Refined experiments in an attempt to show

the differentiation of sex in the pollen tetrad resulted negatively.

However important results were obtained from investigations upon the

sex proportion by cross-pollinating and cutting back plants of Mer-

curialies annua,Melandryum rubrum and ether dioecious forms. A cy-

tological study of the pollen mother cells of various dioecious plants

including Cannabis sativa L. was made in the hopes of finding here an

analogy to the heterochromosomes of the animal kingdom. The results
 9

seem to prove conclusively that no such apparent differentiation is

to be found in the chromosomes of these plants.

A study of the v a ri ability of sex in Rumex acetosa L. and

Cannabis sativa L. was published in 1913 by Andreas Sprecher (57)*

A splendid resume of the ground covered by earlier investigators is

followed by a discussion of various phases of the question of sexuality.

The author's contribution consists of experimental work upon the d i f ­

ference in variation which occurs in the staminate and pistillate plant

of the same species when grown under different conditions. Variations in

length,weight and.osmotic pressure of the two sexes are recorded a c ­

cording to a statistical method.

 10

THE INFLORESCENCES.

Flowering Season.

The flowering season of Cannabis sativa L. varies according

to the climate and the season. In Iowa it is usually from June to

September. The staminate. and pistillate plants differ somewhat in

respect to their time of flowering,the former blooming slightly

earlier than the latter. After the staminate plants have finished

blooming they turn yellow and die while the p i s t illate plants in­

crease in... size and vigor until late in September and remain green

until killed by the frost.

Since Cannabis sativa L. is normally strictly dioecious,the

flowers of the two sexes will be considered separately. As is usual in

anemophilous forms,the flowers are small and inconspicuous displaying

rather primitive characters in the single perianth and hypcgyny. An

essential d i f f e r e n c e .i n .structure.characterizes.the two f l o w e r s - a

condition uncommon among Angiosperms where hermaphroditism predominates.

Staminate Inflorescence.

The staminate inflorescence has been studied by Viylder (67),

Eichler (23),8riosi and Tognini (9) and Fraxn (50) but the accounts of

the authors do not differ materially.

The individual staminate flowers are short-stalked and drooping

arranged in small determinate inflorescences or paniculate cymes. The

cymes are located in pairs in the axils of stipules,usually on special

floral branches but also found at the bases of these branches. The
 11

pairs of cymes both on and at the base of a floral branch show an i n ­

equality in size - one being larger than the other. Prain (50) mentions

this fact in his description of the staminate inflorescence of the

Indian form. The cymes on the floral branches and also the leaves in

whose* axils they are found become reduced and more crowded toward the

apex of the branch. No leafy branch appears between the pair of flower

clusters but Briosi and Tognini (9) have found a third panicle here

which is usually smaller than the lateral o nes,often being reduced to

a single flower. Prain (50) in summing up the discussion of the staminate

inflorescence states that "though simulating a panicle,the male i n f lo ­

rescence is not in reality a spiral of racemes but a spiral succession

of cyme systems in which the axis is defined and gives out two branches,

the lower smaller than the higher,which latter repeats the process".

Pistillate Inflorescence.

The pistillate inflorescence like that of the staminate plant

is subtended by a leaf with two free stipules. In the axil of each

stipule is a single, solitary pistillate flower enclosed in a leaf-like

bract. The branch arising in the axil of the stipulate leaf between the

two basal flowers forms a spiked inflorescence bearing an indefinite

series of pairs of flowers which are located like the basal pair in the ■

axils of the stipules. The inflorescence becomes very compact toward

the apex,the leaves and stipules being much reduced. Toward the base of

the stem the spikelets are often replaced by leafy branches.

The pistillate inflorescence has been studied by Wylder (67),

B. Clarke (12),Sichler (25), Briosi and Tognini (9),'Zinger (69),M o n t -

emartini (44) and others. The opinions of the various writers concerning
 12

its true nature vary considerably. Wylder (67) considers the pistillate

inflorescence as homologous with the lateral inflorescence of the s t a m ­

inate plant. Eiehler (23) ranks the pistillate flower in its bract an

equivalent of the staminate inflorescence which he terms a panicle.

Zinger (69) opposing the view of a definite inflorescence described by

Briosi and Tognini (9) states that the pistillate flowers form no i n ­

florescence but are scattered in the axils of leaves on twigs of several

orders. Montemartini (44) agrees with Zinger (69) and shows that the

pistillate f lowers are developed two by two in the axils of leaves r e p ­

resenting the .first small branchlets of the secondary axillary branch

which develops between them.

The attempts of ftylder (67) to correlate the pistillate i n f l o ­

rescence with the staminate inflorescence and of Eiehler (23) to c o r r e l ­

ate the pistillate flower with the staminate inflorescence seem useless

since the two plants are very different. The view of a definite i n f l o ­

rescence set forth by Briosi and Tognini (9) appears to be offest by

the presence of the flower pair at the base of the spike-like floral

branch. The best explanation,then,of the arrangement of the pistillate

flowers is as Zinger (69)and Montemartini (44) have shown that they are

in pairs in the axils of the leaves of twigs of several orders.

 13

FLORAL ORGANS AND THEIR DEVELOPMENT.

Staminate Flower.

.The individual staminate flower consists of five free perianth

parts opposite each of which is a stamen (figs.1 0 , 11). The sepals are

imbricated,oblong-lanceolate,pale-green,herbaceous segments. The stamens

consist of a terminal,pendulous,oblong anther and a slender filament

shorter than the anther (fig.10). The normal staminate flower shows no

trace of a rudimentary pistil (fig.11).

The first indication of the staminate flower is a rounded

elevation of meristematic tissue (fig.l). The floral organs a r i s e ’

in acropetal succession - sepals and stamens. 'The primordia of the

sepals appear first as little papillae cn the periphery of the r e c e p ­

tacle (fig.2). These elongate curving inward and becoming quite d i s ­

tinct before the primordia of the stamens appear. The latter arise

within and opposite the parts of the perianth appearing first as small

elevations (figs.3,4,5 )* The stamens elongate rapidly (fig.6),the tip

becoming enlarged to form the anther (fig.8) and the filament elongating

at the mother cell stage of the microsporangium. The stamens are straight

in the bud and are roofed over by the imbricated sepals (fig.10).

Pistillate Flower.

The pistillate flower proper,enclosed within the involving l e a f ­

like bract, is composed of a t h i n , membraneous,hyaline,perianth whose m a r ­

gin is slightly higher behind than in f r o n t ,clos e l y embracing but free

from the ovoid unilocular ovary (fig.24). Crowning the ovary are two

filiform deciduous styles of which the posterior internal is slightly

 14

larger than the external. The stigmae which occur around the periphery

of the apex and along the opposing faces of the styles,are brush-like

having their epidermal cells elongated into hair-like projections.

Within the ovary is a solitary uncinate ovule which is pendulous'from

the apex and attached by a short funiculus. In the normal pistillate

flower there is no trace of an androeeium within the perianth (fig.24).

The pistillate flower appears in its earliest stage as a r o u n d ­

ed mass of meristematic tissue in the axil of the young floral bract

(figs.12,13). The floral members develop in acropetal succession -

perianth and carpels. The perianth appears first as two small prim-

ordia which soon lose their identity and develop "en masse". The inner

or posterior primordium opposite the involving bract is the first to

appear (fig.14) and is followed quickly by the outer anterior prim-

ordium (fig.15). The axial growth of these two points is early checked

and the whole zone at. the periphery of the torus shares the growth r e ­

sulting in an annular structure which is somewhat higher behind than

in front. Within the perianth the carpels appear (fig.16) one posterior

and internal,the other external. These soon fuse laterally forming a

cup-like structure. The ovule appears slightly lateral to the organic

apex (fig.1 8 ) . It lies in the axil of the posterior carpel uniting

with it (figs.19,20) so that as the posterior wall of the ovary grows

upward the ovule is carried with it,reaching the top of the ovarian

chamber before it is closed (figs. 21,22,23). 'The two styles appear at

points representing the margins of the fused carpels (fig.21).These

become united at their bases thus closing the ovarian chamber (fig.24).
 15

Perianth of Pistillate Flower.

The inconspicuous perianth of the pistillate flower was un­

known to earlier writers who recognized the involving bract as the

perianth. Payer (49) and at about the same time B. Clarke (12) first

discovered the true perianth and it has since been recognized in des.-

criptions of Cannabis. The.opinions of different authors vary s o m e ­

what in regard to its nature and development. Bentham ana Hooker (5)

and others report that the perianth is weakly developed and is missing

at times. However Prain (50),who examined many pistillate flowers of the

Indian hemp asserts that it is always present in the normal flower e x ­

cept in solitary pistillate flowers at the apices of spikelets,where

the involving bract is also absent. Payer (49) and Zinger (69) d e s ­

cribe the development of the perianth from two leaves or leaflets

which are formed independently of each other. Briosi and Tognini (9),

however,show a different development for the flowers of Indian plants

whose perianth they describe as arising from single ridge resulting

in an almost perfectly horizontal perianth.

The present study of the floral envelope of the American Cannabis

has shown a condition similar to that described by Payer (49) and Zinger

(69) in which the perianth is d e v e l o p e d .from two primordia and is slightly

higher behind than in front (figs. 1 4 - 2 4 )(. Zinger (69) by using a c l e a r ­

ing preparation was able to study the development of the floral o r ­

gans from the exterior,and consequently could more readily observe the

two beginnings of the perianth than Briosi and Tognini (9) who studied

simply sections of the young flower. My observations of many serial

sections of the young flower and also of the external appearance of

the mature flower has led to the conclusion that the perianth agrees
 16

with that described by Payer (49) and Zinger (69) rather than that

shown by Briosi and T o g n i n i (9).

Floral Bract of Pistillate Flower.

The nature of the involving leaf-like bract of the pi s t i l ­

late flower has been the subject of much discussion. It varies in

shape being in the unclutivated Indian plants and the cultivated

European plants usually acute or acuminate while in the cultivated

Indian forms it is g e nerally truncate. The earlier writers to whom

the true perianth was unknown considered this bract as the perianth

of the flower,but later authors have shown it to be a leaf-like

floral bract. Schlaiden (54) was the first to consider it as a bract

and not the perianth. Irmisch and also Payer (49) pronounced it a

bract in whose axil the flower arises. Briosi and Tognini (9) showing

that the flower and bract are closely associated have termed the

structure "brattea perigoniale". Zinger (69) disagreeing with the

view of Briosi and Tognini (9) regarded the bract as a leaf of the

secondary branch in whose axil the flower is formed. Montsmartini

(44) has termed it a leaf of a small floral branch n o t ,however,in

relation with the secondary branch as Zinger (69) believed.

The exact nature of the involving bract is difficult to d e ­

termine since there are numerous gradations in the plant kingdom b e ­

tween foliage and floral leaves. Briosi and Tognini (9) have shown

that a close relationship exists between the bract and the flower

both in their origin and in their vascular supply. The bract and the

floral organs develop from a common prirnordium the bract becoming

\
differentiated first (f i g s . 12,15). Also the vascular trace which in-
 17

nerves the bract is traced from the pedicel of the flower. Hence,

although the bract cannot rightfully be termed a part of the flower

proper.it is as Briosi and Tognini (9) have stated,in very close

relationship with it and perhaps their term "brattea perigoniale" is

aptly applied.
 18

NATURE OF OVARY AND ORIGIN OF OVULE.

The peculiar nature of the gynoecium of. Cannabis has made it

the subject of discussion by many writers. The nature of the ovary

whether formed by one carpellary leaf or two and if two which leaf is

fertile,has been speculated upon extensively. The origin and nature

of the ovule has been equally disputed whether it is of epicarpellary,

hypocarpe^lary or axial origin. Since this subject has been discussed

so extensively in the literature an attempt is made here to bring t o ­

gether the principal theories of the various writers with the facts

upon which the interpretations are based.

B. Clarke (12),probably the first writer on the nature of the

ovary and ovule of Cannabis attempts to interpret the structure m e r e ­

ly from dissections of the flowers. He states that the ovary is mono-

earpellary and that the posterior thickening represents the placenta

formed by the uniting of the margins of the leaf-like carpel. The

ovule is of foliar origin and located at the upper end of the p l a ­

centa. No explanation is given for the presence of two styles.

The next attempt was made by Payer (49) who explains the

nature of the ovary and ovule of Cannabis on organogenic grounds.

According to his theory the ovary begins as two carples which are at

first distinct but afterward unite at the base to form the styles. The

anterior carpel becomes enlarged and forms the ovary while the p os t ­

erior carpel remains rudimentary. The ovule is cauline arising from

the posterior side of the stem apex,according to Payer (49) and is

carried to the top of the ovary by intercalary growth. This theory

 19

.explains the origin of the posterior as well as the anterior style

but does not offer any explanation for the larger size of the post­

erior style and the posterior thickening of the ovarian wall.

Doell (20) advances the monocarpellary theory which is also

held by Clarke (12). However he further explains the second style as

an excrescence on the ventral suture of the carpel.

Celakowsky (11) bases his theory on the views of Payer (49)

describing the development of the ovary from two carpels of which the

inner is abortive and the anterior is d e v e l o p e d . .The ovule is e x p l a i n ­

ed as foliar the growing point being displaced by the greater d e v ­

elopment of the anterior carpel.

Briosi and Tognini (9) made a histological study of the ovary

and ovule and describe the vascular traces of the base of the ovary.

Four vascular traces are found,one anterior,one posterior,and two

lateral. The anterior smallest trace enters the involving bract,the

external lateral trace the anterior side of the ovary,the internal

lateral trace branches to innerve a greater part of the carpellary

wall and the posterior trace branches and passes into a small swelling

opposite the base of the posterior earpelldry wall. This latter trace

innerves the ovule. These data overthrow the views of B. Clarke (12),
/
Payer (49) and C elakowsky (11). However Briosi and Tognini (9) do not

interpret these facts as solving the question as to whether the ovary

is monocarpellary or b i carpellary or as to whether the ovule is

cauline or foliar,but are inclined' to regard the ovule as axial and

the carpel as representing an intermediate form,referable neither to

axis or leaf.
 20

Zinger (69) regards the pistil of Cannabis as formed by two

carpels of which the anterior only takes part in the formation of

the ovary while the posterior constitutes only the back style. The

ovule is formed by the apex of the flower axis and the lifting up of

the ovule is accomplished by the elongation of the internode which

separates the carpels.

Montemartini (44) describes the ovary as b i c a r p e l l a r y ’but c o n ­

trary to the views of Zinger ( 6 9 ) , Payer (49),and C elakowsky (27),

shows that the posterior carpel is fertile and the anterior a b o r ­

tive and sterile. The ovule is described as independent - a cauline

formation supported by the dorsal suture of the posterior carpel.

The author also traces the course of the vascular traces which c o r ­

respond in great part to that given by Briosi and Tognini (9).

Prain (5(9) who is the most recent author on the subject agrees

with Montemartini (44) concerning the nature of the ovary and origin

of the ovule of Cannabis , a l th o u g h apparently ignorant of his work.

He explains the ovule as originating lateral to the apex and in the

axil of the posterior carpel. The styles are developed on the mar­

gins of the fused carpels. He shows that the vascular traces d e s c r i b ­

ed by Briosi and Tognini (9) seem to substantiate this theory - the

two carpels being innerved by the two opposite lateral traces and the

primary axial trace ending at the base of the ovarian chamber at a

point corresponding to the position of the ovule. Prain's (50) d e s ­

cription of teratological phenomena shows that the pistil is c o m p o s ­

ed of two carpels of the leaf-tyle which are capable,under abnormal

 21

influences,of becoming leaves and that the ovule is truly axillary

to the posterior carpellary leaf.

In view of the fact that both an anatomical study of the

vascular system of the pistillate flower,and also certain tera-

tological phenomena are in harmony with this interpretation of the

nature of the ovary and ovule,the more recent explanation advanced

by Montemartini (44) and Prain (50) is the- most acceptable.

 22

MEGASPORANGIUM AND FEMALE GAMETOPHYTB.

The Ovule.

The ovule of Cannabis as described above,is cauline and

lateral. It originates as a papilla of meristematic tissue (fig.18)

which soon becomes assoicated with the posterior carpellary wall

(figs.19,20) and with its growth is carried upward to the top of the

ovarian chamber (figs.21,22,23). There is practically no funiculus,

but a slight thickening corresponding to the path of the ovule

along the posterior ovarian wall is distinguishable. The pendulous

position,then,is due,not to the growth of the funiculus,but to the

elevation of the hilum. In the mature flower the ovule fills the

ovarian cavity and is attached above slightly to the posterior side

of the cavity (fig.24).

The direction of the growth of the ovule results in a form

which is usually termed campylotropous. However,it has been v a r i o u s ­

ly described as anatropous because of its curved development,ortho-

tropous because of its terminal position,and campylotropous owing

to its uncinate form. Prain (50) maintains that the latter term is

not accurate since the form of the ovule is due,not to its curving,

but to the elevation of the hilum. He proposes the terra "obcampy-

lotropous". The orthotropous ovule is reported by Coulter and C h a m ­

berlain (17) as common for the Urticaceae and for the Arehichlamydeae

in general. The campylotropous type is rare,being found among Dic-

otyledones in the Chenopodiaceae,Resedaceae,Cruciferae,Capparidaceae,

 23

Caryophyllaceae,etc. - ,more or less specialized families of their

alliances.

After the nucel-lus has become prominent ana as a result of

the growth of the carpels has reached the top of the ovarian chamber

the integuments appear. The inner originates f i r s t ,appearing as an

annular outgrowth around the base of the nucellus (fig.22). Soon the

primordium of the outer integument becomes visible (fig.23). During

their growth the integuments become very massive and when the embryo

sac is formed are completely coalesced over the apex of the nucellus

(fig.35). As Zinger (69) reports for the C a n n a b i n e a e ,the integuments

of the Cannabis grow together at the posterior side of the ovary and

become fused over the tip of the nucellus so as to completely o b l i t ­

erate the micropyle.

.According to Modi lewsky (43) two integuments are common

among the Urticaceae. Shat tuck (56) also observed two and oc c a s i o n ­

ally three for Ulmus americana. Coulter and Chamberlain (17) state

that on the whole this condition is the rule among Monocotyledones

and also prevails among the Archichlamydeae. Since a single integ­

ument is characteristic of the Sympetalae and higher forms the

double integument can sefsly be regarded as a primitive structure.

The Megasporangium.

The material was not favorable for the study of the mega-

sporangium and female gametophyte. However,certain p e c u l i a r i t i e s

were very evident and at least deserve mention. Because of the

oampylotropous form of the ovule sections through the embryo-sac

are difficult to obtain.
 24

According to Briosi and Tognini (9) the archesporiun of

the megasporangiura originates as a single hypodermal cell which

as usual,divides perielinally resulting in the formation of a p r i m ­

ary parietal and primary sporogenous cell. The latter becoming the

megaspore mother cell divides to form a row of three megaspores of

which the inner one functions. They state that the embryo-sac is

formed as is usual among Angiosperms but givs no details in this

connect ion.

The present investigation shows a variation from the,mode of

development described by the Italian botanists. A several celled

archesporium was distinctly observed in a number of cases. The e a r l ­

ier development of the megasporangium,noted but rarely,showed only

one primary arehesporial cell (fig.26). Whether or not this s e v e r a l-

celled sporangium develops from a one-celled primary archesporium

is a auestior.. Treub (65)for Casuarina, Miss Benson (4) for Carpinus,

Conrad for Gijprcus report many-celled archesporia,which according

to Coulter and Chamberlain (17) are certainly not all hypodermal in

origin. Treub (63) reports the archesporium of Casuarina as a group

of hypodermal cells but his account and figures suggest that all of

the sporogenous tissue may not be derived from the hypodermal layer.

The many-celled archesporium is the prevailing tendency

among the Rosaceae and is also common in the Amentiferae. The Ranun-

culaceae show a great irrgedlarity in the number of arehesporial

cells with a tendency to increase their number. Since the Rosaceae,

Amentiferae and Ranunculaceae are recognized as primitive members of

the Archichlamydeae the many-celled archesporium would seem to be

 25

Characteristic of the more primitive Dicotyledones. However,Coulter

and Chamberlain (17) state that this is true only in a general sense,

for no large groups have this general tendency and the same feature

has been reported for members of higher groups.

The divisions of the primary parietal cell result in the f o r ­

mation of 6-8 rows of cells placing the sporogenous cells deep within

the sporangium (figs.26-35)* A conspicuous development of parietal

tissue is noted by Coulter and Chamberlain (17) for certain of the

Monoeotyledones and Archichlamydeae. Since the suppression of the

parietal tissue of the megasporangium is universal in the Sympetalae

as far as they have been investigated a large development seems to

indicate a primitive condition.

The Female Gametophyte.

fiith the growth of the megasporangium the sporogenous cells

increas e in size and take oh the appearance of megaspore mother cells.

The nucleus of each cell shows a distinct spirem on which prominent

chromatin masses appear (fig.2$). This is evidently a presynaptic

condition of the nucleus. A later stage shows one mother cell becoming

dominant at the expense of the others. This one.enlarges rapidly p u s h ­

ing the others aside (fig.30,31)* Th$ latter cells appear shrunken

and deeply stained showing degeneration. A later stage of d e v e l o p ­

ment shows only the dominant mother cell remaining,the surrounding

cells having disappeared (fig.32).

The further development of the female gametophyte was not

observed. It was not determined whether the mother cell divides to

 26

form megaspores or whether it functions directly as a megaspore.

Four-celled (fig.34) and mature embryo-sacs (fig.35) seem to indicate

that the development is as usual in Angiosperms. Modilewsky (43)

reports that the antipodals are weakly developed and that certain

large nourishing cells are present in the antipodal end of the sac.

The embryo-sac is horse-shoe shaped narrowing toward the micropylar

end. The nueellus also narrows at this extremity and becomes d i s t i n c t ­

ly beaked (fig.33)*The integuments as noted before become coalesced

over this beak closing up the micropyle.

Fertilization.

No traces of pollen tubes were found and hence it was not

determined whether or not fertilization occurs here. Briosi and T o g ­

nini (9) failed to find pollen tubes but figure the development of

the embryo. Zinger in 1898 observed the p o l lination of Cannabis and

traced the course of the pollen tubes from the stigma to the embryo-

sac. According to him the tube grows directly downward through the

ovary and perpendicular to the integuments until it reaches the n u ­

ceilus. The tube branches freely and forms blind sacs. However the

phenomenon of fertilization was not observed. Winkler (65) records

that earlier experimenters on hemp regarded it as parthenogenetic

Kirchner (41) states that seed formation in hemp is partly d e p e n d ­

ent upon parthenogenesis and suggests that some races of hemp may be

parthenogenetic while others demand fertilization.

 27

THE MICROSPORANGIUM.

Development of Microsporangium.

The androeeium of Cannabis sativa L. consists of five,free,

o-
shortly stalked stamens zrranged in a whorl opposite and inside the

perianth parts (f i g s . 1 0,11). As before noted the stamens arise late r ­

al to the floral apex and hence may be regarded as foliar structures.

Each stamen consists of an oblong anther somewhat longer than the

slender filament at whose tip it is attached. The anther produces

four microsporangia which later at the time of dehiscence form two

loculi by breaking together of t.wo sporangia. A notable feature

connected with the development of pollen from the spore mother cells,

is the sterilization of potentially sporogenous tissue which occurs

at several stages.

The development of the microsporangium agrees with that d e s ­

cribed by Coulter and Chamberlain (17) for most Angiosperms. The

anther in cross-section is at first a round mass of meristematic

cells surrounded by an epidermis (fig.4). At an early stage the mass

becomes four-lobed (fig.7) and an archesporium is differentiated

in each of the lobes (fig.36). By periclinal divisions the a r che­

sporium gives rise to the parietal and sporogenous tissue (figs.

37,36)* fhe divisions of the parietal tissue result in four layers

arranged concentrically around and completely investing the s porog­

enous7mass (figs.38,39). Of these layers the inner next to the

 28

sporogenous tissue becomes the tapetal or nourishing layer,the two

middle layers disorganize during the divisions of the pollen mother

cell,and the outer layer or endothecium becomes specialized for the

dehiscence of the anther. During the divisions of the parietal layers

the sporogenous tissue divides rapidly in all directions giving rise

to a cylindrical mass of spore mother cells which is about 56 cells

in length ana 10-12 cells in diameter. The total number of sporog-

enous cells in one sporangium is,then,about 5 6 5 5 , and for one stamen

22620.

The Parietal Layers.

The "middle layers",the two layers of the wall between the

endothecium and the tapetum,consist of tabular cells (figs.39,40)

as Shattuck (56) observed in the species Umlus americana L. During

the division of the pollen mother cells these layers become f l a t ­

tened and disorganize. The inner one breaks down first (fig.41)

by a slow process becoming thinner and thinner and finally d is a p ­

pearing at the early tetrad stage of the mother cell*. The outer

parietal layer persists until later in the tetrad stage (fig.42)

breaking down shortly before the tapetal cells disorganize.

The Tapetum.

The tapetum or "nourishing jacket"is a regular layer in c o n ­

tact with and completely investing the sporogenous tissue. The layer

is apparently derived from the parietal tissue as is usual in the

typical Angiosperm (fig.33).

 29

At the mother cell stage of the sporogenous tissue the

tapetum is well organized (fig.39) - the cells appearing blocky

and uninucleate with very granular contents (fig.44). During the

early presynaptic stage stages of the mother cells the nuclei of

the tapetal cells divide mitctically and increase in size (fig.45).

The first nuclear division is regular and results in a uniform

binucleate condition which persists for some time. The later d i v ­

isions are more scattered. The nuclei present often peculiar lobed

and irregular forms due to processes of nuclear fusion and subseq­

uent divisions. Their form may be elongate two nuclei lying side by-

side until fused into a large elongated body,or,two or three s p h e r i ­

cal nuclei may unite resulting in an irregular form (fig.4$). Gates

(26) working on Oenothera lata and Beer (2) on Oenothera longiflora

and Oenothera biennis state that the later divisions of the tapetal

cells are chiefly by amitcsis. However,the recent work of Bonnet

(31) on the tapetal cells shows that the irregular appearance of the

nuclei is due entirely to nuclear fusions and not amitotic divisions.

The author states "Les phenomenes d'amitose ne paraissent pas exister

dans les cellules nourricieres.- Toutes les apparences q u ’on leur

on attribuees s'expliquent par des irregularites mitotiques et

des fusions nucleaires".

The tapetal cells reach their maximum development during

the formation of the tetrads at the time of the disorganization

of the inner of the middle layers. At this time the cells are very

large (figs.47,48,49) and may have 2,3,or 4 nuclei which contain

an abundance of darkly staining m a t er i a l ,presumably chromatin.

 30

In the later stages the nuclei fuse and the cells,shortly before

disorganization,contain one or two nuclei (figs.50,51 )• After the

microspores have separated from their tetrad arrangement the n ou r ­

ishing cells break apart,become vacuolated,take peculiar irregular

forms and stain darkly. Their nuclei become indistinguishable and

the chromatin masses appear scattered about in the cytoplasm of the

cell (figs.52,53)* Their breaking down is coincident with the f o r m ­

ation of the wall of the pollen grain.

At times the tapetal cells present the appearance of d i s ­

organizing at earlier stages than usual. The cells appear as they do

in the later stages being separated,irregular and darkly staining.

This condition is found at almost any stage in the development of

the pollen mother cell. Gates (26) observed a similar appearance

in the tapetal cells of Oenothera.Lamarckiana,and attributed the

appearance to the degeneration of the cells. However,in Cannabis

no m ie r o s p o r a n g i a were observed in which the tapetal cells had

entirely disappeared prematurely as Gates (26) reports in L a m a r c ­

kiana. Since the sterility in the sporogenous tissue of Cannabis

sativa L. is similar to that observed by Gates (26) i-n Oenothera

Lamarckiana,it is significant to note the apparent similarity in

the sterility of the tapetum.

 31

D EVELOPMENT OF MICROSPORE.

The general development of the microspo r a n g iu m was followed

in the preceding chapter with especial emphasis upon the formation

and ultimate fate of the wall tissue,including the tapetum. The p r e s ­

ent chapter deals particularly with the divisions of the pollen

mother cell and the formation of pollen.

Since Strasburger (59) in 1894 placed the alternation of

generations in plants on a chromosome basis and explained the s i g ­

nificance of the reduction divisions,much study has been given to

this subject in both the animal and plant kingdoms.The nature of the

chromosomes and the part they play in heredity have been discussed

in an extensive literature. As a result of the many investigations

the later stages of the reduction division from the end of p ro­

phase onward are generally agreed upon by most c y t o l o g i st s , e s pe c i a ­

lly those who have studied plants. The recent papers have been in

regard to the earlier stages from the resting condition of the

mother cell to synapsis.

Although extensive work has been done upon the structure

and morphology of Cannabis sativa L . ,the study of the reduction

division of the pollen mother cells has been taken up but brief­

ly. Briosi and Tognini (9) mention that the division of the pollen

mother cell is,as in all Dicotyledones,"a simultaneous division

into four parts". Strasburger (62),in a recent paper dealing with

the sex-determining factors of dioecious plants,reports the results

i
of a cytological study of several species including Cannabis sativa L.
 32

The observations were made in an attempt to find a difference in

the chromosomes which might correspond to the sex determining "het­

erochromosomes" or "idiochromosomes" of the animal kingdom. Both

the pollen mother cells and the somatic cells of the root of C a n ­

nabis were investigated and several stages in the reduction d i v ­

ision figured.

The small size of the pollen mother cells which are only

13 microns in diameter renders the study of the divisions somewhat

difficult to follow in detail. The present investigation includes

the chief events of the divisions of the pollen mother cell with

especial reference to the prophase of the heterotypic division.

Heterotypic Division.

The sporogenous tissue of the sporangium at the mother

cell stage consists of a cylinder 10-12 cells in diameter and about

50 cells long. A wide variation in the stage of development of the

sporogenous cells is noticeable especially during the homoeotypic

division which is more rapid than the the heterotypic. In the di f ­

ferent sporangia of an anther and the different anthers of a flower

the variation is much greater and all stages of both divisions were

traced within a single flower. No uniformity exists as to which

end of the loculus contains the younger stages,for progression may

be made from either the upper or lower end of the sac to the o p p o ­

site extremity. In the lower part of one sporangium the mother cells

are in the resting condition at the end of the heterotypic division

while in the upper end they are in telophase of the h o m o e o t y p i c >

An adjacent loculus shows the four cells resulting from the hom-
 33

oetypic division below and the telophase stage of the second d i v i ­

sion above. Some sporangia show zones of development,as for e x a m ­

ple one in which the dyad stage of sporogenous cells appears

in both e n d s , with later stages of metaphase and anaphase in the

central region.

Because of the large mass of mother cells in one s poran­

gium some variation is not unusual as was found by Allen (1) in

Lilium eanadense. However,the lack of uniformity between the anthers

of a single flower seems -greater than has been observed in the n o r ­

mal development of pollen mother cells. Gates (26) who found such

an irregularity in Oenothera lata,a sterile hybrid,connects this

condition with the failure of pollen development. Since sterility

appears in the later stages of the pollen development of Cannabis

similar to that observed by Gates (26) in Oenothera,the coincidence

is worthy of note.

In early prophase the pollen mother cells pass through the

usual growth period resulting in a marked increase in the size of

both the cells and their nuclei (figs.54-56). The cytoplasm at this

stage is exceedingly dense (fig.54) and granular,showing few v a c ­

uoles and staining darkly. The cells are angular exhibiting no

evidence of rounding off as yet. The nuclei are about 7 micrones

in diameter and contain each a relatively large globular nucleolus.

An extremely delicate reticulum composed*of thread-like strands

upon which small chromatin granules are distinguished forms a n e t ­

work around the nucleolus (fig.54).

As the nucleus grows the reticulum becomes more pronounced

 34

and the chromatin granules larger (figs.55,56). The network can

now be seen more distinctly and the chromatin granules become

spread out along its length thus making the linin thread indis­

tinguishable (57). This appearance gives ground for the theory of

Gregoire (41) who maintains that the reticulum is at times composed

of one general material which varies in density so as to give the

appearance of granules. The cytoplasm surrounding the nucleus becomes

less dense (figs.56,57) and somewhat vacuolated.

In a very few instances the mother cells at this stage

appear to be disorganizing as evidenced by their shrunken condition

and dark stain (fig.58). Gregory (32) in a sweet pea hybrid and

Gates (26) in Oenothera lata report sterilization at this early stage.

As the nucleus approaches synapsis the chromatin becomes

more conspicuous and more closely aggregated. The nuclear f r a m e ­

work which now consists of a continuous spirem (fig.58) shows no

evidence of being double. At this time the nucleus has reached its

growth limit and measures 11 microns in diameter being surrounded

by a very delicate nuclear membrane. The c ytoplasm of the cell b e ­

gins to round off slightly (fig.59)*

.Synapsis which as usual consists of the contracting of the

nuclear framework into a knot,was found as a very common condition

in the material studied (fig.59)* Although this stage was long r e ­

garded as an artifact due to faulty fixation of material it is now

generally recognized as a constant character of the mother cell.

Recent careful investigations on both the animal and plant sides

show that the peculiar phenomena connected with this stage are
 35

uniform. Special significance has been attached to this stage by

some who maintain that an exchange of material between the materanl

and paternal chromosomes takes place here. Other investigators

among whom is Gates (27) refute this theory. The latter author in

a paper on chromosome reduction states that synapsis can have no

special significance since the chromosomes are known to be paired

in the somatic tissue of the sporophyte,and because there is no

satisfactory evidence cf the existence of smaller units of struc­

ture than the chromosomes. The theories in regard to the sign i fi ­

cance of this period can perhaps never be substantiated satisfac-

torialy- by observations alone,but will remain to be solved by ex ­

perimentation.

The contracting of the spirem into synapsis gees on rapidly

resulting in a compact mass usually at one side of the nuclear

chamber (fig.59). The nucleolus remains distinct and may be either

within the mass of chromatin or at one side of it. This condition

is doubtless a. prolonged period since it is so commonly seen.

The loosening of the synaptic knot results in a much coiled,

delicate,uniform spirem (fig.60) The chromatin granules can no

longer be distinguished in this homogenous thread,nor was any e v ­

idence found of either a longitudinal split or a pairing of threads.

The duration of the spirem stage is very short and segmentation

soon takes place.

The cutting up of the spirem results in rather long straight,

curved,or b e n t ,segments (fig.61). The mode of reduction here seems

to be undoubtedly telosynaptic. X,V,U figures are formed by the

 36

segments and the end to end arrangement of a pair of chromosomes

is often very evident (fig.61) Soon the segments contract and blocky

bivalent chromosomes appear distinctly (fig.62),taking up their

position at the periphery of the nucleus.

The phylogenetic significance of the mode of reduction has

been somewhat questioned by Gates (27). He concludes that either

mode of reduction may take place and often both modes occur in the

same species. From a study of various species he states that in

forms which have long thread-like chromosomes the pairing may be

expected to take place side by side while in forms with short,

stout chromosomes the pairing is likely to be end to end. The short

blocky nature-cf the chromosomes of Cannabis and their telosynaptic

origin seem to substantiate this generalization of Gates (27).

At the dyad stage of the chromosomes a count showed the r e ­

duced number to be 10 (fig.6 2 ) , a number which was later found to

agree with that of Strasburger (62) who made a count from a plate

view of the spindle in metaphase. He also determined the diploid

number of c hr omosomes,2 0 , in a somatic cell of the root.

Previous to the formation of the spindle the chromosomes

aggregate in the center of the cell. The nuclear membrane and n u ­

cleolus disappear and spindle fibers are seen extending toward the

chromosomes (fig.63). The usual multipolar spindle is distinguished

which,however,scon takes a bipolar form. The chromosomes,which are

at first scattered along the spindle,later’ come together in the

equatorial region (fig.64) and are closely packed together. The

pulling apart of the chromosomes (fig.65) results in a compact

 37

chromatic mass at each of the spindle poles (fig.66)

During the reconstruction of the daughter nucleus s nuclear

membrane is formed around each mass (fig.67) and the nucleus begins

to grow in size. The chromosomes become separated (fig.66) and,as

the nucleus becones larger,seem to stretch out becoming scattered

along the reticulum. The chromatin masses,however,remain distinct

(fig.69) and a count of these shows in no case more than ten p r e ­

sent in one nucleus.

Homoeotypic Division.

Although the process involved in the division of the pollen

mother cell nucleus extend over s long period,the division of the

daughter nuclei is rapid. Practically all the stages of this d i v ­

ision may be traced within a single sporangium.

The daughter nucleus increases in size preparatory to b e g i n ­

ning the second division. The spirem spreads out,the segments being

distinct and located at the periphery of the nuclear chamber. The

nucleus becomes more rounded (fig.69) sr.d at the time of the d i s ­

appearance of the nuclear membrane is nearly spherical.

With the breaking down of the nuclear- membrane cytoplasmic

fibers appear extending toward the chromosomes (fig.70) which are

massed in the nuclear chamber es in the preceding division. The

small spindle figures are formed rapidly and lie either parallel

or at right angles to each other (fig.71)- The chromosomes s e p ­

arate as in the heterotypic division (fig.72) and appear later

massed at the poles (fig.73). The reconstruction of the daughter

 38

nucleus resembles the process described for the first division

and results in four nuclei (74) arranged t e tr a hedrally in the cell.

A noticeable sterilization occurs at the tetrad stage of

development as is shown by the shrinking ana dark stain of whole

tetrads or parts of tetrads. A more detailed account of this p e c ­

uliarity follows in the chapter on sterilization.

Much evidence has been accumulated recently in support of

the individuality and permanence of the chromosomes. The resting

stage of the spore mother cells,their daughter and grand-daughter

nuclei show the chromatin in definite masses which apparently c o r ­

respond to the "chromocentren" of Rosenberg (52) and the "prochrc-

mosomes" of Overton (47).

Special Kail of Yeung Tetrad.

Inside of the mother cell wall another wall is developed

around the protoplast which becomes distinct early in the prophase

of the heterotypic division,being first observed during the f o r m a ­

tion of the dyad chromosomes (fig.6l). A similar wall was observed

by Mottier (45) in Acer negundo L. in the spirem stage of the mother

cell and described as a "soft thick wall". During the succeeding

stages of division this wall becomes thicker and at the tetrad stage

of the mother cell is very conspicuous in the ircn-alum.-haematoxylin

preparations where it stains brown (figs.75,76). Beer (3) who o b s e r ­

ved similar walls in the tetrads of Ipomea describes them as "mass­

ive mucilaginous walls" which give reactions to callose and pectose.

Soon after the formation of the nuclei of the tetrad certain t h i c k ­

enings appear on the inner side of this wall ( f i g . 66) c o r r e sp o n d ­

 39

ing in position to the cell plates between the four nuclei. These

thickenings increase in size becoming ridge-like projections (figs.

69,66) growing toward the center of the cell. Partition walls between

the cells ore formed apparently independent of the investing wall

and connecting with the ridge-like thickenings (fig.76). Each cell

of the tetrad then invests itself with a delicate cell wall (fig.

68) which later becomes differentiated into the intine and exine of

the pollen grain. While the wall of the microspores is being f o rm ­

ed the heavy wall of the tetrad disorganizes (fig.77) and finally

disappears (fig.78).

The young microspores in many cases appear shrunken and

degenerating. A definite period of d e ge neration,which will be d i s ­

cussed. later,occurs at this stage similar to that observed in the

young tetrad.

Kale Gametophyte.

The germination of the microspore beginning with the d i v ­

ision of the nucleus takes place after the spores have separated

from their tetrad arrangement. The microspore nucleus is relatively

large,central in position and exhibits definite chromatin masses in

the reticulum (fig.65). The division of the nucleus was not o b s e r ­

ved. The daughter nuclei resulting from the division (fig.88) are

different in size the tube nucleus being larger than the generative.

The generative cell with its cytoplasm about it lies in the pollen

grain and as far a.s observed maintains its spherical form.

Experiments were tried in germinating the fresh pollen

 40

during the summers of 1912 and 1913* Successful results obtained

by Dr. R. B. Wylie with the pollen of Sambucus canadensis L. and

other species placed in sugar solution,led to these trials. The

pollen of Cannabis sativa.L. was taken from unopened flowers c o l ­

lected in good condition and placed in sugar solutions varying in

strength from 1/2 to 10 per cent. Although the experiment was r e ­

peated frequently no pollen tubes could be grown. Briosi and T o g ­

nini (9) report that the pollen of Cannabis placed in a damp c h a m ­

ber germinated in 4 or 5 hours sending out tubes with heavy walls.

Although the greater number of the pollen grains used for

experimentation appeared shrunken and sterile a few presented the

normal,plump appearance of functional pollen. These latter grains

would be expected to germinate under favorable conditions. The

negative results obtained from the experiments in germinating the

pollen lead to one of two conclusions. Either the medium in which

the grains were placed was not a suitable one for its germination,

or,none of the pollen was functional. Briosi and Tognini (9) s u c ­

ceeded admirably in germinating the pollen of the Italian hemp in

a damp chamber. The pollen of the American Cannabis placed in a

nutritive solution in which the pollen of other species had been

successfully germinated,failed to develop tubes. Whether or not

this indicates sterility in all the pollen of the species inv e s ­

tigated is a question.

Wall of Pollen Grain.

The young microspore is invested with a thin wall (fig.81)

 41

which becomes differentiated into the inner intine and the outer

exine (fig.65) The mature pollen grain (fig.88) is spherical when

normal,20 microns in diameter and has a wall which is very thin in

comparison with the spore costs of most land plants. The outer cut-

inized layer is delicately aculeolate when observed in the dry c o n ­

dition but appears smooth when immersed in a liquid.

At three points equidistant from each other a small papilla-

like protuberance appears on the pollen grain marking a thin spot

in the exine for the exit of the pollen tube (figs.85,88). A similar

number of such points is noted by Coulter and Chamberlain (17) for

the Cu p u l i fe r a e , Fr o t e a ce a e , G eraniaceae,Cnagraceae,Boraginaceae and

Compositae. Each point in section shows a small aperture opening

into a narrow canal that crosses the exine and enters a small c h a m ­

ber between the intine and exine (fig.91). The outer layer is thin

at this point and enlarges around the opening to form an annular

lip. Corresponding with each aperture the intine forms a h e m i s p h e r ­

ical body (fig.91) which Briosi and Tognini (9) state is stored

with cellulose to furnish material for the walls of the pollen

tube.

Dehiscence and Pollination.

The mature anther wall consists of the epidermis and o u t e r ­

most parietal layer or "endothecium",a name given by Furkinje (51)

to the dehiscing anther- wall within the epidermis or exotheciun: and

since become restricted to the outer parietal layer. Briosi and log-

nini (9) have described the reticulate system of lignified t h i c k e n ­

 42

ings and also the method of the dehiscence of the wall. A non-lig-

nified and contractible membrane connects the thickenings,and the d r y ­

ing of the» membrane produces a consequent constipation of cells. The

epidermal layer in drying also contracts more strongly than the endo-

thecium and consequently exerts a force which tends to make the wall

curve outwards and break at its weakest point. The endothecium and

the epidermis become smaller and weaker toward the connective tissue

and as a result of the tension of the anther wall tend to rupture

at these places in a longitudinal line.

Pollination is affected by the wind as is usual in the p r i m ­

itive unisexual flowers of Angiosperms and is the rule in the Urti-

caceae. The flowers of Cannabis display those negative characters

common to anemophilous plants - the absence of light-colored floral

envelopes,perfume and honey. The loose staminate inflorescences,

stalked and drooping staminate flower and pendulous stamens lend

themselves admirably to the shaking of the wind. On the other hand

the large,brush-like character of the stigma,the location of the p i s ­

tillate flowers in the axils of leaves are adaptations which fit the

flower to catch and retain the wind-blown pollen. The character of

the pollen itself fits it for wind distribution. Its abundance, i n­

coherency, dryness, smoothness and lightness render it easily carried

and distributed in the air. Although the sexes are se p a r ated,pollina­

tion is well cared for because of the' great abundance of spores. An

estimation of the number of grains produced by one flower based upon

a count of the mother cells gave about 452,400. Strasburger (60) for
one staminate hemp plant computed 12,500,000 grains. This number is
 43

perhaps conservative since it must include only a count of the flowers

which were on the plant at a given time without regard to those which

had gene or to those which would appear later. Considering that the

plants in the uncultivated state grow thickly and with the sexes

intermixed pollination seems to be well assured.

 44

STERILIZATION OF POLLEN.

Sterilization in the raicrosporangium of Cannabis sativa L.

is a very conspicuous feature. As was noted in the preceding chapter

the degeneration of potentially sporogenous cells occurs at several

stages in the development of the pollen. An attempt is made here to

describe the breaking down of the spores and also to obtain definite

counts showing the ratio of the sterile to the apparently functional

spores.

The development of the mother cells of the microsporangium

proceeds in a normal manner to the tetrad stage with the exception

of a scattered few which degenerate at various early stages (fig.58).

However,at this stage a marked difference in the cells is noted,

some appearing plump and normal (fig.77),others showing evidences

of breaking down or being alomst completely degenerated (figs.79,

80). The latter take a deep stain and are shrunken. Gates (26) in

a study of the mi crosporangium of Oenothera lata and Oenothera

Laraarckiana, found many irregularities in the tetrads and also e v­

idences of their breaking down.

The sterilization does not occur in a regular manner

but varies in different sporangia and different anthers. One sac

may contain only a few degenerating cells while another may be

almost completely sterile. Irregularity is also noted in the d e ­

gree of sterilization within the tetrad,1,2,5 or all the spores

breaking down. The abortion of spores often occurs in regions,one

 45

end of a sporangium showing practically all the cells shrunken while

the remainder contains normal cells. Again the functional and abortive

microspores may be interspersed. At times a whole anther sac and even

a whole flower shows sterility,while on the other hand some exhibit

no evidences of degeneration.
Owing to the irregularity of the sterilization much diffi­

culty is experienced in attempting to determine the exact ratio

of sterile microspores to those whioh appear to be functional.

Numerous counts made from prepared slides to determine the percent

of abortion that occurs at the tetrad stage showed that a p p roxima ­

tely 40 per cent of the number degenerates (Plate XXIII).

The young microspores after they have separated from the

tetrad arrangment resulting from the division of the mother ceil,

show various degrees of degeneration. Some of the spores which be­

gan to break down earlier appear now to be almost completely d i s ­

organized (fig.84). Others are evidently just beginning to deg e n ­

erate (figs.82,85) and appear slightly shrunken.

As the microspores enlarge and become rounded those which

formerly appeared to be entering upon a period of degeneration are

now darkly staining and small (fig.87)- ‘T hese, however, do not co n ­

stitute a large per cent of the spores being only about 10 per cent

of the number (Plase XXIII). The remaining microspores appear almost

normal at this stage but often show slight evidences.of shrinkage

(fig.86) which is doubtless the beginning of the sterilization

•30 evident in the mature pollen grains.

The mature pollen in the fresh condi tion was examined d u r ­

ing the summers of 1912 and 1913 and presented a peculiar app ear­
 46

ance. The greater number of grains was badly shrunken (figs.89,90)

their form being angular and irregular. Even pollen found in u n ­

opened flowers showed this appearance of d egeneration and only a

few were round and plump appearing as normal functioning grains.

The .ratio of shrunken to normal grains was determined to be about

50 to 1 or 98 per cent of the number (Plate XXIII). Counts both
of the fresh material and of prepared slides gave a great p r e p o n ­

derance of badly shrunken*grains. The fact that the fresh pollen

showed evidences of degeneration precludes the idea of apparent

abortion due to faulty fixation of material.

It has been shown that at several stages in the d e ve lop­

ment of the microspore and also in the mature pollen,sterilization

of potentially sporogenous tissue is evident. The proportion of

the sterility that occurs at the various stages is of interest in

this connection. Considering that each mother cell would give rise

to four spores,the abortion of spores appearing at scattered stages

of the mother cell before its division,is a pproximately 1 per cent

of the whole. At the tetrad stage 40 per cent of the remainder

becomes sterile. Another period of degeneration occurs at the e a r ­

ly microspore stage when approximately 10 per cent if the remainder,

or,5 per cent of the whole break down. The mature pollen shows the

greatest amount of sterilization. 98 per cent of the pollen grains

or about 44 por cent of the total number of spores that would have

been forme'd if no sterilization had occurred, become aborted. The

sum total of these various losses through sterilization between

the mother cell and the mature spore,approximates 99 per cent of

 47

the total potential number of spores and leaves but 1 per cent

of apparently functional pollen. (See Plate XXIII).

The significance of this sterilization is taken up in the

chapter on the discussion.

Sterilization of Pollen.

Per cent b e ­ Per cent of total Per cent

Stage
coming sterile potential pollen remaining

Early sporogenous
cells 1% 1% 99%

Tetrad 40# 39.6% 50.4%

Microspore 10% 5% 45%

Mature pollen 98% 44.1% 1%

 48

EMBRYO AND FRUIT.

A study of the embryo development was not attempted since

the material proved unfavorable. Prain (50) reports the presence

of albumen in the seed and that the posterior cotyledon of the

embryo is noticeably larger than the anterior one.

The fruit is a nut enclosed in a leathery husk and c o m ­

pletely filled by a fleshy kernel. In terming the fruit a nut

meaning a one-seeded fruit,enveloped by a husk,with a hard shell

composed of more than one carpel,we accept the theory of a bicar-

psllary ovary. The fruit is smooth,ovoid,and of a greenish gray

mottled appearance. The husk is the remains of the involving bract

and the mottling is due to the remains of the perianth.

A layer of irregular columnar cells in the endocarp is

responsible for the crustaceous character of the shell.

 49

DISCUSSION.

Relative Rank of Species.

As previously noted,Cannabis sativa L. has been classified

in the Urticales one of the primitive orders of the Archichlamydeae.

Accepting the evidence that the simpler flowers are primitive rather

than derived, this classification is warranted by certain features

which may be regarded as primitive as far as our present knowledge

of the lines of descent can determine.

The small inconspicuous flowers are ch aracteristic of prim­

itive forms which are wind-pollinated. The dioecious habit also

prevails among the lower seed-bearing plants and is nearly always

associated with wind pollination. The tendency to dioecism is g e n ­

eral for the Urticaceae and the more primitive groups and hence

may be regarded as a primitive character.

The staminate and pistillate flowers differ from each other •

in many respects. On the whole the staminate flower displays more

primitive characters than the pistillate. The parts of both flowers

are definite and cyclic in arrangement,features which entitle them to

a comparatively high rank. However,the single inconspicuous per­

ianth marks them as less specialized than the higher forms which

have a double showy perianth. The staminate flower has separate

floral parts which are developed independent of each other. The

pistillate,however,has both perianth parts and carpels united or

 50

coalesced,which tendency is more advanced than that of separate

development displayed by the staminate flower. Hypogyny as is

regarded by Coulter and Chamberlain (17) is un doubtedly a prim­

itive feature.

The ovule is campylotropous,which is a rare type and e s ­

pecially peculiar to the Urticaceae where the orthotropous ovule

is common. Its cauline origin and double integument mark it as

primitive. Lower forms in both Monocotyledones and Dicotyledones

have ovules of cauline origin and hence the character is to be

regarded unspecialized. The two integuments found in Cannabis

prevail also among the Urticaceae and Archiehiamydeae in general.

Since a single integument characterizes the Sympetalae and higher

groups the double i n t e g u m e n t .is doubtless a primitive feature.

A noteworthy character brought out by this study is the

many-celled archesporium of the megasporangium. Briosi and Tognini

(9) report a one-celled archesporium for this species which would

indicate a more advanced condition. The many-celled archesporium

noted in this investigation is doubtless a primitive character since

as Coulter and Chamberlain (17) state it occurs quite commonly in

the Amentiferae,Ranunculaceae,Rosaceae and other lower forms. The

occurrence of both the one celled and many celled archesporium in

the same species may indicate that' the plant is shifting its habits.

A conspicuous development of parietal tissue appears in

the megasporangium. This is probably correlated with the many-celled

archesporium and is certainly a primitive charact er since higher

forms have a tendency toward its suppression.

 51

The branching of the pollen tube of Cannabis has been r e ­

ported by Zinger (69). Since this feature, is common among 3ym.no-

sperms and the Ament iferae of Angiosperms it is doubtless c h a r a c t e r ­

istic of lower forms.

Cannabis sativa L. although displaying mostly primitive

characters,has certain tendencies which may be regarded as more

advanced. The primitive characters as noted above are the small^

inconspicuous,dioecious flowers,the single,bracteate perianth,

hypogeny,the cauline ovule,the double integument,the many-eelled

archesporium,the great development of parietal tissue of megas p o r ­

angium and the branched pollen tube. On the other hand the definite

number of floral parts,the cyclic arrangement of parts,the zonal /

development of the floral organs of the pistillate flower and its

zygomorphism are more advanced characters.

Diclinism.

The problem of sex in plants and animals is very old and

has long been o f interest among biologists because of its relation

to the important question of heredity. Sexuality is almost universal

in the organic kingdom being lacking in the lowest forms only.

It is not always present in the higher plants,but,as Schaf-

fner (55) states,their relationships and morphology clearly point

to a sexual ancestry. Sexual difference in plants extends not only to

the sexual generation but to the sporangium and entire sporophyte

as well. The staminate and pistillate flowers differ in number,

position and structure,the former being specialized for the pro-

 52

duction and distribution of pollen,the latter for the protection

and ripening of the ovules and the distribution of seed.

As before noted the strictly dioecious character of Cannabis,

its easy cultivation and world-wide distribution have caused it to

be the subject of much experimental work and discussion in relation

to this question. Problems of the character of the sexual dimorphism,

the difference between the staminate and pistillate individuals,

sex ratio,sex constancy and the time of the determination of sex

in the individual have been studied through investigations upon

Cannabis.

Differences between the staminate and pistillate individuals

of Cannabis are not easily discernible until the time of flowering.

After blooming a marked difference i3 noted. The staminate plants

✓
turn yellow and soon die,while the pistillate become more luxuriant

and persist until after the ripening of the seed. The staminate plant

with its drooping stalked flowers aggregated in loose inflorescences

is specialized for the production and distribution of pollen. On the

other hand the pistillate plant with its sessile flowers arranged

in compact inflorescences is specialized for catching the wind-blown

pollen. The persistence of the latter plant also allows' time for

the ripening of the seed.

Goebel (29) in an article on the sexual dimorphism of plants

states that the staminate and pistillate individuals are rarely

unlike in secondary sexual characters before blooming and cites

hemp as an example. He explains that in Dico tyledones a post -fer­

tilization development of the vascular system in the pistillate

 53

plants is responsible for its longer life,while the weak vascular

supply of the staminate plant causes its early disappearance.

Sprecher (57) made a study of the sexual differences in

Cannabis sativa and Rumex acetosa. He measured the variability in

length,weight and osmotic pressure of the staminate and pistillate

plants when grown under different conditions and concludes that the

variation is much more extended in the pistillate individuals than

in the staminate.

The ratio of sexes'in the lower plants is far from being

equal. In dioecious Angiosperms the ratio is more nearly equal but

/
usually shows a preponderance of the pistillate individuals. Hemp

has been considered by many experimenters in relation to this question.

Heyer (34) examined 40,000 plants of Cannabis and determined the

proportion of sex to be 100 staminate to 114.93 pistillate; Hab-

erlandt (33) is Austria found the ratio of the same species to be

100 staminate to 120.4 pistillate plants; Fisch (25) counting 66,000

plants at Erlangen found the ratio of 100 staminate to 154.24 p i s t i l ­

late individuals. As Schaffner (55) states the case of hemp shows a

ratio which is exceedingly variable. However,the variation in porpor-

tion seems constant for' the various races observed when large counts

are taken.

The question then arises as to what conditions cause this

constancy of sex relation and why is it recognizable in such large

counts ? Until recently it was generally believed that sex in plants

is largely determined by conditions of environment,such as t e m pera­

ture, light, nourishment, conditions of growth,etc. This belief was

 54

founded upon certain successful experiments in Cryptogams where the

sex organs are developed upon smal l,free-living,few-colled prothallia.

Here,by the changing o f ‘external conditions,the thalli could be

induced to produce either male or female sex organs - unfavorable

conditions causing the male organs to appear and favorable the female.

From experiments performed upon various groups of plants it seems

proven that environmental factors do influence either directly or

indirectly the development of sexual organs in plants where both

I

tendencies exist. However recent and more careful experiments show

that in strictly dioecious forms such factors do not determine,at

least in the life history of the individual,which sex shall d e vel­

op.

Perhaps no dioecious Angiosperm. has been investigated as much

as Cannabis in an attempt to solve the problem of the determination

of sex. Haberlandt (33),Saccardo (53),Heyer (34), Hoffmann (35) and

many other experimenters worked with this species in attempting to

determine the influence of external conditions upon sex. Their c o n ­

clusions are as follows: *

1. Sex is not influenced by changes in external conditions

such as temperature,light,soil,etc.

2. Sex cannot be determined by the position of' the seed upon

the mother plant or by any general characters of the seed.

3. Sex of dioecious Angiosperms is already determined in the

. seed.

Noll (46) experimented with dioecious Angiosperms including

Cannabis sativa L. in an effort to determine the cause of the d e f ­

 55

inite ratio of staminate and pistillate individuals and the time

of sex determination. His experiments lead to the conclusion that

sex must be deterimned at the time of f ertilization by the paternal

sex cells. His theory concerning the differentiation in the pollen

grains is that two spores from a tetrad have a dominant and two a

recessive staminate tendency. Hence the union of a sperm from a pollen

grain having a dominant staminate tendency with the egg gives rise

to staminate offspring. Also the union of a sperm from a pollen

grain having a recessive staminate tendency with the egg gives rise

to pistillate descendants.

Correns (16) experimenting with monoecious and dioecious

Angiosperms agrees with Noll(46) that the paternal sex cells deter­

mine the sex of the descendants,that the tendency of the egg is

always pistillate and that the differentiation of paternal sex cells

takes place in the reduction division of the pollen mother cell.

However,the theory of Correns (16) differs from that of Noll (46)

for he argues that two spores resulting from the tetrad have a s tam­

inate and two a pistillate sexual tendency.

Strasburger (62) in a recent publication records the results

of experiments and observations concerning the cause of sex de t e r ­

mination in.dioecious Angiosperms. He attempts to prove the theory

of Noll(46) and Correns (lo) concerning the differentiation of sex

in. paternal sex cells by pollination e x p e r i m e n t s , but his results

are negative. The division processes of the sporogenous cells of

several dioecious forms were investigated in an effort to determine

the cytological basis of sex separation. The work of Stevens (58),

 56

Wilson (68) and others has shown that certain sexual differences
♦ ..

of chromosome groups exist in the spermatozoa of insects. Darling (18)

„ reports the presence of certain "acccessory" or "idiochromosomes"

in Acer negundo which he states are analogous to those found in the

animal kingdom. However,Mottier (45) in a recent investigation of the

pollen formation of the same species denies the existence of such chro-
»

mosomes here. The observations of Strasburger. (62) indicate that there

are apparently no differences in the size or number of chromosomes in

the pollen produced from a give tetrad of microspores.

The present study which follows the divisions of the pollen

mother cell in greater detail than was done by Strasburger (62) c o n ­

firms his theory that in the pollen of Cannabis sativa L. there is

no significant difference in the division products resulting from

one pollen mother cell. Neither in the size or number of chromosomes

is there any indication of a differentiation which might be c o r r e l ­

ated with the observed condition in insects.

The inheritance of sex in plants as well as animals

is a debated question. Correns (16) as a result of c r o s s - pollina­

tion experiments concludes that the staminate plants of dioecious

forms are heterozygotic and the female homozygotic and that sex is
is
inherited according to Mendelian rule. Strasburger (62) concludes

that the staminate tendency is subjected to a weakening in forms

which have resulted from self pollination. The ratio of staminate

to pistillate individuals is somewhat dependent upon the chance

pollination of the species and. the preponderance of pistillate

individuals would be due to the recessiveness of the staminate

 57
r

tendency in more than half of the pollen grains.

Strasburger (62) would agree that both sexual tendencies

exist in the individual one being dominant and one recessive. The

original separation of the sexes is completed sometime in the hap-

loid generation and is not necessarily bound to the reduction div­

ision as Correns (16) and Noll (46) believe. Primary sex separation

then precedes fertilization and its separation in the reduction d i v ­

ision is only secondary.

The strictly diocious character of Cannabis sativa L. has

led to much discussion concerning the nature of its diclinism.

Some regard the unisexuality of this species as inherent and prim ­

itive. Others believe that the flowers have become unisexual by the

atrophy of the androecium in the pistillate and of the gynoecium in

the staminate flowers.

Briosi and Tognini (9) regard the dicliny as a primitive

condition since there are no rudimentary sex organs in either the

staminate br pistillate flower. Furthermore a certain meristematic

activity noted in the flowers seems to indicate a disposition to

produce new organs. They observed certain mersitematic structures

between the perianth and ovary of the pistillate flower and consider

these as beginnings of stamens rather than rudiments. The rare o c ­

curence of hermaphrodite flowers is taken to be indicative of a

condition which is beginning rather than one which is inherent

and primitive.

Praia (pO) agrees with Briosi and Tognini (9) that the

dicliny is primitive supporting his theory by observations of ter-

 58

<
atological phenomena. In the hermaphrodite flowers of Maria hemp

a single stamen is found closely applied to an open mon ocarpe l-

lary ovary. In the Sheria hemp various combinations between stam-

inate and pistillate floral organs exist. Staminody of the pistil

resulting in the substitution of a stamen for a style is the com­

monest phenomenon. The author states that since the alternation

of the sexual organs precludes any comparison between the parts of

the staminate and pistillate flower the diclinisra is in the highest

degree inherent.

Dicliny is generally considered as a derived condition

since many families show transition stages from the monoecious to

the dioecious forms. The experimental studies of N o l l ,Corirens and

Strasburger indicate that both a staminate and pistillate tendency

is present in dioecious Angiosperms. In their phylogenetic d e vel op­

ment the paternal sex cells have become differentiated and a sep­

aration of sexes results.

In this investigation of the American Cannabis no herm­

aphrodite, monoecious or other intergrading c o n d i t i o n s 'have been

noted. The plants observed were growing in waste fields. The accounts

of teratological phenomena reported by various authors indicate

that abnormal conditions appear more commonly in cultivated places.

Since the plants are given more favorable growth conditions in

cultivated fields the tendency toward monoecism would seem to be

a reversion to an ancient type brought about by unusual conditions.

If the dicliny here is derived rather than primitive the differ-

entiation of sex has gone far. The flowers themselves have lost
 59
r

all suggestions of the suppressed parts and differ completely in

form, structure, and function. The differentiation has also extended

to the staminate and pistillate plants which have become specialized

for pollen and seed bearing.

Sterilization of Pollen.

The phenomona of sterilization so evident in the m i c r o ­

sporangium of Cannabis sativa L. lead to the question of its cause

and significance. Sterility in pollen formation has recently received

much attention from cytologists because of its common occurrence in

hybrid forms,and because of its significance in the determination

of the hybrid origin of species. Sterilization has never been reported

for Cannabis sativa L. Hence its appearance here leads to a co m ­

parison of this species with other plants in which sterility has

been observed.

That hybrids are frequently sterile has long been

known and commented upon. According to Webber and Swingle (64)

hybrids arising from the union of widely different parents are co m ­

monly sterile often producing little or no pollen and more rarely

having defective ovules. All students of hybrid forms agree that

the pollen is more likely to be imperfect than the ovules. Hence

the commonest consequence of hybridization is the imperfect f o r ­

mation of pollen grains in the offspring. Often the anthers are

completely empty or they ace small and do not open.

 60

•Juel (39) who made a study of the pollen development in the

sterile Syringa Ro thomogensis,a cross between S.vulgaris and S.

persica,found many irregularities in the reduction division. De­

generation sometimes begins as early as synapsis,but usually the

tetrad divisions are completed.

Rosenberg (52a,52b) in a study of the pollen development

of Drosera longifolia obovata,a hybrid of D.r otundi folia and D.

longifolia,observed that many pollen grains lose their contents

and also that the embryo-sac development usually stops at the bi-

nucleate stage.

Gregory (32) noted an early degeneration in prophase of

the pollen development of hybrid sweet peas.

Tischler (62a) ascribes sterility in the hybrids of Ribes

and Bryonia and also hybrids of Lathyrus odoratus to the influence

of long culture.

Gates (26) made a critical study of the pollen development

in hybrids of Oenothera lata x 0. Lamarckiana. Sterility appears

hers rarely in the early stages of the mother cell but is more

evident in the tetrad stage. Oenothera Lamarckiana also shows ir­

regularly shaped pollen grains and occasionally complete sterility.

Sterilization in the pollen of a number of species of Po-

tentilla was observed by Wullff (66). The author explains the

cause Qf the sterility as a series of influxes of the outer life

conditions of the plant and suggests that perhaps parthenogenesis

may be found in these forms.

Holden (36) records certain peculiarities in Betula pumila

 61

which suggest its hybrid origin. Large areas of abortive spore-

mother cells were observed in the staminate cones and later the

degeneration of three spores of each tetrad was noted. The author

cites other peculiarities of the plant and states that spontaneous

hybrids can be diagnosed as hybrids by the investigation of their

internal anatomy.

Jeffrey (38) in a recent article in which he attempts to

prove that the genus Oenothera is subject to spontaneous hybrid­

ization, states that sterility is a noticeable characteristic of

species crosses or true hybrids. He observed the spores and pollen

of liverworts,mosses,ferns,lycopods,selaginella,quillworts,lepi-

dcdendrons,equiseta,cycads,ginkgo,conif ers,gnet ales,monocotyledon-

ous and dicotyledonous Angiosperms and always found known hybrids

to be characterized by abortive spores which have little or no

protoplasmic contents. He considers the sterilization of spores

as an indication of the hybrid origin of species. Jeffrey (38)

finds sterility in many species of Oenothera and concludes that

these forms are subject to spontaneous hybridization. Then the

mutation theory of De Vries based upon observations of various

species of this genus,according to this author appears to be only

a useless hypothesis. However .more investigation is needed in order

to establish a cytological and morphological basis for the phenomena

of mutation.
/
The occurrence of sterilization in the raicrosporangium of

Cannabis arouses the question as to whether or not the species is

a hybrid or is giving rise to mutants. According to Jeffrey ( 3 8 ) ?

 62

Holden (36),and others whose works are quoted above the sterility

of pollen in varying degrees is a peculiarity of hybrids. The o b ­

scure and ancient origin of this species makes it impossible to

determine this point; but the marked sterility of potential pollen

as proven for the first time by this study casts suspicion upon

its ancestry. Because of the similarity of the pollen condition to

that of hybrids the study of the plant from the stand point of m u ­

tation would doubtless prove profitable.

In spite of the recent morphological and cytological studies

on the sterilization of hybrids no satisfactory explanation of its

cause has been made. According to Gates (26) some of the suggested

causes are : (1) lack of nutrition in the parts affected causing

failure in the development of the tapetum. and sporogenous cells;

(2) the influence of long culture; (3) irregularities of development

in the germ cells due to some lack of harmony between idioplasffls ;

(4) some more deep-seated phenomena affecting the whole plant.

These catagories are more or less general having no definite li m ­

itations. However,they serve to designate the widely different

points of view with which this subject is regarded.

Lack of nutrition due to the early degeneration of the

tapetal cells is not a possible explanation for the sterility in

Cannabis. The irregularity of the degeneration of the tapetal cells

does not permit of any correlation with the degene ration of the

sporogneous cells,since the tapetum often appears to be breaking

down after the sporogenous tissue has begun to degenerate.

Sterility as a result of long culture appears in many cul-

 63

tivated plants whose origin is unknown. De Vries (19) names many

common garden plants which have long been propagated b y vegetative

means and are sterile,sometimes eve destitute of flowers. Although

Cannabis sativa L. is widely cultivated,it is always propagated by

the seed. If the formation of the seed is dependent upon the act

of fertilization then the pollen must be normal and functional.

Whether or not fertilization takes place in Cannabis is discussed

later.

Irregularities in the pollen development due to a lack

of harmony between idioplasms is proposed by Gates (26) as a pos­

sible explanation of sterility in Oenothera lata. He states that

the intimate union of paternal idioplasms in synapsis might lead

to the development of incompatibilities between the plasms causing

irregularities in the reduction divisions. This explanation is more

or less theoretical based upon the unproven fact of the exchange of

material between maternal and paternal plasms in synapsis. The lack

of harmony between plasms of a hybrid derived from widely different

parents is conceivable. However,since Cannabis is not a known hybrid

this explanation of the sterility of the pollen would hardly suffice.

Furthermore irregularities during the early stages of the mother

cells are very few,the greater part appearing at the end of the r e ­

duction divisions.

Gregory (32) suggests that sterility of the pollen is the

expression of more deep lying phenomena affecting the whole plant.

A similar view is held by Tischler (62a). The American Cannabis

differs from the European and Indian forms in many particulars

 64

evidently due to a difference in climate and soil. It is c o n ceiv ­

able that sterility in the pollen formation might also be caused

by the different climatic and soil conditions.

In many apogamous Angiosperms the pollen is defective or

reduced. Strasburger (6l) observed apogamy in Elatostema sessile,

the pollen development of which is irregular. Miss Pace (48) also

observed apogamy in Atamosco and substantiates the theory of S t ras­

burger (61) that a diploid egg is incapable of fertilization. Wullff

(66) as was previously noted also suggests that this condition may

exist in Potentilla where the pollen Is sterile. Cannabis,according

to Winkler (65),is perhaps parthenogenetic or apogamous. If this

is the case,then a certain correlation may exist between the p a r ­

thenogenetic development of the egg and the defective condition of

the pollen. Hence,as was previously shown by germ ination expe ri­

ments the pollen may be entirely sterile if it is not necessary

for fertilization and seed formation.

Investigation of the embryo-sac of Cannabis is needed

to throw light upon this problem. The peculiar shape of the ovule

renders a study of the female gametophyte difficult,and it is

doubtless due to this fact that not much is known concerning its

development.
 65

SUMMARY.

1. The floral organs of the staminate flower appear in acropetal

succession - speals and stamens.

2. The floral parts of the pistillate flower arise in acropetal

succession - perianth and carpels.

3. The perianth of the pistillate flower is developed from two

primordia and is slightly higher on the adaxial side.

4. The floral oract of the pistillate flower is closely associated

with the flower proper.

5. The ovary is bicarpellary in origin.

6. The ovule arises lateral to the apex of the stem,becomes as­

sociated with the posterior carpellary wall and is carried

upward to the top of the ovarian chamber where it assumes a

pendulous position.

7. ‘T he ovule is campylotropous and has two integuments.

8. The styles are developed on the margins of the fused carpels.

9. The species is dioecious and the flowers show no indication

of suppressed organs.

10. The archesporium of the megasporangium is many-celled,but one

cell functioning.

11. The wall of the microsporangium consists of an endothecium,

two middle layers and a tapetum.

12. The tapetal cells divide mitotically in the early stages. Later

fusions of the nuclei result in irregularities in their number

and form.
 66

13. The disorganization of the tapetal cells occurs normally at

the early microspore stage of the pollen,but frequently at

various earlier stages.

14. A large mass of sporogenous tissue is developed in the micro-

sporangium, a single stamen having about 22,620 spore mother

cells.

Ip. The reduction divisions of the pollen mother cells reveal no

pecularities in the form and behavior of the chromosomes.

16.. The chromosomes remain distinct in the resting phase between

the heterotypic and homotypic divisions.

17. A special wall is developed around the pollen mother cell which

persists and becomes conspicuous at the early tetrad stage,

later disappearing.

18. The pollen grain at maturity is invested by a comparatively

thin wall in which are three exit points for the pollen tube.

19. The mature pollen contains a tube nucleus and a spherical g e n ­

erative cell.

20. All attempts at germination of the pollen resulted negatively,

no pollen tubes being formed.

21. The dehiscence of the anther is in a longitudinal line,but

there are no special lines of cleavage.

22. Pollination is affected by the wind.

23. Sterilization of potentially sporogenous tissue in the m i c ro­

sporangium occurs at all stages of pollen formation.

a. Early sporogenous cells -- 1 per cent.

b. T e t r a d - - - - - - - - - - - - - - - - - 39*6 per cent.

c. m i c r o s p o r e :- - - - - - - 10 per cent.
 67
r
d. Mature p o l l e n - - - - - - - - - 44.1 per c e n t .

24. The pollen remaining unshrunken and apparently functional c o n ­

stitutes 1 per cent of the total potential pollen.

 68

BIBLIOGRAPHY.

1, Allen,0. E.

1905. Nuclear D ivision in the Pollen Mother -cells of Liliura

canadense. Ann. Bot. 19: 189. 1905.

2. Beer, Rudolf.

1905. On the development of the pollen grains and anther of

some Onagraceae. Beih. Bot. Centralbl. 19: 286 - 513.

1905.

5. - - - -
1911. Studies in Spore Development. Annals of Botany 25: 199.

1911.
4. Benson, Margaret.

1894. Contributions to the Embryology of the Amentiferae. I.

Trans. Linn. Soc. London 31: 409 - 424. 1894.

5. Bentham and Hooker.

1880. Genera Planatarum . London. 1880.

6. Bonnet, Jean.

1912. Recherohes sur 1 * evolution des cel lules- nourricisres du

pollen,chez les Angiosperm.es. Archiv. fur Zellforsch-

ung. 7: 605 - 710. 1912.

7. Boodle and Fritsch.

1908. Solerederi's Systematic Anatomy of Dicotylsdones. Vol.

II.
8. Braun.

1872. Ueber eine monoecisehe Form des Hanfes - Sitzungsber

 69

der Gesssllseh. Naturforsch. Freunde,Berlin. 1872.

9. Briosi, G. and Tognini, F.

1894 - 96. Interne all" anatomia della Ganapa. Milan,(i) 1894;

(ii) 1896.

10. Britton, N.

1907. Manual of the Flora of the Northern States and Canada.

New York. 1907.

11. Celakowsky.

1874. Ueber die morphologische Bedeutung der Saraenknospen.

Flora. 1874.

12. Clarke, 3.

1835. Ann. iviag. Nat. Hist. 11. 1853«

13. - - - - -
1866. New Arrangement of Phanerogamous Plants. 1866.

14. Conrad, A. H.

1900. A Contribution to the Life History of Quercus. Bot.

Gaz. 29: 408 - 418. 1900.

15. Correns, C.

1907. Die Bestimmung U. Vererbung d. Geschlechtes n oh neuen

Versuchen mit hoheren Pflanzen. Berlin. 1907.

16 . ------
190S. Die Rolls der mannlichen Keimzellen bei der Geschle-

chtsbestimmung der gynodioeischen Pflanzen. Ber.

d. deutschen bot. Gesellsch, Bd. XXVI a,Jahrg. 1908.

17. Coulter, J. M. and Chamberlain, C. J.

1903. Morphology of Angiosperms. Part II.

 70

18. Darling, 0. A.

1909. Sex in Dioecious Plants. Bui. Tor. Bot. Club. 36:

177 - 199. 1909.

19. De Vries, Hugo.

1910. The Mutation Theory. Vols. I and IT. 1910.

20. Doell.

1857 - 62. Flora des Grossherzogthums. Baden;Garlsruhe.

1857 - 62.
21. Dusing, 0. A.

1885. Die experimsntelle Prufung der Theorie von der Reg-

ulierung des Geschlechtsverbaltnisse. Jenaiehe Ztschr.

f# naturwiss. 19. Suppl. II. 1885*

22. Eckler, 0. R. and Miller, F. A.

1913* A Study of American grown Cannabis i n Comparison with

Samples from Various other Sources. Lilly, Sci. Bui.

Series l,No.3: 93 - 101. 1913*

23. Sichler, A. W.

1975 - 78. Bluther-diagramms. Leipzig. 1875 - 78.

24. Engler, A. anf Prantl, K.

1887. Die Naturlichen Pflanzenfamilien. Leipzig. 1887-

2p. Fisch. C. -

1887* Ueber die Zahlenverhaltnisse des G eschlechtes beim

Hanf - Ber. d. d. bot. Bd. V. 1887.

26. Gates, R. R.

1907. Pollen Development in Hybrids of Oenothera Lata x 0.

Lamarckiana,and its Relation to Mutation. Bot. Gaz.

43: 81 - 110. 1907.

 71
r

27. - - - - -

1911. The Mode of Chromosome Reduction. B o t . G a z . 51:

321 - 342. 1911.

28. Goebel, C.

1887* Outlines of Classification and Special Morphology.

Sng. Trans. Oxford. 1887.

¥

29. Goebel, K.

1910. Ueber sexuellen Dimorphismus beiPflanzen. Biol. Cbl.

30. 1910.

30. Gray, A.

1908. New Manual of Botany. Sev. Ed. 1908.

31. Gregoire, V.

1907. La formation des gemini heterotypiques dans les vegetaux

La Cellule. 24. 1907.

32. Gregory, R. P.

1905. The abortive development of the pollen in certain sweet

peas. Proc.Camb. Phil. Soc. 13: 148 - 156. 1905.

33* Haberlandt, Fr.

1877* Welche Einflusse bedingen das Geschlecht der Hanfp-

flanZen? Fuhlings landw, Ztg. 881. 1877.

34. Heyer, F.

1884. Untersuchungsn uber das Verhaltniss des geschlechtes bei

einhausigen und Zweihausigen Pflanzen. Ber. aus a.

physiolog,. Laborat. d. landw. Instituts d. Universi-

tat Halle. 1 Bd. 5. Heft. 1884.

35. Hoffman, H.

1885- Ueber Sexualitat. Bot. Ztg. 43- Jhrg. I885.

 72
f

36. Holden, R.

1913* Anatomy as a means of Diagnosis of Spontaneous Plant

Hybrids. Science. 1913*

37. Holuby, H.

1878. Cannabis sativa monoica. Oesterreich Bot. Zeitsctirift.

28. 1878.

38. Jeffrey, E. C.

1914. The Mutation Myth. Science 39: 488 - 491. 1914.

39. Juel , H. 0.

1900. Beitrage zur Kenntniss, der Tetradentheilung. II.

Die Tetradentheilung bei einer hybriden Pflanzen.

Jahrb. fiiss.Bot. 3 5 : 658 - 649. 1900.

40. Kerr.

1893- Report on the Cultivation of,and Trade in,Ganja in

Bengal. 1893*

41. Kirchner, 0.

1905. Parthenogenesis bie Blutenpflanzen. Jatresh. d. Vereins f.

vaterl. Naturk. in Wurttemberg. 54. 1905..

42. Masters.

1869. Vegetable Teratology. London. 1869.

43. Modilewsky, J.

1908. Zur Samentwicklung einiger Urtifloren. Flora 98. 1908.

44. Montemartini, L.

1902. Sul valore morfologico dell r ovario e del 1 ovulo

della Canapa. Rendic. Congr. Botan. di Palermo. 1902.

4p. Mot tier, D. M.

1914. Mitosis in the Pollen Mother Cells of Acer negundo L.

 73

and Staphylea trifolia L. Annals of Botany. 28: llp-

130. 1914.

46. Noll, F.

1907* V o r l a u f i g e r A b s c h l u s s der V e r s u c h e u b er die B e s t u m m u n g

des G e s c h l e c h t s bei d i o e c h e n P f l a n z e n . S i t z u n g s b e r .

der n i ed e r r h. G e s e l l s c h . f. N a tu r - U . Heil k u n de .

Bonn. 1907.

47. Overton, J. B.
190p. Ueber Reduktioristeilung in den Po llmutterzellen einiger

Dikotylen. -Jahrb. Hiss. Bot. Bd. 42: 121 - 153. 1905-

48. Pace, Lula.

1913- Apogamy in Atamosco. Bot. Gaz. 56. 1913•

49- Payer, J. D.

1557. T r a i t e d 1 o r g a n o g e n i e de la fleur. Paris. 1857*

50. Prain, D.

1904. On the M or phology,Teratology,and Diclinism of the

Flowers of Cannabis. Sci. Mem. Gov't. India,No. 12.

Calcutta. 1904.

51. Purkinje, J. E.

1830. De Cellulis Antherarum fibrosis nec non de granorum

po ll inarium formis comraentatio phytotomico. Vrat-

islaviae. 1830.

52. Rosenberg, 0.

1909. Cytologische una morphologische Studien uber Drosera

longifolia x D. rotundifolia. Kungl. Sv. Vet. Akad.

H o n d l . 43. 1909.
 74
r

52a. --------

1903* Das Verhalten der Chromosomen in einer hybriden P f ­

lanze. Ber. Deutsch. Bet. Gesells. 21: 110 - 119.

1903.
52b. -------

1904. Ueber die Tetradentheilung eines Drosera - Bastardes.

Ber. Deutsch. Bot. Gesells. 22: 47 - 53« 1904.

53- Saccardo, P. A.

1879. Sülle cause determinant! la sessualita nslla Canapa - Bul.

Soc. Veneto - Trentina Sei. Nat. Padova. 1879*

54. Schleiden.

1849* Principles of Scientific Botany. Eng. Transi, by E.

\

Lankester. London. 1849-

55- Schaffner, J. H.

1909. The Nature and Development of Sex in Plants. Proc.

Ohio. Acad. Sei. 5 : 327 - 350.. 1909.

56. Shattuck, G. H.

1905. A Morphological Study of Ulmus americana. Bot. Gaz. 40:

209 - 223. 1905.

57* Sprecher, Andreas.

1913* Recherches sur la Variabilité des Sexes chez Cannabis

sativa L. et Rurnex acetosa L. Ann. Sei. Nat. 17:

255 - 352. 1913«

58. Stevens, N. M.

1906. Studies in Spermatogenesis. II. Carnegie Inst.,Wash­

ington. Pub. 56. 1906.

 75

59» Strasburger, E.

1894. The Periodic Re duction of the Number of Chrmomsoraes

in the Jtife History of Living Organisms. Annals of

Botany 8: 281 - 316. 1894.

60 . ----------------

1909. Zeitpunkt der Bestimmung des Geschlechts,Apogamie,

Parthenogenesis U. Reductionsteilung. Jena. 1909.

6l. --------------------

1909 - 10. Sexuelle und apogame F o rt p flanzung bei Urticaceen.

Jahrb. Wiss. 'Bot. 38. 1909 - 10.

62 . , -------

1910. üeber geschlechbestummende Ursachen. Jahrb. Wiss. Bot.

48: 427 - 520. 1910.

62a. Tischler, 3.

1906. üeber die Enturicklung des Pollens und der Tapeten-

zellen bei Ribes-Hybriden. Jahrb. Wiss. Bot. 42:

545 - 579. 1906.

63. Treub, M.

1891. Sur les Casuarinees et leur place dans le systeme

naturel. Ann. Jard. Bot. Buitenzorg 10: 145 ~ 231.

1691.
64. Webber, H. J. and Swingle, K. T.

1897. Hybrids and their Utilization in Plant Breeding. Yr.

Bk. U. S. Dept. Ag. 1897.

.65. Winkler, H.

Uber Parthenogenesis und Apogamie in Pflanzenreiche.

Prog. Rei. Bot. L.otsy.

 76

66. Wullff, E.

1909. Ueber Pollensterilitat bei Potentilla . Oesterr.'pot.

Ztschr. 59: 384 - 595. 475 - 423. 1909.

67. Wylder.

1844. Zur Kenntniss der Inflorescence von Ca n n a b i s , HueuIus,

Urtica and Parietaria. Flora. 1844.

68. Wilson, E. B.

1909. Recent Researches on the Determination and Heredity

of sex. Science II. 29: 53* 1909.

69. Zinger, N.

1898. Beitrage zur Kenntniss der weiblichen Bluthen und

Inflorescenzen bei Cannabineen. Flora. 85: 189 -

253. 1898.
 77

EXPLANATION OP PLATES.

All figures were drawn with Spencer camera lucida. Figures

24 and 25 were made with Bausch and Lomb 1 1/2 objective and Zeiss

ocular 4; figures 10 and 11 with Bausch and Lomb 5/4 objective and

Zeiss ocular 4; figures 1 - 9 , 1 2 - 2 3 with Spencer 2/3 objective

and Spencer ocular 8; figures 26 - 33, 36 - 43, 92 with Spencer 1/6

objective and Spencer ocular 8; figures 34, 35, 44 - 54, 90, 93 -

101 with Bausch and Lomb 1 1/2 objective and Spencer ocular 8. The

approximate magnification in diameters is as follows: figures 24 and

25, x 80; figure 10 and 11, x 120; figures 1 - 9 , 12 - 23, x 168;

figures 26 - 33, 36 - 43, 92, x 825; figures 34, 35, 44 - 54, 90,

93 - 101, x 1725.
 78

ABBREVIATIONS USED.

a. sp. e. - - - - - - - - - active sporogenous cell.

an. - - - - - - - -- - - - - — - anther

ar> ------ . arehesporium.

at. -- antipodals.

c. - - - - - - - carpel.

cyst. - - - - - - - - - - - - - cystolith.

d. sp. c. - - - - - - - - - degenerating sporogenous cell.

e. - - - - - - - egg.

smb. s.-- - - - - - - - - - - embryo sac.

end. - - - - — -- - - - - - - - endotheciura.

ep. - - - - - -- ------ - epidermis.

fil. - - - - - - — filament.

fl. b. - - - - - - - - - - - - floral bract.

i. int.-- - - - - - - - - - - inner integument.

i. mid. 1. - - - - - - - - inner middle layer.

mid. 1. — -- - - - - - - - middle layer.

n . - - - - - ------- -- nucellus.

o. int. - - - - - - - — — outer integument.

o. mid. 1. ------ outer middle layer.

o v . - - - - - - - - - - - - - - - ovule.
p. - - - - - - - - - - - - - - - - perianth.

par, - - - - - -- - - ----- parietal tissue.

 79

p, ar. c. - - - - - - - p rimary archesporial cell.

pol. gr. - - - - - -- pollen grain.

p. sp. c. - - - - - - - primary sporogenous cell.

s. — sepal.

sng. — synergid.

sp. c.-- -- - - - - - - sporogenous cell.

sp. m. c. - - - - - - - spore mother cell.

st. - - - - - - - - - - - stoma.

stg. - - - - - - - - - - - stigma.

stm. - - - - - - - - - - - stamen.

tet. - - - - - - - - - - - tetrad.

trc. - - - - - - - - - - - trichome.
 PLATE I.

Staminate Flower.

Figure 1. Longitudinal section through beginning of flower.

Figure 2. Appearance of sepals.

Figure 3. Appearance of stamens.

Figures 4,5. Transverse and longitudinal sections of later stage.

Figures 6,7. Longitudinal and transverse sections of still later stage.

P la t e I
 PLATE II.

Staminate Flower.

Figure 8. Longitudinal section of advanced stage showing enlarg-

ment of anther.

Figure 9. Transverse section of same stage.

p la te II
 PLATE III.

Staminate Flower.

Figure 10. Longitudinal section of mature flower bud showing erect

stamens roofed over by the imbricated sepals.

j
plateIII
 PLATE IV.

Staminate Flower.

Figure 11. Transverse section of mature flower bud.

p l a te I V
 PLATE V.

Pistillate Flower in Bract.

Figure 12. Longitudinal section through papilla which is to give

rise to the flower and bract.

Figure 13. Appearance of floral bract.

Figure 14. Appearance of posterior perianth papilla.

Figure 15. Appearance of anterior perianth papilla.

Figure 16. Later stage showing appearance of carpels.

Figure 17. Transverse section of flower and bract at stags

shown in figure 15.

Figure 18. Beginning of ovule.

Figure 19- Later stage showing association of ovule with posterior

ovarian wall.

Figure 20. Transverse section of stage shown in figure 19.

p la te V
 PLATE VI.

Pistillate Flower In Bract.

Figure 21. Longitudinal section through later stage of flower showing

the closing of the ovarian cavity and the beginning of the

stigmas.

Figure 22. Later stage showing beginning of inner integument of ovule.

Figure 23. Beginning of outer integument of ovule.

*
plateVI
 PLATE V I I .

Pistillate Flower in Bract.

Figure 24. Longitudinal section of nearly mature flower showing

brush-like stigma and final pendulous position of ovule.

p la t e V II
 PLATE V I I I .

Pistillate Flower.

Figure 25. Longitudinal section through ovary after fertilization.

J
p la t e V III
 PLATE IX.

, Megasporangium.

Figure 26. Longitudinal section through young ovule showing primary

archesporial cell.

Figure 27. Slightly later stage showing primary sporogenous cell

divided.

Figure 28. Later development o? megasporangium showing several-celled

archesporium.
p la t e IX
 PLATE X.

Megasporangium.

Figure 29* Section of ovule showing group of sporogenous cells.

p la te X
 PLATE XI.

Megasporangium.

Figure 30« Section through ovule showing one sporogenous cell b e ­

coming dominant.
p la t e X I
 PLATE X I I .

Megasporangium.

Figure 31* Section through ovule showing one sporogenous cell greatly

enlarged and others disappearing.

i
plate XII
 PLATE X I I I .

Megasporangium.

Figure 32. Section through ovule showing embryo- sac initial.

plateXII
 PLATE XIV.

Megasporangium.
*

Figure 33* Section through tip of ovule at mature stage of e mbryo-

sac showing integuments fused over beaked nucellus at mic-

ropylar extremity.
p la te X IV
 PLATE XV

Female Gametophyte.

Figure 34. Four celled embryo-sac.

Figure 35* Mature embryo-sac.

p la te X V
 PLATE XVI.

Microsporangium.

Figure 36. Trnasverse section through portion of stamen showing

young archesporium and parietal tissue.

Figure 37. Later stage of same.

Figure 38. Section of m icrosporangium'showing differentiation of

tapetum.

Figure 39* Later stage of same.

Figure 40. Section of sporangium at synaptic stage of mother cell.

i
p la te X V I
 PLATE X V I I.

Microsporangium.

Figure 41. Section through sporangium showing beginning of d e gen­

eration of inner "middle layer".

Figure 42. Degeneration of outer "middle layer".

Figure 43. Mature anther w a l l ’ showing endotheciura with thickening

bands.
p la te X V II
 PLATE X V I I I .

Tapetal Cells of Microsporangium.

Figure 44. Early uninucleate stage of tapetum.

Figure 45. Nuclei dividing at synapsis stage of spore mother cell.

Figure 46. Uninucleate cell at dyad stage of spore mother cell.

Figure 47* Three nucleate cell at same stage.

Figures 48,49. Tapetal cells at tetrad stage of mother cells show­

ing peculiarly shaped fusion nuclei.

Figures 50,51. Tapetal cells at microspore stage showing beginning

I
of degeneration.

Figures 52,53* Tapetal cells almost completely degenerated.

p la te X V III
 PLATE XIX.

Pollen Mother Cell.

Figure 54. Nuclei in a resting condition.

Figures 55,56,57. Nuclei showing reticulum and chromatin becoming

more distinct.

Figure 58. Nuclei with definite spirem before synapsis.

Figure 59. Nuclei in synapsis.

plateXIX
 PLATE XX.

Pollen Mother Cells.

Figure 60. Nucleus at spirem stage.

Figure 61. Segmentation of spirem showing telosynaptic formation

of chromosomes.

Figure 62. Nucleus with dyad chromosomes.

Figure 63. Formation of multipolar spindle.

Figure 64. Bipolar spindle with chromosomes grouped in equatorial

region.

Figure 65. Pulling apart of chromosomes.

Figure 66. Massing of chromosomes at spindle poles.

Figures 67,68,69. Reconstruction of daughter nuclei.

p la t e X X
 PLATE XXI.

Pollen Mother Cells.

F igure 70. Daughter nuclei entering into homoeotypic division.

Figure 71. Appearance of bipolar spindles.

Figure 72. Pulling apart of chromosomes.

F i g u r e 73- C h r o m o s o m e s m a s s e d at p o l e s of s p in d l e s .

Figure 74. R ec o n struction of grand-daughter nuclei.

F i g u r e 75* Y o u n g t e t r a d showing prominent special wall.

Figure 76. Nuclei of tetrad walled apart. Development of ridge-like

t hickenings on inner side of wall.

Figure 77. Later stage of tetrad showing outer w a ll beginning to

disorganize.

Figure 78. Tetrad whose wall has disappeared. Microspores breaking

apart.

F i g u r e 79. S a m e s t a g e s h o w i n g tone cell of t e t r a d d e g e n e r at i n g .

Figure 80. Whole tetrad degenerating.

plateXXI
 PLATE X XI I.

Pollen.

Figure 81. Young microspore after separation from tetrad.

Figure 82. Microspore at same stage slightly shrunken.

Figure 83. Microspore at same stage smaller and shrunken.

Figure 84. Microspore nearly completely broken down.

Figure 85. Later stage of microspore.

Figures 86,87* Microspores at same stage showing different degrees

of sterilization.

Figures 88. Normal mature pollen grains.

Figures 89,90. Pollen grains,shrunken and sterile.

Figure 91* Section through point of pollen tube exit ( after Briosi

and Tognini),
p la te X X II
 PLATE X X I I I .

Diagram showing the amount of sterilization that occurs in

the sporogenous tissue of the microsporangium.

p la te X X III
 PLATE XXIV.

Figure 92. Section of mature leaf.

plate X X IV
 PLATE XXV.

Figures 93,94. Beginning of development of glandular hair.

Figures 95,96. Transverse and longitudinal sections of later stage

of development.

Figures 97,98. Sections through 4-colled glandular hair.

Figure 99* Transverse section of multicellular hair.

Figure 100. Longitudinal section of multicellular glandular hair.

Figure 101. Unicellular hair.

p l a t e X X V
 APPENDIX.

Economic uses and Vegetative Structures.

A conspicuous development of bast is noted in the stem of

Cannabis. The fibers are very long often reaching a length of 22

millimeters and are multinucleate. It is for the bast of the stem

that the plant is cultivated so widely.

Cannabis sativa L. is the commercial hemp which is cultivated,

in many parts of the world for its products. Although a native of

Asia it spread westward throughout Europe and southward through the

Indian psnninsula. The cheif continental hemp producing countres

are Italy,Russia and France and it is also grown in several parts

of Canada and the United States and India. The plant is cultivated

in different countries for three products - the bast fiber of the

stem,the resinous secretion developed in hot countries u p on its

leaves and flowering heads,and its oily seeds.

The fiber of hemp has long been used extensively in the

rope industry and is also used in the production of yarns,the m a n ­

ufacture of sail cloth,sheeting,sacking,etc. The coarser quality is

made into ropes and similar material and the finer into cloth. The

plant is cultivated for its fiber in almost all European countries

and most extensively in European Russia which is the chief hemp-

exporting district. The finest hemp fiber is grown in the province

of Piedmont,Italy where the climate and soil are such as to permit

 ii

a rapid growth of the plant at first,thus resulting' in the f or m a ­

tion of long fibers.

The preparation of the fiber for the market involves several

processes. The plants are pulled or cut,then bound into sheaths,

dried,and stacked. A' process of combing removes the seeds and the

stalks are tied into bundles. The retting or rotting of the stems

is accomplished by immersing the bundles in water where they remain

from 10 days to 2 weeks,until the fiber is separated from the woody

core. The stalks are then dried and the fibers are separated and

cleaned by hand or machinery .after which the material is ready for

the market. The average yield of hemp fiber per acre is about 1000

pounds.
The ordinary unicellular superficial hairs which are common

to Dicotyledones are very conspicuous in Cannabis where they often

reach a great length (fig. 101). These are common in all parts of

the plant especially upon the leaves,bracts and stem. The hairs are

simply prolongations of epidermal cells and appear dead and devoid

of cell contents at maturity. Boodle and Fritsch (7) report that

they are common in the Urticaceae.

Glandular hairs which are mul tice>ll ular, epidermal outgrowths

with a head and more or less definite stalk,are noted more particularly

on the floral parts. When mature the cells of the head vary from

one,two or four cells (figs. 97,98) to many (figs.99,100). The

resinous secretion for which the species is cultivated in the East

is secreted from these glands especially from those of the p i s t il ­

late inflorescence.
 iii

Cystoliths which are aggregates if' cal ci u m carbonate are

common on the upper side of the leaf. They are contained in t r i ­

chomas the basal portion of which is strongly swollen (fig.92).

Longer hairs on the lower side of the leaf also contain smaller

cystoliths. Of these the stalk of the cystolith may be said to be

formed by the body of the hair which has become solid owing to

calcification and silification.

The medicinal and intoxicating properties of hemp have been

known in the oriental countries since an early time. In Persia,

Northern India,Arabia,and parts of Africa and Brazil it is grown

partly and often mainly for the resin which is produced by the

glands upon the leaves and the flowering heads. The plant as a

drug or intoxicant for smoking and chewing occurs in three forms -

"bhang" which is the dried leaves and small stalks,"ganja" or the

flowering heads of the pistillate individual and "charas",the resin

collected as it exudes from the plant. The com p os i ti o n of the drug

is obscure but it acts typically as an intoxicant resembling alcohol

in many features of its action but varying widely in its effects

upon different individuals and .races. The early symptoms of its use

are highly pleasurable and for these it is largely consumed in the

East. The drug is employed extensively for medicinal use in all

parts of the world.

The seeds of hemp are combed from the pistillate plants

when the fibers are prepared for the market and sola for various

uses. The finest ones are kept for sowing,a large quantity is sold
 iv

for cage birds,and the remainder is sent to the oil mills to ’os

crushed. From the extracted oil of the seeds soap is manufactured

while the solid remains or oil cake is a valuable food for cattle.

Cannabis indica,a special variety of Cannabis sativa L . ,

differs from the American grown Cannabis in several particulars

owing to the difference of climate and soil. The plant grows to a

much larger size and it is the. resin of the Indian form that has

the medicinal and narcotic properties before mentioned. Much c o m ­

ment on the subject of the American Cannabis has been recently

caused by the increased cost of the Indian drug and the question

as to whether or not an active variety can be successfully cultivated

in this country. A study of the American grown Cannabis in c o m p a r ­

ison with samples from various other sources was carried on by C.

R. Eckler and F. A. Miller (22). Hemp seeds from different l o c a li ­

ties and of different varieties were grown and their medicinal

effect tested upon animals. From these experiments it was shown

that the soil,climate and geographical location have a decided in­

fluence upon the activity of the American and the Indian Cannabis,

since repeated plantings from selected seeds of both forms did not

yield a product testing over 65 per cent as active as the Indian

grown drug. Consequently if the American Cannabis is made officinal

much difficulty will be experienced in obtaining active lots which

will compare favorably with the Indian drug.

 v

Teratologies! Phenomena.

Abnormalities in Cannabis have been observed to coour c o m mo n ­

ly in certain localities ana their presence is taken to bear upon

the origin of the dioecious character of the species. The most

striking and common of these have been reported by various writers -

the most recent accounts being made by Briosi and Tognini (9) ana

later by Prain (50)*

In the cultivated hemp fields of Bengal where all the s t a m ­

inate shoots are pruned away from the pistillate plants there is

always present a considerable number of pistillate plants which are

termed"Khasia". These individuals are marked by a total absence

of narcotic resin for which the crop os grown and although they

flower as copiously as the normal pistillate plants they never set

seed. The flowers are similar in appearance to the normal flowers

except that the glandular hairs are not discernabls from the non-

giandular owing to the absence of resinous material. Althoug these

plants are recognizable at a fairly early stage by their robust

growth and darker color they are left standing in the field because

their differentiation is not effected sufficiently early to admit

of their being replaced. This condition is reported by Prain (50)

and Kerr (40) from the locality of Bengal but is not known in other

regions. Prain (50) examined carefully hundreds of these plants and

found the ovule enclosed within a perfect ovary but without any

trace of fertilization. The author suggests that this may be due to

impregnation by imperfect pollen or the result of an over stimulation

 vi

of the vegetative functions since the condition is found e x clusive­

ly in cultivated fields.

Monoecious conditions of hemp have been reported as quite

common in certain regions. Prain (pOl states that this condition

is not unusual in the Indian Cannabis and describes several d i s ­

tinct monoecious flowers. Plants of staminate habit and appearance

have staminate flowers on the lower portion of each branch and

only pistillate on the upper part. The converse condition is also

found where plants of the staminate type have the lower one-third

to one-half of each floral branch bearing nothing but pistillate

flowers and the upper portion only staminate. The last condition is

so common in India that it has received the special name of "Sheoria"

hemp. Pistillate plants are often observed with one or mors branches

bearing only staminate cymes which in cultivation are carefully

removed. The "Moria" hemp of India is a common monoecious c o n d i ­

tion in which the axillary spikelet occuring between the basal

pair.of pistillate flowers develops stamens.