Digitized by tlie Internet Arcliive
in 2007 witli funding from
IVIicrosoft Corporation
littp://www.arcli ive.org/details/contributiontoouOOIubbuoft
7'
SEEDLINGS
CONTRIBUTION TO OUR KNOWLEDGE
OF
SEEDLINGS
^A.u ^ ^^
THE EIGHT HON. LORD AVEBURY
P.O., F.R.S., D.C.L., lyL.D., D.L.
WITH 282 FIGURES IN TEXT
Popular Edition
LONDON KEGAN PAUL, TRENCH, TRUBNER & CO. Ltd.
DRYDBN HOUSE, GERHARD STREET. W.
1907
{^he rights of translation and of reproduction are reserved)
PEEFACE
The germination of plants is certainly not the least interesting portion of their life-history, but it has not as yet attracted the attention it deserves. The forms of cotyledons, and the fact that they differ so much from the subsequent leaves, had of course been alluded to more or less fully in botanical works, but no explanation had been offered, and Klebs ' in a recent memoir expressly states that the problem is still an enigma.
Under these circumstances it seemed to me that the subject was very promising, and it was evident' that Kew would afford unrivalled opportunities for such an investigation. I applied, therefore, to Sir Joseph Hookek, and I cannot too cordially thank him, as well as his successor, Mr. Thiselton Dyer, and indeed the whole of the staff, for the facilities they have offered me, and for their valuable assistance in many ways.
I have also to thank Mr. Cahruthers and the Trustees of the British Museum, Mr. Lynth and the
' Beitr. zur Morphologie u. Biologie der Keimung: 'Im Allge- meinen sind una diese Verschiedenheiten in den Blattformen hinsichtlich ihrer biologischen Bedeutung fiir die Pflanze einKathsel.' UniersJich. Bot. Instil, zu Tubingen, 1884.
vi ON SEEDLINGS
authorities of the Cambridge Botanic Gardens, and other friends, especially Mr. Hanbury, for the gift or loan of many rare or interesting specimens. By degrees a large store of material accumulated, which I made use of for several papers which the Linnean Society has done me the honour of publishing in their Journal. I thought, however, that it would be well to publish descriptions and figures of the more interesting species, especially as many of them are not often grown from seed, and are therefore not easily procurable.
The full particulars are given in a larger work ' A Contribution to our knowledge of Seedlings,' published in two volumes by Messrs. Kegan Paul & Co. I have here given those parts which seemed of most general interest, and must refer those who wish for more detailed observations to the above-mentioned work.
The seedlings were drawn in most cases either by Mr. Henry, or by my assistant, Mr. Fraser, to whose skill and ability I am greatly indebted. In the classi- fication, &c., I have followed Bentham and Hooker's great work, the * Genera Plantarum.' My time has latterly been so much occupied with other matters that Mr. Rendle has been good enough to see the book through the press. Sir Joseph Hooker has also most kindly looked through the proofs, and made many valuable suggestions, for which I beg to offer him my very warm thanks.
High Elms, Down, Kent.
A CONTKIBUTION
TO OUR
KNOWLEDGE OF SEEDLINGS
Forms of Leaves.
I HAVE elsewhere ' called attention to the forms of leaves, and discussed the causes to which we may ascribe the endless differences which they present. Vertical leaves, for instance, are generally long and narrow, horizontal ones have a tendency towards width, which brings the centre of gravity nearer to the point of sup- port. "Wide leaves, again, are sometimes heart-shaped, sometimes palmate. The former shape is obviously that' which would arise if a linear leaf were gradually widened at the base ; and I have pointed out that in many species with palmate leaves — for instance, species of Passiflora, Cephalandra, Hibiscus, &c. — the first, or few first, leaves are entire and more or less cordate.
' Flowers, Fruits, and Leaves (Nature Series), Macmillan & Co. See also various papers in the Journal of the Linncan Society.
B
2 ON SEEDLINGS
The cordate form, then, appears to be the early, the palmate a later form. But how has the palmate form arisen ?
The origin is perhaps connected with the manner in which the leaves are folded up, more or less like a fan, into the bud, so as to save space.
Another advantage perhaps is that in cordate leaves with veins following the curvature of the leaf — as, for
Fig. 1. — Leaf of Tamus, to show the curved course of the veins.
Fig. 2. — Leaf of Sycamore (Acer), to show the straight course of the
instance, in Tamus (fig. 1) — the vascular bundles pursue necessarily a curved course ; while in palmate leaves, as in Acer (fig. 2), the veins are straight ; and it is clearly an advantage that the main channels which convey the nutritive fluid should hold a direct course. In such cases the leaf naturally assumes the lobed form with the vein running to the point of each lobe. There has indeed been some question whether the path of the sap lies mainly in the cell-walls or in the cell-cavities ; but
FORMS OF LEAVES 3
the evidence seems to point strongly to the latter view.^ The tracheids of, say, the Yew ' are at least seventy or eighty times as long as they are broad, so that in travelling transversely the length of a single tracheid the water-current has to traverse seventy cell- walls instead of one.' ^
In reply, then, to the question why some plants should have cordate leaves with curved veins, while others have palmate leaves with straight ones, I suggested that the first retains the old form of leaf, while the latter has assumed one which presents certain advantages.
We can also see a relation between the shape and size of the leaf and the length of the internode. For instance, in the common Lime the leaf-stalks are arranged at an angle of about 40° with the branch, and the upper surfaces of the leaves are in the Fame plane with it. The result is, they are admirably adapted to secure the maximum of light and air.
' In the Beech (fig. 3) the general plane of the leaves corresponds with that of the branch, but the leaves them- selves are ovate in form, and smaller, being only from two to three inches in length. On the other hand, the dis- tance between the nodes is also smaller, being, say, one and a quarter inch against something less than two inches. The diminution in length of the internode is
' See, for instance, Darwin and Phillips, ' On the Transpiration Stream in Cut Branches,' Proc. Canhridge Phil. Soe. vol. v. * Zoc. nt. p. 364.
b2
4 ON SEEDLINGS
not, indeed, exactly in proportion to that of the leaf, but, on the other hand, the leaf does not make so wide an angle with the stem. To this position is probably- due the difference of form. The outline of the basal half of the leaf fits neatly to the branch, that of the upper half follows the edge of the leaf beyond, and the
Fio. 8. — Beech. Fig. 4. — Spanish Chestnut.
form of the inner edge being thus determined, decides the outer one also.' '
In the Spanish Chestnut (fig, 4) the terminal branches are stouter in proportion, and can therefore carry a greater leaf-surface. The width of the leaf, however, being determined by the distance between the intemodes, the leaf is, so to say. compelled to draw itself out, and hence the peculiar sword-like shape. • Lubbock, Florvert, Fruits, atid Leaves.
THE OAK LEAF 5
We are so accustomed to the form of the oak leaf that it does not strike us as anything peculiar, and comparatively few of us, perhaps, have ever asked why it should be as it is. And yet it is peculiar, unlike that of any of our forest trees, and those of the evergreen oaks so abundant in hotter countries.
In botanical phraseology it is ' deciduous, oblong- oblanceolate, or oblong-elliptical, sinuated, with blunt lobes extending not more than halfway down to the mid- rib.' The sinus between the lobes is generally rounded off at the bottom.
Again, though I have not found this mentioned in the botanical works which I have consulted, they are rarely symmetrical, the lobes of the two sides not corre- sponding.
The three points, then, which give the oak leaf its peculiar form are : —
1. The deep rounded sinuses.
2. The want of symmetry of the two sides.
3. The obovate or oblanceolate outline.
The explanation which I would suggest is as follows. The leaves of the evergreen oak are entire, and small in comparison with those of the English oak. During the winter and early spring they are protected by a series of brown scales, inside which they lie, and with which they form the buds which are so familiar to us, and which are both small and short in proportion to the size of the leaves themselves.
In cooler and moister regions, on the contrary, there
6 ON SEEDLINGS
is^ as we know, a tendency for leaves to become larger and deciduous. I will not now enter into the reasons for this, but the fact will not probably be denied. These influences do not, however, affect the outer scales, which remain as before without any increase of size. But as the leaves have increased in size, and the buds have not, the leaves can no longer retain their original arrange- ment in the bud. If, for instance, we compare the buds of the oak and of the beech we see that while the leaf of the oak is longer than that of the beech, the bud of the oak is, on the contrary, shorter than that of the beech.
Under these circumstances, what must happen? The leaf grows and becomes longer than the bud ; it is therefore necessarily bent into a curve. But an entire leaf, if thus thrown into a curve, would necessarily fall into folds, the number being determined by the number of ribs or veins. For such folds, however, there would be no room within the narrow limits of a bud, or rather, perhaps, they would be inconvenient because they would leave more or less empty spaces.
This may be rendered more clear by taking a piece of cloth or paper, folding it up, and then throwing it into a curve. It will then necessarily fall into one or more folds. If it were strengthened, as an oak leaf is, by three or four side ribs, there would be a fold between each two ribs. As a matter of fact, however, from the absence oi space the membrane where the fold would be is not actually developed. We may imitate this by removing them. If this be done, the result will be the
LEAF OF TULIP-TREE 7
formation of sinuses, rounded at the base, closely resem- bling those so characteristic of the oak leaf. These sinuses are due, then, as I believe, to the curvature of the leaf, owing to the shortness of the bud in comparison with the length of the leaf.
The young leaf is not only curved, it is wrapped round the interior leaves. The result of this is that one side of the leaf is folded within the other ; the one therefore has more space than the other. The two sides of the leaf are in fact differently situated, and this, I believe, accounts for the second point — namely, the want of symmetry.
The obovate form is an advantage in consequence of the way the leaves diverge from the stalk.
I think, then, that the explanation I have suggested accounts for all these points, and beautifully explains the peculiar torm assumed by the leaf.
The leaves of the Tvi[ip-tree'(Liriodendron) have long attracted attention from the peculiarity of their form.
They are saddle-shaped, abruptly truncate at the end, or, in the words of Bentham and Hooker, ' sinuato- 4-1 oba.' I have often wondered what could be the purpose or the advantage to the tree of this remarkable shape. One idea which occurred to me was that the difference of form might enable insects to perceive the tree at some distance, just as the colours of flowers are an advantage in rendering them more conspicuous. I then looked closely to see whether the peculiar forms could in any way be explained by the position of the
8
ON SEEDLINGS
leaves on the tree. I believe, however, that the cause is of a different nature, and has reference to the peculiar character of the bud.
Each youtig leaf is, as in the family MagnolieaB generally, originally enclosed in, and sheltered by, the stipules of its predecessor. These are in Liriodendron oval, or in form resembling a shallow dish or spoon, so that, when placed face to face, they form a hollow almond- shaped body. The first of these neat little boxes which I opened showed the young leaf in the stage shown in fig. 5, p being the petiole, I the lamina of the leaf, and
Fig. 5. — Liriodendron. Position of young leaf in bud. x 12.
Fig. 6. — Liriodendron. Very young leaf and stipule. x20.
Fig. 7. — Liriodendron. Very young leaf and stipule. Second stage, x 20.
8 the stipule. This threw me off the scent, as it seemed to give no clue whatever to the peculiar form of the leaf. Eventually, however, on examining younger buds, I found what I believe to be the true
LEAF OF TULIP-TREE
explanation. Within the stipules of such a bud as that represented in fig, 5 are contained several younger buds, one within the other. The youngest stage which need here be mentioned is represented in fig. 6. The petiole is short and thick, the lamina is widest in the middle, and tapers regularly towards both base and apex; the sti- pules are in the form of a hemisphere. Gradually the stipules become more oval (fig. 7), assuming an almond-like form, some- what thicker in the middle, tapering away to the sides, and more gradually to the apex. The petiole has also elong- ated, and the lamina of the leaf is more abruptly turned downwards ; so that the petiole passes up one side of the stipule, while the midrib passes round the
Fig. 8. — Liriodendron. Young leaf and stipule in a rather more ad- vanced stage. /< 20.
Fig. 9. — Almond-shaped body on and under a glass.
tip of the stipule and down again on the other side. The lamina itself has considerably enlarged, is conduplicate
10
ON SEEDLINGS
or folded on itself, and lies on one side of the stipule, between its own stipule and one of those by which it is enclosed. It has also considerably modified its form, and for the following reason. The young stipule at its central part touches the surrounding older stipule, so that the young leaf cannot find room between them ; while, on the other hand, at the sides and towards the end of the young stipule there is plenty of space for growth. This, I would suggest, accounts for the particular shape it assumes. Suppose, for instance, we lay an almond on a table, and place a piece of glass over it, the glass will touch the almond on a ^, — .^ surface, a (fig. 9). In the case
of Liriodendron, from the form of the bud the surface of con- tact occupies not only the oval (fig. 1 0, a), but also the space lying below the line h. The young leaf therefore cannot retain its original form of lamina, because it is stopped by want of space along the line a, which it will therefore follow, as shown in fig. 10. Moreover, the thickened rib or vein (figs. 5 and 8) will be arrested sooner than the thin lamina. When stopped by meeting the stipule, it bifurcates ; and this perhaps is one reason why the leaf permanently retains the form which it is thus compelled to assume. The lamina grows for a while
Fig. 10. — Diagmm showing arrangement of the young leaf.
CONIFER LEAVES 11
somewhat more rapidly than the stipule, then the stipule more rapidly than the lamina ; and, lastly, the growth of the stipule is arrested while the leaf attains a considerable size. The terminal portion of the young leaf seems narrower in fig. 8 than it is in the leaf ; but it must be remembered that it is to some extent curved round the inner stipule, so that in the figure it is some- what foreshortened.
It might be suggested that the form of the leaf determines the bud. But in fact the form of the bud is not that of the leaf; the leaf follows, not the form of the bud, but that of the vacant space left in the bud.
In the Firs and Pines the leaves differ little in form, but some are much longer than others. Why is this ? I puzzled over it for some time. The governing con- sideration is, I believe, to secure a suitable amount of leaf surface. They are all evergreen, but in some cases the leaves of one year fall soon after those of the next are developed. In others they are longer lived. In the Scotch Pine they last three or even four years ; in the Norway Fir eight to ten ; in the Pinsapo even as many as sixteen to eighteen. No\m, speaking roughly, and having regard also to the stoutness of the terminal shoots, it will be found that the shorter leaves have longer lives, and vice versa.
In my book on ' Flowers, Fruits, and Leaves,' I have also given some account of the causes which have deter- mined the form and structure of seeds and fruits.
12
ON SEEDLINGS
Fig. 11. — Seedling of Mustard (Brassica nigra), x 3.
Fig. 12. — Seedling of Cress {Lepidium sativum), x 3.
Fig. 13.— Seedling of Pink {Dian- Fig. 14. — Seedling of Mouse-ear thus Gary ojphyllus). Nat. size. Chickweed (Ceras^iMw). Half nat. size.
COTYLEDONS
13
These considerations naturally led me to the study of cotyledons.
No one who has ever looked at seedlings can fail to have been struck by the contrast which the cotyledons afford, not only with the final leaves, but even with those by which they are immediately followed.
Let us then take certain plants (especially, as far as possible, the commonest and most familiar), and see what light can be thrown on the varied forms which their seed- lings present. Look, for instance, at the familiar Mustard and Cress ; the first (fig. 11) has kidney-shaped coty- ledons, one of them rather larger than the other : while the Cress (Lepidium sativum) (fig. 12), on the other hand, has the cotyledons divided into three lobes. The Pink (fig. 13) has broad cotyledons, the Mouse-ear Chick weed (Gerastium) (fig. 14) narrow ones ; those of the Beech (fig. 15) are fan-shaped in outline ; those of the Sycamore (fig. 16) shaped almost like a knife-blade; those of Eschscholtzia (fig. 1 7) divided like a hay-fork ; those of the Bean or Acorn tliick and fleshy.
Mustard and Cress were the delight and wonder of
Fig. 15.— Seedling of Beech {Fagus sylvatica). Half nat. size.
14
ON SEEDLINGS
our childhood ; but at that time it never occurred, to me at least, to ask why they were formed as they are,
Fig. 16. — Seedling of Acer Pseudo-Platanus. Half nat. size. In this and some other figures, one (or more) of the leaves is given
in outline only.
Fig. 17. — Seedling of Eschscholtzia californica. -Nat. size.
and why they differed so much. So they grew, and beyond that it did not occur to me, nor I think to most, that it was possible to inquire.
I now propose, how- ever, to suggest reasons which may account for some, at any rate, of these differences.
In previous mem oirs I have discussed the causes which regulate
V
f
OVULE AND SEED 15
the forms of seeds, and will not now, therefore, enter into them. I may, however, observe that the shape of the cotyledons seems to have little, if any, influence on that of the seed.
Ovule and Seed.
The seed of a flowering-plant contains a more or less highly developed embryo, which, after a certain time, will, under favourable conditions, resume growth, emerging from the seed-coat to produce the young plant or seedling. In the section of flowering-plants known as Gymnosperms, and including the Conifers and Cycads, the seeds are borne naked on an open scale, while in the far larger section of the Angiosperms they are protected till ready for dispersion in the closed cavity of the Fruit.
The seed is the result of the fertilisation of the ovule by the pollen. The effect of fertilisation, how- ever, extends in the Angiosperms beyond the ovules to the ovary in which they are contained, causing often a great increase in size, as well as important struc- tural changes. The ovary thus becomes the fruit as the ovule has become the seed.
Sometimes the effect extends still farther to other parts of the flower, which thus persist, usually with increased size or change of structure, and form part of what is then termed a pseudocarp or false fruit. In the apple, for instance, the edible portion consists of
16 ON SEEDLINGS
the greatly developed floral-receptacle, wliich includes the ovary as its core. The ovules are borne usually on some definite part of the ovary wall or walls known as the placenta. They consist of an internal portion, the nticelhis, one cell of which grows at the expense of the rest to form an embryo-sac^ which again contains the egg-cell or oosphere, and one or two integuments entirely surrounding the nucellus except at the apex, where a small aperture, known as the micropyle, is left. In the process of fertilisation the pollen- tube passes down the micropyle to the oosphere which, enclosed in the embryo-sac, approaches very close to the outer wall of the ovule at this point. After fertilisation the oosphere develops into the embryo, the rest of the embryo-sac becoming filled with a cellular tissue, the endosperm, containing a store of food for the young plant. This is, however, sometimes reabsorbed before germination, so that the embryo occupies the whole interior of the mature seed. The original substance of the ovule, the nucellus, is often completely dis- placed by the embryo-sac, and the subsequent develop- ment of endosperm in its interior. One or two layers, however, may persist and share with the integuments in the formation of the seed-coats. Occasionally, as occurs in the Nymphaeaceae, Piperaceae, and some of the Scitaminese, a considerable portion of the nucellus remains in the mature seed, sharing with the endo- sperm the function of storing reserve material, or it
POSITION OF OVULE 17
may quite take the place of the endosperm. It is known as perisperm.
There is considerable variety in the position and arrangement of the ovule and subsequently of the seed. They are attached to the placenta by a stalk or funicle, the point of connection with the funicle being called the hilani. Vascular tissue supplying nutriment runs up the funicle into the integuments, and the point reached in the latter before the nutritive vessels ramify is the chalaza.
In some cases, usually where there is only one ovule in the ov^ary or ovary-cell, as e.g. in Buckwheat or Nettle, the ovule stands erect, and the micropyle is opposite the hilum, which in this case coincides with the chalaza. The ovule is then said to be oi-thotropous ' or straight. But this, though the simplest, is the least common form. More often, as in the Bean, Chick- weed, and many others, the ovule during its growth is curved on itself, so that the apex and the micropyle are brought down very near to the base. Such ovules are known as campylotropoiis or curved.
In a third category the ovule itself is straight, but it stands as it were at right angles to the base of attachment. These are called half-aiiatropous or half- reversed.
' Figures of the various positions of the ovule are given in most text-books; see, for instance, Vines' StudeiM Text-hook of Botany, p. 437, fig. 28i; or Asa Gray's Structural Botamj, p. 278.
C
I
18 ON SEEDLINGS
Lastly, in a very large number — in fact the largest — the ovule is quite inverted on its base ; the funicle, or stalk, growing with its curvature, and forming a sort of ridge or raphe^ which is very conspicuous in the ovule, but gradually becomes less and less apparent, finally becoming merged in the seed-coat. The chalaza is thus carried right away from the hilum to the opposite end, and the whole ovule or seed is reversed, so that the true base is removed from, and the true apex brought close to, the point of union with the ovary. Such are called anatropous or reversed.
Now this seems a very curious and roundabout arrangement. It is described in all works on general Botany, but those which I have seen do not give any explanation of its object or purpose. I will for the moment omit any consideration of campylotropous and half-anatropous seeds, and confine myself to ortlio- tropous and anatropous forms.
In most orthotropous species, as, for instance, in the Buckwheat, the ovule is straight, upright, and at- tached by its base to the funicle and base of the ovary. At the free end is the micropyle, and immediately below it the egg-cell, which, when fertilised by the pollen, becomes the rudiment or embryo of the future plant. When the pollen falls on the stigma it soon pushes out a little tube which rapidly elongates, passes into the cavity of the ovary, and entering the micropyle of the ovule fertilises it.
Bat ovules thus constituted are, as I have already
POSITION OF OVULE 19
mentioned, exceptional. In a great many plants the ovule, instead of being upright and attached to the base of the ovary, is, on the contrary, attached to the summit and pendent. Now in a pendent orthotropous ovule the micropyle is turned away from the pollen-tube, and the object and effect of the ovule being reversed or ana- tropous seems to be to bring it back into a convenient position for fertilisation. So also, when there are many ovules, the result of the anatropous arrangement is again to bring the micropyle into a suitable position.
The structure and arrangement of the ovule have been the subject of various important memoirs, which, however, have been almost exclusively descriptive, Cal- mer^ indeed quotes Schleiden, who, speaking of Berberis, pointed out that among the normal anatropous ovules here and there one occurs which is orthotropous, and that these never develop into seeds. The pollen-tubes, however, would naturally follow the regular course ; and we could not, I think, deduce a general conclusion from such rare and abnormal cases. Dalmer himself does not seem to have done so ; for, after referring to the different forms of ovules, he observes that ' some- times the form of the ovule appears to be adapted so as to facilitate the entrance of the pollen-tube.' Even here, then, the explanation I venture to suggest seems referred to rather as an exceptional occurrence than as a general explanation of this remarkable arrangement.
' ' Ueber die Leitung der PoUenschlauche bei den Angiospermen,' Jer.aisclt Zeits. 1880, p. 530.
20
ON SEEDLINGS
It must be admitted that there are some cases in which the anatropy of the ovule appears at first sight rather disadvantag-eous than otherwise. Most of these can, I
believe, be explained, while in some it is possible that some plants retain, even per- haps to their present dis- advantage, an arrangement inherited from an ancestral condition in wliicli it was beneficial. I hope, however,
B e
Fio. 18. — Fraxinus excel- sior. A, samara with half removed and tlie seed pulled out, nat. size : pa, placental axis ; ao, aborted ovules ; y, funiculus ; ra, raphe ; ch, chalaza. B, longitu- dinal section of seed, x 2 : /, funiculus ; p, endosperm ; e, embryo ; ra, raphe ; ch, chalaza.
Fig. 19. — Ricinus sanguineus. A, longitudinal section of seed, nat. size. B, transverse section of seed, nat. size, c^ cotyledon ; m, caruncle ; others as in fig. 18.
to enter into this question more fully on a future occasion.
The mature seed contains the more or less differen- tiated embryo, in some cases, as in Larkspur (De/- phintum) (fig. 23), very small, consisting only of a small basal portion, the radicle, which on germination gives rise to the primary root, and one (Monocotyle- dons), two (Dicotyledons), or several (Conifers) seed- leaves or cotyledons, sheathing a microscopic bud,
CONTENTS OF SEED
21
the plumule — from which the future stem is developed. The rest of the space is occupied by the mealy, oily, fleshy, or horny endosperm. In other cases, as in the Ash (Fraxinus) (fig. 18) or Castor Oil (Ricinus) (fig. 19), the embryo is much larger, and we find a little miniature plant, generally white, sometimes green, but still embedded in endosperm. In others, aofain, the endosperm is reduced almost to a film, as in Hippophae (fig. 20), and finally, as in the
O
Fig. 20. — Hippophae rham- noides. A, langitudinal section of seed, x 4. B, transverse section of seed, x i.
Fig. 21. — Phaseolus multiflorus. Section of seed parallel and vertical to cotyledons, three- fourths nat. size: rt, a, auricles; ar, arillode ; r, radicle.
Bean (fig. 21), the embryo occupies the whole seed, and the nourishment intended for the young plant is stored up in the cotyledons themselves. Similarly, in a Walnut, the two halves of the seed are filled by the two cotyledons, and attached to them may be found the little plantlet with a delicate white radicle, and a little plumule bearing five or six minute rudiments of leaves, often just tipped with green.
The Bean and Walnut are instances of exalhuminous
22 ON SEEDLINGS
or non-endosperm ic seeds, while those containing endo- sperm (or perisperm) are termed albuminous. The seed-coat is double in cases where the ovule was pro- vided with two integuments, the outer coat, or testa, developed from the outer integument, being the stronger and often crustaceous in texture. The inner coat or tegmen is thin or soft and delicate, conforming closely to the surface of the endosperm or embryo. It is developed chiefly from the inner integument of the ovule, and is therefore wanting in ovules with only one coat.
The testa is often provided with appendages or out- growths of very diverse form and nature, but generally having reference to dissemination of the seed, as in the case of wings or tufts of hair. A more or less incomplete accessory seed-coat is sometimes formed between the time of fertilisation and ripening of the seed. This is known as an aril. Good instances are seen in the Water Lily {Nymphcea^ and the Yew. Crest-like or wart-shaped appendages developed during the same period are known as strophioles or caruncles. The Castor Oil (Bicinus^ supplies a good instance of the latter (m, fig. 19).
Number and Size of Seeds.
As regards the size of seeds, if we could imagine a state of things in which every seed grew and attained 'maturity, it would be suflScient to keep up the number of any given species existing at any time if each
NUMBER AND SIZE OF SEEDS 23
plant produced but one or two seeds during its whole life. There is, however, an enormous destruction of seeds. The great majority are eaten by animals, or fail to secure a suitable site for germination ; of those which do germinate, again, many are crowded out by their fellows. Darwin observed that out of 357 seed- lings which came up in a space of 3 feet by 2, no less than 295 were destroyed by slugs and insects. Now the greater the chance against any given seed reaching a suitable locality and attaining maturity, the larger number of seeds must the plant produce in order to maintain its numbers, and, as a general rule, the smaller will the individual seeds be. On the contrary, the greater the chance that each seed enjoys of arriving at maturity, the smaller the number of seeds that is necessary, and in such cases it is an advantage that the seeds should be large.
Hence parasitic plants generally produce a large number of very small seeds, though there are exceptions due to other considerations, as, for instance, in the Mistletoe (I believe, indeed, in all the Loranthacea3), where the seeds are carried by birds.
An interesting illustration is afforded by certain species which produce two kinds of pods, as, for instance, Cardamine clienopodifolia of Br.izil. Besides ordinary pods, which resemble those of any other Cardamine, and contain several seeds, this plant pro- duces a second sort of pod underground. Now in the ordinary pods the number of seeds increases, of course,
24 ON SEEDLINGS
the chance that some one will find a suitable place. On the other hand, the subterranean pods are sown, as it were, by the plant itself. In this case, if there were a number of seeds they would only get in one another's way, and hence, perhaps, the fact that the subterranean pods only produce one or two seeds.
In most species the seeds vary somewhat in size ; but in such cases we cannot contrast the produce of large seeds with that of smaller ones, because it might fairly be said that the former were better nourished, and inherently, perhaps, more vigorous. In Cardamine chenopodifolia, however, the seeds from the underground pods are larger than the others, and Grisebach found that they produced more vigorous seedlings.'
Some species of Vetch also have two kinds of pods.
The seeds of Atriplex hortensis are of two sizes, the small ones being black, with a rather thick, crustaceous, smooth and shining seed-coat, while the larger ones are brown and more orbicular, with a thin, membranous, smooth but not shining coat. The large brown ones germinate much more quickly than the small black ones, which would seem, under natural conditions, to be more adapted to remain in a resting condition in the ground during the winter, and germinate in spring. If such is the case they would enable the plant to exist in a colder climate than the large ones would. Of a sowing made on October 29, one black seed germinated
' ' Der Dimorphismus der Fortpfl. v. Cardamine chenopodifolia,* Gottinger Nachrichtungen, 1878.
HETEROCARPISM 25
in four days, and three da\-s later no more had come up, whereas thirty-two of the large brown seeds germinated in the seven days. On November 16, a sowing was made of thirty-five each of the black and brown seeds. In the course of six days, five of the brown seeds germinated, and only one black seed in the course of thirteen days, whereas by that time twenty-three of the brown seeds had germinated. Twenty-three days after being sown, ten of the black seeds had germinated, and twenty-six of the large brown ones.
There are, on the contrary, other considerations which may make it an advantage that the number of seeds produced by a flower should be reduced, as, for instance, in the case of the Compositae, where the agglomeration of a number of flowers into a single head, as in the J)aisy, and their consequent diminution in size, renders it an advantage that each floret should produce but one seed.
Heterocarpism.
The production of more than one kind of fruit by a plant, as in the Brazilian Cardamine just mentioned, I have proposed to call Heterocarpism.
A species of Corydalis supplies another example. Some fruits are slightly flattened, short and broad, with rounded angles, while others are elongated, hooked, and shaped like a shepherd's crook with a thickened staff.
26
ON SEEDLINGS
The hook perhaps serves for dispersion. It is possible that the alveolate surface of the seed may serve the same purpose — the depressions imprisoning, as it were, a certain quantity of air, and thus lowering the specific gravity. Again, in Caucalis nodosa (Umbellifera?) the fruits are remarkable for their dimorphic character. Those on the circumference of the umbel have the outer or most exposed mericarp furnished with a double row of spreading, curved, and hooked bristles on each of the secondary ridges, while the correspond- ing more protected and centripetal car- pel as well as all the rest of the fruits are only muricate on the secondary ridges.
The polymorphic forms of the fruits of Calendula even on the same recep- tacle are veiy remarkable. In the Marigold (C. officinalis) at least three very distinct and salient types may be noted. Two or three rows round the periphery or margin of the receptacle consist of greatly elongated curiously constructed, curved achenes, muricate on the back, often beaked at the apex and produced into a foot-like process near the base on the inner face. The middle set of fruits is furnished with broad wings involute at the margin, but neither beaked at the apex nor produced into a foot at the base. Those about the centre of the receptacle are greatly incurved, often
Pig. 22.
Calendula officinalis, x 8.
MARIGOLD 27
forming a complete ring, and have narrow wings in- volute at the margin. The seeds at first sight closely resemble a small curled grub or caterpillar. The embiyo is cylindrical in all three cases, curved, gradually tapered into the radicle, and otherwise simi- lar except that in the middle series it is not so strongly curved.
It would seem, then, as if we have here three devices for promoting dispersion — one by means of a wing, one by the presence of a hook, and one by resemblance to an insect.
During germination all three kinds behave in a similar manner, except that in the broadly winged fruits of the middle series (fig. 22) the radicle runs along the groove formed by the strongly involute wings till it reaches the apex of the fruit when it turns downwards and enters the soil, firmly fixing the fruit to the ground while the elongating hypocotyl extricates the cotyledons. In all three cases the radicle emerges at the base of the fruit which splits longitudinally by three veJves, thus facilitating the exit of the embryo. The splitting of the fruit is brought about by the energy or force of the rapidly swelling cotyledons.
The fruits of C. pluvialis are dimorphic. Those at the periphery of the receptacle are broadly obovate or suborbicular, flattened, and smooth with broad flat wings. The embryo is straight in conformity with the fruit. The other fruits are obovoid, not flattened, and muricate all over the surface. In C. hybrida the fruits
28 ON SEEDLINGS
are also dimorphic. The outer ones of the receptacle are obovate, much flattened with broad flat wings. Those of the centre of the receptacle are obovoid, slightly curved or straight, trigonous and toothed on all the three angles. The embryo is straight or nearly so in both cases. The fruits of C. gracilis are also somewhat dimorphic, judging from the specimens observed. One kind is curved so as to form a complete annul us, and it is muricate in transverse wavy ridges. In the other case the annulus is not complete and the muricate ridges on the back are nearly obsolete.
There are three or four different forms of fruit in C. algarviensis. Those on the periphery cf the re- ceptacle are oblong, straight or slightly curved, not winged but prolonged at the apex into a slender beak. The next series consists of fruits with narrow, strongly involute wings, muricate along the back with three rows of teeth, and coiled so as to form a complete annulus. In a third set of fruits the annulus is less complete, the wings are absent, and the back is merely rugose in transverse obtuse ridges. In the fourth or central series the fruits are very small, apparently imperfect, and not completely annular.
Size of Embryo.
As already mentioned, there are many cases, in fact many whole orders, in which the ripe seed is entirely occupied by the embryo ; in other cases, again, as in
SIZE OF EMBEYO
29
Delphinium (fig. 23), the embryo is very small, and examples of every intermediate stage might be given.
Where it is an advantage to the plant that germina- tion should be rapid, this of course can be more readily secured if the embryo is large. In fact, we find that species with large embryos, such, for instance, as the Cabbage or Pea, germinate much more rapidly than
Fig. 23. — Longitudinal and transverse sections of seed of Delphinium Staplnjsagria, x 12 ; E, embryo ; P, endosperm.
those, such as Umbellifers, Ranunculaceae, &c., in which the embryos are small.
On the other hand, in some cases, time is less im- portant, and here other considerations come into play. The protection of the embryo is mainly effected by the outer coverings ; but the endosperm itself contributes also, especially where it is hard, as in the Date and
30
ON SEEDLINGS
Vegetable Ivory (Phi/telepJias)^ and hence a small embryo is less liable to injury.
Arrangement of the Embryo.
In albuminous seeds the arrangement of the embryo presents no special difficulties, as the endosperm simply fills up all vacant spaces. In exalbuminous, on the
coatrary, Nature has to exercise much in- genuity, and adopts various devices to fill up the whole space.
One plan is to arrange the cotyle- dons face to face, and then roll them up in a ball. This is adopted, amongst other cases, in the Sycamore (fig. 16) and hence the long strap-like form of the cotyledons. An- other is to arrange the cotyledons face to face, and then double them up, as in the Cabbage, Mustard, and Radish (fig. 135, B-D).
In a third the cotyledons are convolute edgeways, as in Calycanthus.
Fig. 24. — Cordia subcordata. Half nat. size.
ARRANGEMENT OF EMBRYO 31
In Lepidium sativum the cotyledons are trifid (fig. 12); in Cordia they are thrown into plaits (fig. 24). In others we have still more complex folds, as in the Beech.
In such cases as the Lupine the cotyledons become so fleshy and thickened that they almost lose the ap-
FiG. 25. — Cordia stihcordata. Ti-ansverse section of nut, x 4: T, testa; C, cotyledon ; TB, TB, testa of barren &eeds in empty cells ; En, endo- carp ; B, barren cell with testa, T ; S, sutural line.
pearance of leaves ; in this instance they are set free by the splitting of the testa.
In the arrangement of a large embryo in a seed there are two ways in which the radicle may be bent over the cotyledons.
Sometimes it is turned up over the back of one of the cotyledons, as in figs. 2G and 27, and is said to be incumbent ; while sometimes it is turned along the
32
ON SEEDLINGS
edge of the cotyledons, as in figs. 28 and 29, wliicli are then termed accumbent. The divisions of the Cruci- ferae are based to some extent on this character, some groups being accumbent and some incumbent. I puzzled for some time over the reasons which could account for this difference, though the explanation which I would suggest is very simple when once stated.
Fig. 26. Fig. 27.
Jt
Figs. 2G and 27. — Sections of seed of Hespe^-is matronalis, x 10; 11, radicle ; C, cotyledons.
Fig. 29.
Fig. 28.
Figs. 28 and 29. — Sections of seed of Wallflower {Cheiranthus Cheiri), x 10.
I scarcely feel justified, however, in doing more than throwing it out as a suggestion.
Seen in section the two forms would be as shown in figs. 26 and 28. Now if from the form of the pods, or for any other reason, it is an advantage that the seed should be compressed, as in fig. 28, then, the thickness of the cotyledons remaining the same, it is better that
I
POSITION OF EMBRYO 33
the radicle should be accumbent ; while, on the other hand, in a thicker or globular seed, as in fig. 26, the incumbent arrangement is most convenient. In fact we find that in groups, such as the Arabidese, where the seeds are as a rule compressed, the radicle is almost always accumbent ; while in incumbent groups, such as the Sisymbriege, the seeds are, on the contrary, more or less turgid. As an actual example of an incumbent form I give Hesperis matronalis (figs. 26 and 27), and of an accumbent, Cheiranthus Cheiri (figs. 28 and 29).
Position of the Embryo in the Seed.
As a general rule, the arrangement and position of the embryo in the seed is approximately the same within the limits of any one genus. There are, how- ever, many exceptions.
In Carum Carvi (fig. 30), for instance, the seed is five-angled, and closely conforms to the interior of the mericarp which it fills.
The straight embryo, which is comparatively large for the size of the fruit and also for the Order, is embedded in a copious fleshy endosperm. The cotyledons are oblong, obtuse, entire, plano-convex, and slightly longer than the radicle, lying with their backs or their edges to the axis of the fruit, or obliquely. All these positions seem to be taken indifferently, and two of them fre- quently in the same cremocarp. In another species of the same genus, C. Ajowan (fig. 31), where the conditions
D
84
ON SEEDLINGS
Fig. 30. — Carum Carvi, x 10. A, longitudinal section of fruit : St, etylo- podium ; E, embryo ; Pc, pericarp ; T, testa ; En, endosperm. B, C, and D, transverse sections of the mericarp showing the cotyledons [Cs) in three different positions, transverse at right angles, and oblique in relation to the axis of the fruit : V, vitta or oil canal.
Fro. SI. — Carum Ajowan, x 24. A, longitudinal section of fruit Fc, peri- carp ; T, testa ; En, endosperm ; E, embryo St, Btylopodium. B, trans- veree section of fruit : Cs, cotyledons.
PLANTAGO 35
in the seed are very similar, the position of the cotyledons with their backs to the axis seems to be constant.
In the genus Plantago the cotyledons sometimes have their faces and sometimes their edges to the placenta. This difference is not indeed mentioned either by Barneoud or Decaisne in their respective monographs of the family. Bentham and Hooker, however, say ('Genera Plantarum,' vol. ii. p. 1223): — ' Embryo rectus v. rarius hippocrepicus, hilo parallelus v. in fructu monospermo ^
erectus v. transversus,'
In P. media the
fruit is capsular, dry,
membranous, 2-celled
and 2-4-seeded. The
E
seeds (fig. 32) are pig. s^.— Plantago media. A, longitxi- .„1«„^ «^„,,«^ «- ^.,U dinal section of seed, x8: ^, embryo,
plano-convex or sub- B, transverse sectiou of seed, x 8: C*,
concavo-convex, pelt- cotyledons. ate, small, with equal obtuse ends ; or with the basal end slightly the broader ; the testa is thin, pale brown ; the hilum a little below the middle on the ventral aspect, round, and deeper brown than the rest of the testa ; the raphe tapers from the hilum obliquely towards the upper end of the testa. The endosperm is copious, fleshy, and white. The embryo is straight, narrow, white, a little shorter than the endosperm, and em- bedded in it, a little nearer the dorsal aspect of the seed and somewhat oblique to the median axis ; the cotyledons are linear-spathulate, tapering towards the
D2
86 ON SEEDLINGS
base, obtuse, entire, and with their faces towards the placenta ; the radicle is inferior, obtuse, and shorter than the cotyledons.
In P. lanceolata (fig. 33) the capsule is also 2-celled, with one seed in each cell. The seed is con- cave on the ventral side, at first pale green, at length becoming yellow. The hilum is oval, forming a whit© or pale spot about or a little below the middle on the ventral aspect. The endosperm is abundant, fleshy, or almost horny when dry, and semitransparent. The
embryo is straight, white, \ embedded in the endosperm,
and a little shorter than the seed. The cotyledons are narrowly oblong or linear, T^ „„ ^, , T' , , obtuse, plano-convex, closely
Fi«. 33. — Plantago lanceolata. ' r ' ^
Transverse section of seed, applied face to face, and with
X 12 : (7, cotyledons. *^ ^
their edges to the placenta. The radicle is narrower than the cotyledons, inferior, and tapering downwards.
In P. Coronopus the capsule is many-seeded. The seeds are oblong-oval, suddenly tapering to an obtuse point at the lower end, small, in transverse section somewhat diamond-shaped, with the angles rounded off, and attached to the placenta considerably below the middle. They are much smaller than those of P. media, and differ much among themselves. The embryo is comparatively large, straight, central, nearly equalling the endosperm in length ; the cotyledons are
PLANTAGO 87
linear, obtuse, entire, plano-convex, thick, closely applied face to face, and with their edges to the placenta.
In P. maritima the fruit is narrowly ovoid, 2-celled, 2-seeded. The seed is oblong-lanceolate, biconvex or flattened on the ventral side. The embryo is straight, large, and nearly fills the seed; the cotyledons have their edges to the placenta.
In P. Cynops the fruit is green, with a pale line where the two carpels come together, and a darker one along the middle of the carpel, giving it in a young state the appearance of consisting of four carpels, 2-celled, 2-seeded. The seed is ovate, obtuse, peltate, compressed dorsally, concave on the ventral side, smooth, shining, deep green when young, and suffi- ciently transparent to show the embryo by transmitted light. The embryo is straight ; the cotyledons linear, obtuse, entire, closely applied face to face, with their edges to the placenta.
In P. arenaria and P. major the cotyledons are also placed with their edges to the placenta.
I was for some time much puzzled as to why the cotyledons in P. media should be placed differently from those of the other species examined ; though the reason seems in reality very simple. At first I thought it might have reference to the mode in which the em- bryo emerges from the seed ; but this does not seem to have any bearing on it. In P. lanceolata, however, and its allies the cotyledons are narrow and thick ; and the seed being somewhat comi^ressed, it will be seen from
88 ON SEEDLINGS
fig. 33 that if the embryo had been placed with the faces of the cotyledons to the placenta, it would not have had room to develop.
On the other hand, in P. media (fig. 32) the reverse is the case : the cotyledons are thin and comparatively wide ; their width, in fact, is greater than their thick- ness. It follows that, if they had been arranged as in the other species, they would not have had room to develop. The difference of position is therefore ex- plained by the fact that in P. media the width of the cotyledons is greater than the thickness ; while in P. lanceolata &c., on the contrary, the thickness of the two cotyledons, taken together, is greater than their breadth.
The normal arrangement of an embryo in the seed is to have the faces of the cotyledons turned to the placenta. There are, however, not a few cases in which, as in these species of Plantago, the cotyledons have tlieir edges to the placenta. When this is the case, it may be suggested as possible that the position is due to the fact of the seeds being more or less, in some cases very much, flattened ; and that the embryo is twisted round at right angles to its normal position, so that the cotyledons may lie in the broad way of the seed, as in Ailanthus, Euonymus, Passiflora, linum, Fraxinus, Diospyros, Heliotropium, and many Crucifers, Legu- minosa3, and Rosacese.
On the other hand, in the case of Claytonia (fig. 34) this explanation will not apply. There would appear
CLAYTONIA
39
no reason, so far as the seed is concerned, why the cotyledons should not lie in the usual position.
It has occurred to me that perhaps the arrangement of the cotyledons may have reference to their exit from the seed. If we examine a germinating seedling of Claytonia we shall see that the testa splits vertically from the micropyle, and the cotyledons from their posi- tion, when they separate, act with greater advantage in
Fig. 34. — Claytonia perfoliata. A, transverse section of seed, x 15: C, cotyledons ; B, radicle. B, seedling, x 6.
enlarging the orifice, and thus securing their exit, than they would if they occupied the more usual position. This, however, I only throw out as a suggestion which requires further investigation.
When the seed is flattened laterally, the embryo must either be narrow or lie with the edges of its cotyledons to the placenta.
For instance, in Heliophila pilosa the seeds (fig. 35) agree closely in form with thofce of the Wall-
40
ON SEEDLINGS
flower (Cheiranthns Clieiri) (figs. 28 and 29) ; they are oblong-obtuse at each end, compressed dorsally, with a notch at one end, and in section are naiTow elliptic ; but while the cotyledons of Cheiranthus are broad, in Heliophila they are long and linear. The
reason of this may be that while in Cheiranthus and other Arabideae the pods are flattened dorsally and the coty- ledons are accumbent in the broad way of the seed (figs. 28 and 29), those of the Sisymbrieae, to which Heliophila belongs, have (fig. 35, A and B) the coty- ledons incumbent, so that they lie across the seed, and it is consequently an advantage that they should be linear. Similar cases occur in other Orders, as, for instance, in Caryophyllaceae and Solanaceae.
Germination.
In the great majority of cases the cotyledons escape from the seed and are carried above the ground, becom- ing green and taking part in the processes of assimi- lation like foliage leaves.
When, however, the testa does not readily split, and where in large seeds there is no endosperm, the diffi-
FiG. 35. — Heliophila jnlosa. A, longitu- dinal section of seed, x 12. B, trans- verse section of seed, x 12.
GERMINATION 41
culty of unfolding the cotyledons and extricating them from the seed beccnnes great, and we arrive at cases where Nature seems to have abandoned the attempt, and, as in the Oak and Horse Chestnut, the cotyledons never quit the seed. Thus, among the Juglandeee, Pterocarya has leaf-like cotyledons, while those of the Walnut never quit the shell. Everyone, however, must have observed the elaborate folds into which the two cotyledons are thrown — folds which seem to have no significance or importance now, and which carry us back to a time when the Walnut, like Pterocarya, had foliaceous cotyledons.
If these suggestions be correct, we should expect that species with non-emerging cotyledons would generally have large seeds and be exalbuminous. This certainly appears to be the rule ; among the species with refer- ence to which I have notes, there are 37 genera in which the cotyledons are subterranean or remain in the seed. The seeds themselves are notably large, and all but three are exalbuminous. Some liupines and Beans present us with an intermediate stage, the cotyledons being aerial, and more or less green, but fleshy and by no means leaf-like. For other cases see pp. 60-90.
Cotyledons which appear above the surface of the soil are said to be epigeal, those which remain beneath the surface and never leave the seed, hypogeal.
In cases where the seed is well covered, or the primary root has secured a good hold in the soil, the growth of the hi/pocotyl or portion of the axis below the
42
ON SEEDLINGS
cotyledons, is sufficient to draw the latter out of the seed. Our native hedge-plant, Galium Aparine (fig. 36), will serve as an example.
A B
F o. 86. — Galium Aparine. A, longitudinal; B, transverse section of seed, x 8: e, embryo; pc, pericai-p; <, testa; en, endosperm; in, micropyle. C, germinating seedling, x 4. D, seedling, x 2.
WATER PLANTAIN 43
During germination the endosperm is gradually- absorbed, and the cotyledons increasing at the same time fill the internal cavity of the seed. The radicle pushes its way into the soil where it throws out lateral rootlets and gets a firm hold. The hypocotyl is the first to ap- pear above ground, and if the seed is buried sufficiently deep is able to pull the now largely developed cotyledons out of the soil without the testa ; if this is not the case, the testa is carried up on their tips.
In the Water Plantain (Alisma Plantago), where the radicle does not at once elongate, the peculiarly thickened hypocotyl produces numerous root-hairs by which the seedlingr is fixed in the soil or mud. The single awl-shaped cotyledon frees itself by its own growth.
Seeds sown in a greenhouse, in soil sunk in water, germinated in three days.
The hypocotyl first emerges, ending bluntly or even truncately in the very short radicle, with a thick- ened margin around the tip (fig. o7, A).
Three days after germination (fig. 37, B) the thick- ened margin at the base of the hypocotyl gives out numerous root-haii-s which fix the plantlet in the soil. The place whence . the plumule presently emerges is dimly visible at the base of the slender cotyledon. The carpel or achene is now suspended at the tip of the cotyledon. The hypocotyl is very short. The radicle has not yet elongated.
Four days after germination (fig. 37, C) the coty-
Fig. 87. — Alisma Plantago, flViowing various stages in germination and growth of the young seedling. A, B, and C, x8; D, xi; E and Ff nat. size. C, cotyledon.
WATER PLANTAIN 45
ledon rises up, freed from the acliene, but is still slightly twisted at the end.
Seven days after germination (fig. 37, D) the coty- ledon becomes much elongated, narrowly subulate or filiform, tapering slightly upwards and straight. The root-hairs are still copious at the base of the short hypocotyl, and the radicle has become considerably elongated. The plamule is also considerably advanced, having broken through the sheathing base of the cotyledon by which it has been hitherto protected.
Twelve days after germination (fig. 37, E) the coty- ledon has not altered ; but the first leaf has appeared. It is linear, acuminate, thin, membranous, and shows a distinct midrib. An adventitious rootlet is seen just pushing from the first node above the hypocotyl. The radicle is elongated and furnished with root-hairs nearly to the tip. ,
Nineteen days after germination (fig. 37, F) the cotyledon is still unaltered. A second leaf similar to the first has been developed and is also sheathed at the base by the cotyledon. The radicle is elongated and fur- nished with root-hairs. Adventitious roots well furnished with root-hairs have also sprung from the first node.
The escape of the cotyledons is often facilitated by their being narrow, as probably in the species of Clay- tonia to which I have already referred ; or divided as in Eschscholtzia (fig. 17) and Schizopetalon (fig. 185). Where the seed is enclosed in a stony endocarp which does not burst in germination, their escape is
IG
ON SEEDLINGS
rendered possible by their narrow shape. Symplocos paniculata (fig. 38), a member of the tropical family Styracese, may serve as another illustration.
The endocarp of the fruit is obovoid, vroody, and does not burst during germination. The radicle emerges by a small hole at the narrow end, the
Fig. %%.— Symplocos paniculata. A, seedling, nat. size. B, germinating seedling with vertical section of endocarp and seed, showing mode of exit of embryo, x 2.
hypocotyl elongates, becoming curved and finally straightening, carrying up with it the endocarp con- taining the seed. As the cotyledons elongate they push out at the small hole in the endocarp, finally free themselves and spread out to the light.
The cotyledons owe their shape to that of the seed which is reniform or curved, and contains a large quan-
SYMPLOCOS
47
tity of fleshy endosperm. If a seed is cut open during germination the cotyledons are found to be linear, obtuse, and together with the apex of the hypocotyl
B
Pig. 89. — Veronica hedercefolia. A, seedling, x 4, sliowing mode of ger- mination wlien the seed is buried. B, another stage of A, x 4. C, seed germinating when insufficiently buried : H, radicle ; vc, ventral cavity of seed, x 4. D, another seedling, x 4.
curved like the head of a shepherd's staff. Their shape then might be accounted for by the shape of the seed, the quantity of endosperm originally surrounding them,
48 ON SEEDLINGS
and, thirdly, by the difficulty that broad cotyledons would experience in getting out of the indehiscent endocarp.
As the figures of Alisma and Symplocos show, the seed or one-seeded fruit may be carried up on the hypocotyl or tip of the cotyledons in germination. This is generally due to an insufficient covering of soil. For instance, in the Corn Speedwell (Veronica hederoe- folia), when the seed is well buried the radicle pushes out first and fixes the plant in the soil, then the hypo- cotyl arches and in straightening pulls the cotyledons out of the seed.
Two days later the cotyledons get quite clear, an easy process when the seed is properly covered with soil. The hypocotyl elongates greatly meanwhile (fig. 39, A, B).
When the seed has been insufficiently buried, it is carried up with the cotyledons, which have then, as it were, to wriggle their way out (see fig. 39, 0 and D).
Dipsacus ferox (fig. 40) supplies another instance. The ovary of the Dipsaceas is simple, being inferior and one-celled, with a solitary pendulous anatropous ovule; but in addition to the true calyx with which it is surmounted, each flower of the capitulum is sur- rounded by an involucel which completely encloses it, and appears like a second or double calyx. The fruit is an achene surmounted by the persistent rarely deciduous (as in D. ferox) calyx, and is completely
DIPSACUS
49
enclosed by the involucel, wliicli is also persistent and either surmounted by a lamina varying greatly in size, or truncate at the apex. The seed is solitary and suspended from the apex of the loculus, is covered by a thin mem- branous testa, and contains a layer of fleshy endosperm. The embryo is straight with a superior radicle and oblong or ovate cotyledons.
In D. ferox the radicle emerges at the apex of the fruit through the mouth of the truncate in- volucel. Numerous root-hairs are produced on the radicle and the base of the hypocotyl
The fruit enclosed in the in- volucel is usually carried up with the cotyledons (fig. 40, about the second day after their first appearance above ground), which by widening have split the involucel along one side.
Nearly all the seedlings carry up the fruit during germination, but eventually the energy of the expanding cotyledons generally, if not always, suffices to extricate them from their investments and they spread out to the light. They are oblong in shape, and entire with a slightly prominent, apical, colourless tooth ; narrowed into a short petiole, with n, few obscure lateral nerves; at first pale green in colour, but soon becoming darker.
E
Pig. 40. — Difsacus ferox. The fruit enclosed in the involucel carried uj) by the cotyledons in gcrini- niition. Second day after first appearance abova ground, x 3.
50
ON SEEDLINGS
The fruit and involucel of Scabiosa australis are closely similar to those of Dipsacus ferox, but the behaviour in germination is different.
The achene closely oc- cupies the interior of the involucel, so that on germi- nation the radicle gets out- side immediately and pushes down straight into the soil for a considerable depth, at the same time giving off nu- merous root-hairs (fig, 41). If the involucel is fairly well covered with soil the cotyledons, after the long radicle has established itself, are easily and readily pulled out clear of the seed and its investments. The coty- ledons in this species are not very broad, because the seed itself is not very thick.
The base of the hypocotyl has a thickened projecting ledge which presses against the rim of the involucel and pins it to the earth while the arching upper part grows upwards and extricates the cotyledons (tig. 41, B).
Fig. 41. — Scabiosa australis, x 6. Commencement of germination. In A tlie radicle (?•) only has pro- truded ; in B the cotyledons (c) are nearly free.
SCABIOSA
51
A third type is represented by Scabiosa Gram- muntia (fig. 42). The involucel is furnished with a well-developed cup-like lamina, and the radicle on germinating pierces it, and holds it firmly to the soil by means of a thickening at the base of the hypocotyl while the latter on lengthening draws out the cotyledons. This thickening is sym- metrical or equal all round, not unilateral as in S. australis. The fruit of S. Columbaria is rather thicker and shorter, while the coty- ledons are also shorter, but in other respects the seedlings agree with those of S. Grammuntia.
In a fourth type the radicle of the germinat- ing seedling pushes through the sides of the involucel, beneath the thick- ened rim at the base of the lamina. The base of the hypocotyl also has a symmetrical, annular thickening as in the third group. The involucel of Scabiosa palsestina (fig. 44) is salver-shaped and attains a much greater size than that of the other species which I have examined. The lamina is large, membranous and traversed by straight radiating nerves excur- rent at the margin, forming a deeply fringed border.
e2
Fig. 42. — Scabiosa Grammuntia, x 4. Commencement of germination, r, radicle; c, cotyledons.
52
ON SEEDLINGS
Beneath the thickened rim at the base of the lamina are numerous perforations, through one of which the radicle finds its way, fixing the fruit to the ground.
The radicle in germinating emerges through the opening at the apex of the involucel, easily rupturing the thin mem- branous pericarp. If the invo- lucel with the fruit is placed mouth downward in the earth (an unnatural position), and but lightly buried or lying on the surface of the soil, the whole may be carried up with the cotyledons. If, on the other hand, the involucel is lying
Pig. 43. — Scaliosa palcEstina. Germination, x 2.
Fig. 44. — Seahiosa palccstina. Invo- lucel, containing fruit, x 3.
on its side or with its mouth upwards, the radicle pro- ceeds freely a short distance, and then turning down-
SCABIOSA
53
wards passes through one of the openings, and grows deep down into the earth. Some seedlings in the course of germination showed a radicle measuring 20-40 mm. in length, seven days from the time of sowing. This conduct of the radicle is of great importance in a country with the dry climate of Syria, to which the plant is indigenous, not only in its depth of pene- tration, but in effectually fixing the involucel, peri- carp and seed-coat to the earth while the coty- ledons emerge and rise up free. The cotyledons after- wards become more or less spread out, oblong and pe- tiolate.
S. graminifolia has like- wise a perforated invo- lucel. The thickening of the bypocotyl is well shown in S. atropurpurea (fig. 45). The lamina of the involucel is smaller than in S. palaestina, bat there are no perforations in any part of it. The radicle easily pushes its way through the membranous portions at the
Fio. 45. — Scabiosa atropurpurea. Germination, x 3. Cal, the five segments or bristles of the calyx.
Fig. 46. — Scahiosa caucaaica, X 3. A, commencement of germination; B, day after commencement. JB, thick* ened base of hypocotyl ; O, cotyledons.
VALERIANELLA
55
sides between the greatly thickened ribs of the involu- cel. A peculiar anomaly presents itself in the ger- minating seedling of Scabiosa caucasica (fig. 46). The base of the hypocotyl is more or less thickened as in the other species, but it generally if not always pushes itself right through one or other of the mem- branous portions of the involucel, and is consequently functionless so far as its original purpose is con- cerned. The lamina of the involucel is reduced to a rim crowned by numerous coarse bristle-like seg- ments. The seedling is different from all others I have noticed. The hypocotyl remains short, the oblong entire cotyledons are sessile, and the first pair of leaves are ob- lanceolate and entire.
In all of the above-mentioned
seedlings the cotyledons are sessile Fig. n.—Vaierianella-
coronata, x 4. on leaving the seed, but afterwards
petiolate, except in Scabiosa caucasica, which is peculiar
in several respects.
The germinating seedling of Valerianella coronata
(fig. 47) shows a similar contrivance for fixing the fruit
or achene to the soil and thus enabling the cotyledons
to make their escape. The embryo swells up and the
radicle usually bursts through the side of the fruit
56 ON SEEDLINGS
beneath the calyx limb ; it then piercer the latter, entering by the outer face and pushing on through both sides, enters the ground, thus pinning the fruit to the soil, while the cotyledons are pulled out in a com- paratively short time and spread to the light.
One embryo only out of twenty-eight made its exit at the apex of the fruit, and from the centre of the calyx limb. The radicle in this case pushed through the calyx from its inner face outwards, and not from the outer face inwards, as in the usual method.
The fact of the radicle bursting through the side of the fruit is accounted for by two cells of the three-celled ovary being small and empty, and the third containing a seed which would accordingly be excentric, and there- fore in germination the radicle would normally push through one side of the fruit.
In Valerianella Auricula when the fruits have been moderately well covered with soil the embryo in most cases gets clear of the fruit simply by the elongation of the hypocotyl.
The genus Hedysarum (Leguminosoe) affords an interesting comparison with Scabiosa and Valeriarolla. The fruit is a lomentum, that is, a pod divided into one-seeded segments which become separated but do not open to let the seed escapa.
In germination the radicle pushes itself through the end of the lomentum or through one suture, while the cotyledons make their exit by the other, as in H. coro- narium (fig. 48) and H. obscurum. The wall of the
HEDYSAKUM
57
fruit in H. denti- ctilatum (fig. 49) is thin or mem- branous and reti- culate, and the radicle generally pierces one valve, while by the swell- ing of the coty- ledons and the
elongation of the p^ruarp and te^ta- hypocotyl the lo- mentum is burst open and the upper valve pushed up- wards, permitting the exit of the seedling. The
testa is frequently can-ied up on the cotyledons, but the breadth of the latter and their increasing size soon split it open and get rid of it. Whatever the mode of exit of the radicle and cotyle-
Fio. 48.- Bize.
■Hedysarum coronarium. Nat c, cotyledon ; r, primary root.
58
ON SEEDLINGS
dons, the wall of the fruit nearly always gets pinned to the ground while the seedling rises clear above it. Thick-walled fruits last a long time in good condition after the germination of the embryo.
Fig. 1 shows the embryo germinating and splitting open the pericarp of a segment of the fruit. The radicle has pierced the lower valve of the pericarp.
Fig. 49. — Hedysarum denticulatum, x 2^.
Fig. 2. A case in which two segments have remained united. One embryo is dead ; the other has forced open the pericarp of both segments, and its radicle has pierced the lower valve, a frequent occurrence.
Fig. 3. Another segment of the fruit where the radicle has already advanced considerably through the
DESMODIUM 59
wall ; the two halves of the fruit have just commenced to split open.
Fig. 4. Another mode of germination similar to what occurs in H. coronariura, the radicle emerging at one side of the fruit and the cotyledons at the other. One day after germination.
Fig. 5 shows the ragged margin of the fissure made by the radicle through the walls of the fruit during germination.
In an allied plant, Desmodium canadense, the radicle in germination emerges at the dorsal suture of the fruit, or breaks through the side walls. It fixes itself in the ground, and the lengthening hypocotyl generally carries the lomentum and seed up with it. All the coverings, however, finally drop off, leaving the seedling free. Where the radicle pierces the sides of the lomentum the latter is fixed to the soil. The coty- ledons on becoming free are ovate-oblong, obtuse, sub- falcate, or rounded on the upper edge and with a shallow sinus on the lower edge, subfleshy, pale yellow- ish-green, shining and glabrous.
Medicago orbicularis affords an instance of a very rare case whei'e a many-seeded pod does not dehisce before germination, but is pinned to the ground by the seedlings in their efforts to escape.
In germination the spirally coiled legume splits readily along the dorsal suture, and the radicles of the various seeds seem all to make their exit along one side, while the cotyledons push out along the opposite side.-
Fig. 50.~Medicago orbicularis, x 2. Nine seedlings, all from one fruit.
MEDICAGO
61
By this means the fruit is effectually fixed to the soil while the embryos have no difficulty in getting out. The testa is, however, sometimes carried up by the cotyledons and splits irregularly as these expand.
Fig. 51. — Phaseolus vulgaris. Half nat. size, c, cotyledons.
Twelve to sixteen seedlings from one fruit are not infrequent. A few instances occur (two to three out of fifty-six) where the radicle and cotyledons emerge together at one side of the fruit, and consequently stand
62
ON SEEDLINGS
out free from it. One instaiice occurred amongst fifty- six seedlings where the cotyledons got clear away from the fruit and seed, but the radicle failed to get out of the seed and consequently the seedling did not make the same progress as the rest and would most likely perish.
Pig. 62. — Lonchocarpus latifolius. Half nat. size, c, cotyledons.
The competition among the seedlings from a single pod is in such cases very severe.
Phaseolus vulgaris (French bean) presents us with an intermediate stage, the cotyledons being aerial and green, but fleshy and by no means true foliage-leaves.
Lonchocarpus latifolius (fig. 52) is another instance.
Fig. 63. — Hymenaa Cour- baril. Nat. size, c, coty- ledons.
64
ON SEEDLINGS
The cotyledons are falcate, obtuse, subsessile, and do not spread horizontally ; they are aerial but fleshy, and enlarge but slightly after germination. They fall at an early date, before which they become bright yellow.
s^^fess-:sa
Fig. 54. — Buchanania latifolia, showing thin edge of cotyledons. Nat. size.
We may compare also Hymenoca Courbaril (fig. 63), where the thick, fleshy, plano-convex cotyledons are pale brownish on the back, light green above (the green part being bordered by a narrow rim).
Again, in Copaifera oiflcinalis the cotyledons are
ANACARDIUM
65
aerial, but very fleshy, oblong, sessile, rouuded on the back, flat ou the inner face or becoming concave by shrivelling, and reddish or tinged with yellow,
A transition from the strictly aerial and foliaceoua cotyledons to the fleshy and subterranean is also met with in Anacardium occidentale. Here they are large, fleshy, falcate, and directed to one side of the young stem, yellowish-green in colour, plano-convex, three- nerved, and slightly reticulate on the back, but showing
Fio. 55. — Ardisia pohjceph(da. Half nat. size.
no trace of nerves on the upper surface. Another some- what similar case occurs in Buchanania latifolia (Ana- cardiaceae) (fig. 54), having obovate, unequal-sided, fleshy cotyledons with a thick and a thin edge. They conform to the interior of the seed and that again to the endocarp, to which they owe their peculiar shape. Immediately after germination they are both directed to one side of the stem as in subterranean cotyledons : but they ultimately spread out right and left. "J'hey soon drop off.
F
66
ON SEEDLINGS
In four species of Ardisia (Myrsineae) examined, the cotyledons differ in form, as does also more or less the mode of germination. The cotyledons of A. polycephala (fig. 55) are elliptic, emarginate, foliaceous, and persistent, with an incurved pinnate venation like that of the leaves, the first two of which are similar, but larger and entire.
In A. mamillata (fig. 56) the hypocotyl is short but very stout and fleshy. The cotyledons are aerial but small, transversely oblong, bifid and hairy. Those of A. crenulata (fig. 58) are small, spathulate, and petiolate, but never leave the seed. The latter if fairly well covered with soil remains underground, when the petioles of the cotyledons attain a few millimetres in length ; but if the fruit with its seeds is uncovered the cotyledons are almost sessile, and the seed gets carried up by the elongating hypocotyl. This is more decidedly the case in A. japonica (fig. 57), the hypocotyl of which varies from 13-30 mm. in length. The cotyledons are small, ovate and never leave the seed, which therefore gets carried up with the growth of the seed- ling. The petioles of the cotyledons are both directed to one side of the axis, flattened, pubescent, and be-
FiG. 56.
Ardisia mamillata.
' Seedling showing emarginate cotyle- dons (C). Nat. size.
AEDISIA
67
come undulated as tliey lose substance and fade. In the two species last mentioned the cotyledons are to all intents and purposes subterranean, though car- ried above ground by the elongation of the hypo- cotyl.
The seed of A. japo- nica presents the remark-
FiG. 57. — Ardisia japonica. Germinating seed showing six embryos, x 2.
able peculiarity of often containing several em- bryos, as many as six being present on some occasions. The radicle in such a case points in vai'ious ways, and in germination each embryo makes its exit at a different place (fig. 57).
F 2
Fig. 58. — Ardisia crenulata. Nat. size. 6', seed containing cotyledons.
68
ON SEEDLINGS
In Trichosanthes cucumerina (CucurbitacesB) (fig. 59) the fleshy plano-convex cotyledons conform in outline to the shape of the seed and are strictly sub- terranean, but burst the testa during germination and frequently become more or less expanded beneath the soil.
Fig. 50. — Trichosanthes cucu- merina. Nat, size. S, seed C, cotyledon.
Fig. 60. — Trichosanthes palmata Nat. size. C, cotyledon.
LUCUMA
69
In T. palmata (fig. 60), on the other hand, the hyix)Cotyl is 3-4 cm. long, thrusting the cotyledons well above the soil. These are broad and spreading, and described as rather persist- ent, though only yellowish- green on the upper and whitish on the lower surface. In T. anguina the hypocotyl is still longer, though the aerial cotyledons are thick.
In a species of Lucuma (fig. 61), a genus belong- ing to the tropical family Sapotacea?, the hypocotyl is very short, stout, and sub- terranean, while the very large, fleshy cotyledons are also generally subterranean, but sometimes appear just at the surface of the soil, burst- ing the woody testa into two halves and spreading horizon- tally, becoming deep green on the upper surface when exposed to light. Often one cotyledon becomes suberect while the other is deflexed.
The seedling of Lucuma mammosa is very similar
Fio. 61. — Lucuma sp., showing the two halves of the testa attached to the cotyledons. Half nat. size.
70 ON SEEDLINGS
to the last in all leading particulars. The cotyledons are subterranean and fleshy, but split open the woody testa during germination. The stem is very stout and is notable for the length of the first internode which measures 7-10 cm., and compensates for the want of the hypocotyl. The leaves are sometimes falsely opposite, but as a rule alternate, and the first six are closely aggregated above the first internode, oblong or lanceolate-elliptic in shape, and feather-nerved.
The cotyledons of Butyrospermum Parkii (Sapotacea?) are fleshy and perfectly subterranean, very thick, and applied to each other face to face and (in old seeds at least) inseparable, each occupying half the space in the seed, and, like it, of variable outline, but always broadest at their bases, which rest on the base of the seed.
In Melittis melissophyllum (Labiatae), again, ac- cording to Irmisch,^ the fleshy cotyledons generally re- main in the seed, and are held together by the testa : but they sometimes burst the shell, and stand out from one another. Like true subterranean cotyledons, they have no stomata.
When the cotyledons are large, thick, and fleshy they often contain sufficient nourishment to render the plants for a time independent of any fresh supply. In such cases the seedlings sometimes push up for a while \srithout any fully-developed leaves, the first few being reduced to a very small size or almost obsolete.
' • Zur Naturgeschichte von Melittis vielissojfhyllum,' Bot. Zcit. 1858, p. 233.
CLEMATIS
71
In the following pages I have brought together some instances of truly hypogeal cotyledons.
In Clematis recta (fig. 02) the cotyledons are fleshy and do not leave the seed. As usual in such cases the first few leaves are reduced to scales.
This species is of interest from being quite exceptional in the Order Ranunculacege in its way of germi- nation.
In all the other genera observed the cotyledons were truly aerial, and this is also the case iir Clematis gra- veolens (fig. 63), where the radicle in germination pushes ont at the apex of the achene, which splits rather deeply into two valves, thus allow- ing the embryo free exit. The l)rimary root develops numerous root-hairs at an early stage, which fix the seedling firmly in the soil.
The achene is rarely if ever carried up by the cotyledons, but is securely fixed in the soil by the long, feathery, persistent style.
In the family Guttifera? the em- bryo is large, filling the seed, to which it confonns : it presents a remarkable and abnormal case, inasmuch as it often.
Fici ()2. — Clerintin recta. Nat. size. The scule- like leaves of Nos. 1-3 pairs inclusive had fallen off. S, seed.
Fig. 63. Young plant of Clematis gravco- lens, var. orientalis. Nat. size. C, cotyledon.
GUTTIFER.E 73
consists mainly of a swollen fleshy radicle with small scale- like inconspicuous cotyledons at its apex as, for instance, in the tribe Clusieae, while cotyledons are said to be absent in the Moronobeae and wanting or very minute in the Garciniege. The radicle is always inferior, and in the Calophyllege it is very short with large cotyledons sometimes grown together in one mass. In Quiina they are large and fleshy but distinct, while the radicle is very short.
In the seedlings observed the cotyledons where present are subterranean in germination, and this gfenerally obtains where the seeds are large, exal- buminous, and contain a fleshy embryo. The cotyledons of Calophyllum Inophyllum (fig. 64) occupy a large globular seed, and their fleshy petioles are just long enough to allow the plumule to get clear out of the seed. They are also subterranean. The first pair of leaves are reduced to scales, and generally the second pair also, or these are imperfectly developed. Then follow two pairs of narrowly elliptical, obtuse leaves, after which the seedling stops growing for a season. The feather-nerved venation is remarkable on account of the close arrangement of the veins in parallel lines.
Xanthochymus pictorius (fig. 65) and Mesua ferrea agi*ee with the above in all general particulars ; but in the former the first four pairs of leaves are reduced to small, ovate, or subulate, black scales, and only one pair of lanceolate acuminate leaves are
74
ON SEEDLINGS
produced the first year by the seedling. In Mesua ferrea the first four pairs of leaves are small, scale-like, and caducous, followed by two pairs of lanceolate-oblong
Fig. 64. — Calophyllum Inojifiyllum. Half nat. size. S, seed.
very finely feather-nerved leaves, before growth ceases for a time.
The tea-tree (Camellia theifera) (Ternstroemiacese) offers a remarkable exception amongst its congeners in
TEA-PLANT 75
having a large, fleshy, globular embryo, with the small radicle and plumule completely surrounded or covered by the hemispherical cotyledons. The radicle of Camellia reticulata, on the contrary, projects beyond the cotyle- dons considerably.
Pig. 65. — Xantlwchymus pictorius. Half nat. size.
The cotyledons of the tea-plant (fig. QQ) swell ud somewhat during germination, bursting the woody testa, but otherwise do not alter in size and shape. They form a hemispherical mass occupying the interior of the seeds, and are subterranean. They assume more or less of a green colour should they be exposed to light during germination by the bursting of the testa and by removal of the soil.
76
ON SEEDLINGS
Fig. 66. — Camellia theifera. Half nat. size. S, seed.
OCHNA— SAPINDUS
77
In Ochna Kirkii (OchnacesB) (fig. 67) the fleshy cotyledons fill the seed, which in turn fills the carpel. They remain in the ground attaching the seedling to the fruit.
The first and second leaves
Fig. 67. — Ochna Kirkii. Nat. size. Fig. GH.— Sapindvs ina-qvalis. S, fruit. Half nat. size. S, seed
are reduced to small, subulate scales, No. 3 is small and foliaceous, No. 4 and those following much larger.
In Sapindus inasqualis (Sapindacege) (fig. G8) the thick fleshy petioles are pushed out of the seed or
78
ON SEEDLINGS
elongated so as not to inconvenience -the radicle and
C
plumule in their exit.
Here also leaves Nos. 1-4 are linear-lanceolate, acute, small and scale-like.
Thick, fleshy, and truly subterranean cotyledons oc- cur also in Rhus Thunbergiana (Anacardiaceas) (fig. 69),
a South African species with almost dry, one-celled, one- seeded fruits from three- quarters to one inch broad. The first five leaves are minute scales ; while all succeeding ones are simple, and feather-nerved.
In Lathy rus Nissolia (Leguminosae) (fig. 70) the hypocotyl is subterranean, fleshy, colourless, and very short (about 1-75 mm. long). The plano-convex, fleshy cotyledons, each occupying half the interior of the seed, are unequal-sided, so that the petioles and point of attach- ment to the plumule are at one side. Tne petioles are unequal or the lower one is undeveloped, while the upper one is developed and curved just as much as will allow the cotyledons to lie horizontal. The latter do some- times stand erect on their ends and then both petioles are equally developed.
Fig.
9. — Bhus Thunbergiana. Half nat. size.
LATHYRUS NISSOLIA
79
The first two leaves are small, subulate, scale-like and adpressed to the stem. The succeeding ones are
Fig. 70. — Lathyrus NistoHa. Nat. size. S, seed
narrowly linear and convolute in bud enclosing the younger leaves.
80
ON SEEDLINGS
Eugenia bracteata (Myrtaceee) (fig. 71) isanotlier ex- ample. The first and second pairs of leaves are reduced to small green scales, ultimately becoming brown, while the third pair are oval and foliaceous ; the fourth pair are
elliptic and much larger.
Bignonia insignis (fig.
72) is of interest in having
the thick fleshy cotyledons
amalgamated in one piece,
which remains in the seed
after germination. The
first pair of leaves are small
and scale-like; the second,
tliird and fourth pairs are
cordate, foliaceous and
simple ; while the fifth pair
have each a pair of leaflets
and a terminal three-parted
hooked tendril.
In the Walnut (Juglans regia) (fig. 73) the primaiy
root emerging from the apex of the nut is very stout,
woody, tapering downwards, and furnished with lateral
rootlets.
The very short, stout, woody hypocotyl is scarcely distinguishable from the root. The cotyledons are large, fleshy, bilobed and crumpled, filling the whole cavity of the seed and remaining there after germination, attached to the young plant by their short fleshy petioles.
Fig. 71. — Eugenia bracteata. Four-fiftlis uat. size. 6', seed.
HOLM OAK
81
The ultimate leaves are compound and imparipinnate, the first eight in the specimen drawn being reduced to scales and opposite or subopposite in four pairs.
The Oak and Hazel are also common instances of this mode of germination.
The cotyledons of the Holm Oak (Quercus Ilex) (fig. 74) are oblong, piano- coti vex, fleshy and sub- terranean. They are auricled at the base and petiolate,
Fig. 73.
Juglans regia.
One-tenth nat. size.
n, nut.
Fio. 72. — Bignonia insignis. Nat. size. S, large winged seed.
and although the shell of the acorn splits longitudi- nally when they swell during germination, they never emerge. The first five leaves are linear - subulate, minute, scarious, brownish and caducous. The sixth is somewhat larger, cuneate and trifid at the apex ; the ssventh and eighth are foliaceous but small. The
G
82
ON SEEDLINGS
ninth is much larger, lanceolate-oblong, and distantly serrate. The tenth to the thirteenth inclusive are elliptic-oblong, irregularly penninerved and acutely serrate. The above characters of the primary leaves vary according to the vigour of the seedling and the conditions under which germi- nation takes place. The primary internodes have not the power of elongating, so that when an acorn happens to be deeply buried with dead leaves in a wood or planta- tion the primary leaves are ar- rested in growth and scale-like, while those leaves or most of those (probably never all) which reach the light become green and foliaceous.
The cotyledons of Q. pedun- culata are also subterranean. Several of the primary leaves are reduced to scales as in the last case and for the same reason. The first green foliaceous leaf is obovate-
FiG. li.—Quercus Bex. t ■, ^^ ^ ^ ^ ^
Half nat. size. oDiong and shallowly iobed at
the top. Succeeding ones are obovate or obovate- oblong, and irregularly Iobed with alternate nerves running into the lobes.
In CycadacesD the seed is large, variable in shape, and often variously angled. The outer layer of the testa is
CYCADS
83
orange or red, and the inner crusta- ceousorbonj. Endo- sperm is copious and fleshy, embedding the subcylindrical em- bryo in its npper part ; and the cotyle- dons are conferrumi- nate or amalgamated in one piece, or free only at the base to permit the exit of the plumule.
The seeds are large, and in all the seedlings observed the cotyledons are sub- terranean, remain- ing in the seed till the endosperm is ab- sorbed, when they decay. The plumule consists of a mass of stout, fleshy scales resembling a resting or winter bud, which emerges from the fis- sure at the base of
Fig. 75. — Cycas Beddomei. A, first leaf ; B, second leaf; C, C, petioles ot cotyledons; S, seed. Nat. size.
o2
84 ' ON SEEDLINGS
the cotyledons during germination. A solitary perfect leaf is soon after developed from the centre of the bud as seen in Zamia integrifolia. This leaf consists of two pairs of Jeaflets, crowded together at the apex of the petiole ; the terminal pair is much the smaller, and all have numerous, longitudinal, parallel nerves.
The seed of Cycas Beddomei (fig. 75) is globose- oblong, and the bud formed by the plumule is rather small, consisting of one or two scales. The method of germination corresponds to that of Zamia integrifolia. The primary perfect leaf is pinnate with eleven to nineteen linear, entire leaflets having a distinct midrib only. Dioon spinulosum has a comparatively large, oblong seed and a large bud of four scales. The first leaf is finely pectinate or pinnate with twenty-three to twenty-five spiny-serrate leaflets, the veins of which are longitudinal, slender, and numerous. In all three cases the leaflets are circinate in vernation like most Ferns.
In most of the above species it is clearly an object to the young plant to grow upwards quickly, and the first few leaves are reduced to scales. In the garden Nasturtium (^Trofceolmn majvs) (fig. 76), however, the habit is quite different, the first pair are large and green, differing from the normal adult leaves only in being opposite, not alternate, and in their shallowly trilo- bulate outline, the adult form of leaf being seven-lobulate.
In the Scarlet Runner (Fhaseolus multiflorns) the first two leaves are also opposite, forming a pair of true foliage leaves. They differ, however, from the form of
TROPJ-OLUM
8^
the subsequent leaves in being simple while the latter are trifoliolate.
Fig. 76. — TropcBolum majus. Nat. size. S, seed containing cotyledons
In Hevea Spruceana (Euphorbiaceas) (fig. 77) the cotyledons are very large and broad, remaining in the
86 ON SEEDLINGS
seed beneath the ground till they decay after absorbing tlie endosperm. The first two leaves, which are here also opposite, contrasting with the alternate arrange- ment of those following, are large and compound.
In an Australian plant, Castanospermum australe (Leguminosae) (fig. 78), the cotyledons are enormously large, semiglobose and fleshy, filling the seed and con-
FiG. 77. — Hevea Spruceana. One-fourth nat. size. S, seed.
forming to it in shape. They remain in the testa, which splits irregularly during germination. The plumule is very large and well developed. It gives rise to an erect stem, ultimately becoming woody, and forming a tall tree.
The embryo of Lecythis Zabucajo (fig. 79), a member of the family Myrtaceae, consists of a solid piece while still in the seed, to the interior of which it strictly con-
LECYTHIS ZABUCAJO
87
forms. It is narrowly ellipsoid or fusiform, and is evidently homologous with the hypocotyl, which has become thickened and fleshy in this remarkable way in order to constitute a storehouse of reserve material. By
I
Pig. 78. — Castanospermum australe. Nat. size. P, developing plumule.
remaining in the thick woody testa both during and after germination it is protected from the depredations of animals. Cotyledons are absent or minute. This being the case it followa that the seedling must behave
88
ON SEEDLINGS
differently from those having cotyledons which are also generally if not always petiolate when subterranean or remaining in the testa in germination, as in Eugenia
Fig. 79. — Lecythis Zahucajo. A, transverse section of seed : OT, testa ; IT, tegmen ; C, amalgamated cotyledons. B, germinating embryo removed from seed. C, seedling. S, seed ; B, radicle ; P, plumule ; L, scale-like leaf. Half nat. size.
BRAZIL NUT
89
bracteata &c. The petioles in such cases allow the embryo to escape from the testa, with exception of the fleshy
Fig. 80. — Bertholletia excelsa. Nat. size.
90 ON SEEDLINGS
laminee of the cotyledons. On the contrary the fleshy hypocotyl of Lecythis Zabucajo is held fast while the plumule splits the testa and pushes its way out at one end of the seed and the radicle at the other. The allied Bertholletia excelsa (Brazil nut) (fig. 80) behaves in the same way.
In this plant a great number of the primary leaves are very small and completely cover the lower part of the stem owing to the non-development of the internodes. They gradually change from sheathing organs to perfect leaves, of gradually increasing length. The first eight are colourless, the next five are oblong, followed by one or more lanceolate ones, while the fifteenth and sixteenth are oblong-obovate, but very moderate in size. Only three or four of the first leaves of Lecythis Zabucajo are scale-like. In both species the colourless leaves would act as a protection to the plumule while pushing its way through soil, or whatever else might have accumulated above the seeds in a state of nature. The following four leaves of Lecythis are roundly cordate followed by four that are ovate, and then by others which are lanceolate. The ultimate form is oblong-lanceolate with serrate margin, and the leaves are alternate and distichous.
Forms of Cotyledons.
No one who has ever looked at young plants can have failed to be struck by the contrast they afford to older specimens of the same species. This arises partly
L
FORMS OF COTYLEDONS 91
from differences in the leaves, partly from the contrast which the cotyledons afford, nob only to the final leaves, out even to those by which they are immediately followed.
This contrast between the cotyledons and true leaves is so great that one might almost be pardoned for asking whether they can be brought into actual correlation.
So far, indeed, are the cotyledons from agreeing with the forms of the leaves, that the difficulty is to find any clear resemblance. In one species of Ipomoea (T. Pes-caprce) both are, as the name denotes, somewhat like the foot of a goat ; but the leaves vary considerably, and it is probable that the resemblance may be accidental. A clear case is, however, afforded by the Onagrarieas, where in (Enothera and some allied genera the form of the mature cotyledons is evidently related to that of the leaves. Even here, however, the likeness is confined to a basal portion which makes its appearance subsequent to germination, and no trace of it is shown in the coty- ledons themselves when they first appear.
The forms of the cotyledons in many species have been the subject of special memoirs by Tittmann, Irmisch, Wichura, Winkler, Tscherning, and other botanists ; but they have not given any reasons for the variou.s forms assumed.
Klebs, indeed, one of the most recent writers on the subject, in his interesting Memoir on Germination,*
' ' Beitrage zur Morphologie und Biologie der Keimung,' Uhter' such. Botan. Iiist. v. TuMngen, Band i. p. 636.
92
ON SEEDLINGS
refers to this diversity of form, and expressly says tliut these differences are an enigma (' Sind gewiss diese Verschiedenheiten in den Blattformen hinsichtlich ibrer biologischen Bedeutung fiir die Pflanze ein Rathsel "), He observes, however, that on the whole the forms of cotyledons are much simpler than those of leaves, and he suggests that while in some cases perhaps, like the first leaves, they retain the form which cha- racterised the species in bygone ages, we may rather, as a more generally applicable explanation, apply to them the suggestion of Goebel with reference to stipules, and regard them as simplified by arrest.^ Another suggestion has been that cotyledons are ' a survival of the universal foliage of deciduous trees in olden geo- logical days, ere time had differ- entiated them into their present various forms.' Even, however, if this suggestion were the real explanation of the comparative simplicity, it would throw no light on the differences between the cotyle- dons of different species.
Though cotyledons do not present nearly such ex- ' ' Beitrage zur morphoL' &;c. I.G. p. 613,
Fig. 81. — Seedling of Fceni- culumvulgare, Halfnat size.
NAKROW COTYLEDONS
93
tensive variations as leaves, still they do differ con- siderably from one another.
Some are narrow, in illustration of which I may refer to Foeniculum (Fennel) (fig. 81), Coreopsis (fig. 82),
-^^
//
4^
3^
Fig. 82. — Seedling of Coreopsis filifolia. Half uat. size.
Fig. 83. — Seedling of Platanui. Nat. size.
Fig. 84.— Seedling of Acer Pseudo-Platanus. Half nat. size.
and Ferula (in the hollow stalk, or ferule, of which Prometheus brought down fire from heaven), &c., where the ultimate leaves are much divided ; Platanus (fig. 83),
94
ON SEEDLINGS
Fig. 85. — Seedling of CJienO' podium Bonus-Henricus Nat. size.
Fig. 86. — Seedling of Bicinus sangui- neus. One-fourth nat size.
Fig. 87.— Seedling of Impatiena Fig. 88. — Seedling of Beech {Fagui
balaamina. Hall nat. size. sylvatica). Half nat. size.
BROAD COTYLEDONS
96
and Acer (fig. 84), where the ultimate leaves are pal- mate ; and Chenopodium (fig. 85), where they are more or less triangular.
Some cotyledons are broad, in illustration of which I give figures of Castor-oil plant (Ricinus) (fig. 86),
Fio. 89. — Seedling of HippophaS rhamnoides. Half nat. size.
Fig. 90.— Seedling of Rivina
Icevis. Nat. size.
Irapatiens (fig. 87), Beech (Fagus) (fig. 88), Brassica (fig. 11), Hippophae (fig. 89), Rivina (fig. 90), Asy- stasia ' (fig. 91), Rhus typhina (fig. 92), and Flax
' In Asystasia coromandoliana there is an interesting peculiarity. The first pair of leaves of each branch, or at any rate of the lower branches, approximate to the form of the cotyledons.
Fig. 91. — Soedlingof Asystasia coroman- Pig. 92. — Seedling of Bhus tijphina, deliana. Half nat. size. Half nat. size.
Fig. 93. — Seedling of Menispermum canadense. Half nat. size.
Fig. 94. — Seedling of Linum monogy num. Nat. size.
FORMS OF COTYLEDONS
97
(Linum) (fig. 91). We find some species with narrow cotyledons and broad leaves, as Meaispermura (fig. 93), while in others the cotyledons are broad and the leaves
Fig. 95.— Seedling of Olf-a cuspidata. Two-thirds nat. size.
Fig. 96. — Seedling of Hakea acicularis. Half nat. size.
1
Fig. 97.— Seedling of Pink (Dinn- thus Caryofhyllus). Nat. sizu.
Fig. 98.— Seedling of Ccras- tium. Half nat. size.
H
98
ON SEEDLINGS
narrow, as in an Australian Flax (Linum monogynum) (fig. 94), Hakea (fig. 9G), and the Pink (Vianthiis) (fig. 97).
In some cases we find instances of broad and narrow cotyledons in the same family, as in Cerastium (fig.
Fig. 99.— Seedling of Galium saccharatum. Nat. size.
Fig. 100. — Seedling of Geraniujn sanguinemn. Nat. size.
98) and Pink (fig. 97), belonging to the Caryophyl- lacese ; sometimes even in the same genus, as Galium saccharatum (fig. 99) and Galium Aparine (fig. 36).
In some cases the two cotyledons are unequal, as in the Mustard (fig. 11), Cabbage, Radish, Cereus (fig. 134) : in some the two halves of each cotyledon are
FOEMS OF COTYLEDONS
99
unequal, as in the Geranium (fig. 100); or they are otherwise unsymmetrical, as in the Lupine and the Laburnum (fig. 101). Sometimes they are sessile, as in Acer (fig. 84), Hakea (fig. 96), Laburnum (fig. 101), &c. ; sometimes they are supported on petioles, as in Microloma
Pig. 101. — Seedling of LaJbarmim {Cytistis vulgare). Nat. size.
Fig. 1C2. — Seedling of Micro- lotna, X 1^.
(fig. 102), which again are occasionally connate. These differences may occur in very closely allied species ; for instance, in Delphinium Stapliysagria (fig. 104) the cotyledons are sessile, while in D. elatum (fig. 103) they are petioled, and in D. nudicaule (fig. 105) the petioles are connate.
H 2
100
ON SEEDLINGS
Fio. 103. — Seedling of Delphinium elatum Fig. 104. — Seedling of Velphinium Two-thirds nat. size. Staphysagria. Half nat. size.
Fig. 105. — Seedling of Delphinium Fig. 106. — Seedling of Cordia sub' nudicaule. Half nat. size. cordata. Half nat. size.
C, cotyledons.
-i.ii^i.AJ*js*.-ea^igma^ititJiixJii3jLiLiiHm'n~ih<>tsMSsxz'-
FORMS OF COTYLEDONS
101
Fig. 107. — Seedling of Pelargonium australe. Half nat. size.
Fig. 108. — Seedling of Malva moscliata. Nat. size.
Fig. 109. — Seedling of Spathodea cam-
panulata. Nat. size.
Fig. 110. — Seedling of Catalpa Keemp- feri. Half nat. size.
102
ON SEEDLINGS
Fig. Ill- — Seedling ot Eucalyptus so. Hal! nat. size.
Fig. 112. — Seedling of Eschacholtzia califomica. Nat. size.
Pio. 118. — Seedling of Pterocarya caucasica. Nat. size.
Fm. 114. — Seedling of Poterium Sanguisorha, Twice nat. size.
FORMS OF COTYLEDOl^S
103
Generally the cotyledons are entire, but sometimes crenate, as in Cordia (fig. 106), or lobed, as in Pelar- gonium (fig. 107) and Malva (fig. 108). Often they are emarginate, as in Impatiens (fig. 87), Mustard (fig. 11) and Cabbage, Ipomoea (fig. 159), Convolvulus,
B
Fig. 115. — A, fruit of Platanus, longitu- dinal section, x 6. B, transverse section, X 12. xv, woody part ; p, endosperm ; r, radicle ; t, testa.
jxt
Fig. 116. — A, samara of Frax- inus excelsior, nat. size, with one half removed and the seed pulled out : pa, placental axis ; ao, abortive ovules; /, funicle; ra, raphe ; ch, chalaza. B, longitudinal section of seed, X 2 : jp, endosperm ; e, em- bryo. In germination the embryo grows to th^ whole length of the seed.
Galium Aparine (fig. 163), and Eucalyptus (fig. 111). They are sometimes even bifid, as in Eschscholtzia (fig. 112) and Ipomoea dasysperma (fig. 160); trifid, as in the Cress (Lepidium) (fig. 12) ; or in four long lobes, as in Pterocarya (fig. 113). Sometimes auricled at the base,
lOi
ON SEEDLINGS
as in Poterium (fig. 114) andCupliea (fig. 191). Some- times tliey are large ; sometimes small. Generally they are leaf-like ; bat sometimes, as we have already seen in studying germination, they are thick and fleshy.
Narrow Cotyledons.
Let us now begin with such species as have narrow cotyledons, and see if we can throw any light on this characteristic. The problem is simple enough in such cases as the Plane (Plataniis), where we have on the one hand narrow cotyledons (fig. 83) and on the other hand a long narrow seed fully occupied by a straight embryo (fig. 115). Again, in the Ash (Fraxinus) (fig. 116) and
Ursinia (fig. 117), the coty- ledons lie parallel to the longer axis of the seed, which is narrow and elongated. Such cases, however, are comparatively few ; and there are a large number of species in which the seeds are broad and even orbicular, while yet the cotyledons are narrow, as for instance in Chenopodium (fig. 85) and Menispermum (fig. 93).
In these it will generally be found that the cotyle- dons lie transversely to the seed. In Menispermum (fig. 118) the fruit is laterally compressed and horse-
Fi(!. 117. — Achene of Ursinia spe- ciosa. A, longitudinal section, X 24. B, transverse section, x 10.
NARROW COTYLEDONS
105
shoe- shaped, with a crest along the edge ; the seed conforms to the shape of the fruit, and the embryo is curved and linear, the cotyledons being applied to one another face to face, and at right angles to the
Fig. 118. — Menispermum canadense.
A, germinating seedling, x 2. B, vertical section of seed, x 2. C, transverse section of seed, x 4.
Fig. 119. — Chenopodium Bonus-Henricus, x 8. A, seed. B, vertical section of seed. C, transverse section of seed : H, hilum ; B, radicle ; C, cotyledon ; P, endosperm ; T, testa.
plane of the seed, so that their edges touch the walla of the seed at each side.
The seeds of Chenopodium Bonus-Henricus (fig. 119) are reniform, small and black with a crustaceous testa. The embryo entirely surrounds the periphery of
106
ON SEEDLINGS
^
the endosperm except a small portion at the hilum. The cotyledons are linear, plano-convex, lying in the narrow plane of the laterally compressed seed, and
SYCAMORE 107
incumbent, of the same length as the radicle, but slightly wider.
In Spinacia glabra (fig. 120) the narrow cotyledons ultimately attain a length of 30-80 mm.
The Sycamore (Acer Pseudo-Plat anus) (fig. 84) has also narrow cotyledons ; but the arrangement is very different. The fruit (fig. 121) is winged, the seed somewhat obovoid arid exalbuminous, the embryo
Fig. 121. — Acer Pseudo-Platanus. Fruit, nat. size. A, B, embryo, showing two modes of arrangement of cotyledons.
occupying the whole cavity of the seed. Now if we wished to pack a leaf into a cavity of this form, it would be found convenient to choose one of a long strap-like shape, and then roll it up into a sort of ball. This is, I believe, the reason why this form of cotyledon is most suitable in the case of the Sycamore. The mode of folding, however, as shown in fig. 121, A and B, is not always the same. I shall suggest a reason for the difference further on.
In other species the narrowness of the cotyledons is perhaps an advantage in facilitating the exit from
108
ON SEEDLINGS
the seed. See, for instance, the cases of Galium (p. 115) and Symplocos (p. 46).
Broad Cotyledons.
I now pass to species with broad cotyledons. The Acorn, Nut, Scarlet Runner (fig. 122), and Pea afford
Fig. 122. — Phaseolus multiflorus. Sec- Fia. 123. — Bicinus sanguineus.
t.oii of seed parallel and vertical to cotyledons, three-fourths nat. size : a a, auricles ; ar, orillode.
B
A, longitudinal section of seed, nat. size. B, transverse section of seed, nat. size : t, testa ; p, endosperm ; c, coty- ledons ; m, caruncle.
e
Fig. 124. — Hippophae rhamnoides. A, longitudinal section of seed, X 4. B, transverse section of seed, X 4.
Fig. 125. — Moscharia pinnatifida. A, longitudinal section of seed, X 8. B, transverse section of seed, X 8 : a a, auricles.
familiar cases, in which the two broad, fleshy, thickened cotyledons conform to and occupy the whole seed. In the Castor-oil plant (Ricinus) (fig. 123) the seed is
BROAD COTYLEDONS
109
ovoid-oblong, somewhat compressed dorso-ventrally, and beautifully mottled, while the projecting knob (caruncle) at the hilum gives it very much the appearance of a beetle or large tick. The endosperm is abundant, fleshy, white, and surrounds the embryo. The latter is straight, flat, large, central, and white ; the cotyle- dons broad, obtuse-oblong, and approximately following the general outline of the seed. In Hippophae (fig. 124-)
Fig. 126. — Euellia longifolia. A, longitudinal section of seed, x 10. B, transverse section of seed, x 10 : FH, fringe of hairs ; LAu, larger, and SAit, smaller auricle ; R, radicle ; C, cotyledons ; P, endosperm.
we have a somewhat similar case ; but the cotyledons are fleshy and occupy almost the whole of the seed. In Euonymus, again, the seed is obovoid, and slightly compressed laterally. The endosperm is abundant, fleshy, firm, and white, entirely surrounding the em- bryo, which is straight, flat, central, pale green in colour, and extends very nearly from one end of the seed to the other. In the Flax the seed is ovate, obliquely pointed, plano-convex, and laterally much com-
110
ON SEEDLINGS
pressed, placed edgeways on the placenta, and the cotyledons lie parallel to the flattened axis. In other cases the seeds are still more flattened, as in Ailanthus, Passiflora, Cobaea, and Stephanotis.
The Compositae generally have narrow cotyledons. In Moscharia, however, they are somewhat broader, the see.I (fig. 125) being obovoid, with the cotyledons lying the broad way.
Fig. 127. — Cerastium, x 15. C, cotyledons; B, radicle.
In many cases, seeds of the same shape produce cotyledons of very different form.
Compare together, for instance, Ruellia (fig. 126) and Cerastium (fig. 98 seedling, and fig. 127 seed). Both h^ive compressed, nearly orbicular seeds, but in Ruellia the cotyledons are broad, in Cerastium they are narrow. If, however, we make sections of the seeds, the cause of this difference becomes obvious, because in one (fig. 123) the cotyledons lie parallel, in the other (fig. 127) transversely, to the seed.
PINK 111
The form of the cotyledons often differs greatly even in the same family.
The Caryophyllaceae, for instance, afford us an interesting illustration. The cotyledons in this family- are placed with their backs to the placenta, and in most species are narrow, as in Cerastium. In some of them, however, such as the Pink itself (Dianthus) (fig. 97) and Tunica, they are wide.
Now in most genera, as in Stellaria, Spergularia, Cerastium, &c., the seeds are laterally compressed ; the cotyledons consequently lie transversely to the seed, and their width therefore is limited by the thickness of the seed, as in fig. 127. The case is, however, somewhat complicated by the fact that the seed and embryo are both curved.
On the other hand, in the Pink (fig. 128) the seeds are not laterally but dorsally compressed, and attached to the columnar placenta (fig. 1 28, pi) by the middle of the interior face, so that the cotyledons are straight, parallel to the seeds, and have in consequence plenty of room to widen out.
In Solanum the fruit is roundish, indehiscent, and many-seeded. The seeds are kidney-shaped, much compressed laterally, and placed with their narrow edge to the placenta, surrounded by a paler margin, glabrous and white ; the hilum is small, and on the middle of the ventral edge. The embryo is, in the mature seed, much curved, embedded in, but lying near the outer edge of, the endosperm. The radicle occupies the lower
112
ON SEEDLINGS
and narrow part of the seed. The cotyledons are linear, not broader than the radicle, curved, with their tips close to the hilum, and their backs to the placental axis, and at right angles to the plane of the seed, the whole width of which accordingly they occupy, so that they cannot grow any wider. On the other hand, while the fruit of Cestrum is not very unlike that of Solanum, the seeds are very different in shape, being peltate, and
Fig. 128. — Dianthiis CarijojihijUus, x 15 ; pi, placenta ; s, seed ; C, cotyledon ; B, radicle.
(In figs. 127 and 128 the letters a, b, and c, d indicate the directions of the sectiouB.)
moi'e or less obovate, with the broad end towards the apex of the seed, so that the cotyledons have room to widen themselves.
Sometimes we meet with species having both narrow and broad cotyledons, even in the same genus. For instance. Coreopsis filifolia has narrow. Coreopsis auriculata broad, cotyledons. If, however, we examine the seeds we find that those of C. filifolia are naiTOW
GALIUM
113:
or subcylindrical (fig. 129), while those of C. auriculata (fig, 130) are broadly obovate; and as in both cases the embryo fills the seed, this difference sufficiently accounts for the dissimilarity in the cotyledons.
The genus Galium is an interesting case. Here also we find some species with narrow, some with
Fig. 129. — Coreopsis filifolia, x 14. A, acliene. B longitudinal section, C, transverse section. D, embryo.
broad, cotyledons ; but the contrast seems to be due to a very different cause. G. Aparine has broad, G. saccharatum narrow, cotyledons. In G. saccharatum the fruit (fig. 131) is deeply lobed, two-celled, two- seeded, indehiscent, and densely covered with tuber- cles. The seed is globose, deeply hollowed on the
114
ON SEEDLINGS
ventral surface. The embryo (fig. 131, A, r, c) is curved, following the concavity Of the hollow, with the
A B
D
Fig. 130. — Coreopsis auriculata, x 14. A, achene. B, longitudinal section. C, transverse section. D, embryo.
larger part of the endosperm lying towards the peri- phery. The cotyledons are linear and obtuse. Fig. B shows that, so far as the form of the seed is concerned,
GALIUM
IIJ
there is no reason why the cotyledons should not be much broader than they are. The explanation may
--A_.
Fig. 131. — Galium saccharatum. A, longitudinal section of seed, X 8. B, transverse section of Beed, X 8. C, germinating seed- ling, X 4.
Fig. 132. — Galium Aparine. A, longitudinal section of seed, X 8. B, transverse section of seed, X 8. C, germinating seed* ling, X 4.
p.c, pericarp ; r, radicle ; t, testa ; en, endosperm ; c, cotj-ledons.
i2
116 ON SEEDLINGS
perhaps be found in the structure of the pericarp (p.c), which is thick, tough, and corky. It is very impervious to water, and may be advantageous to the embryo by resisting the attacks of drought and of insects, and perhaps if the seed be swallowed by a bird, by protecting it from being digested. On the other hand it does not sjilit open, and is too tough to be torn by the embryo. The cotyledons, therefore, if they had widened, as they might otherwise have done, would have found it impossible to emerge from the seed. They evade the difficulty, however, by remaining narrow (fig. C). On the other hand, in G. Aparine (fig. 132) the pericarp is much thinner, and the embryo is able to tear it open (fig. C). In this case, there- fore, the cotyledons can safely widen without endanger- ing their exit from the seed.
The thick corky covering of G. saccharatum is doubtless much- more impervious to water than the comparatively thin testa of G. Aparine. The latter species is a native of our own isles, while G. sacchara- tum inhabits Algiers, the hotter parts of France, &c. May not, then, perhaps the thick corky envelope be adapted to enable it to withstand the heat and drought ?
In all these species the cotyledons are flat or nearly so, but a large number are enabled to widen themselves l)y being more or less folded. One form of this is afforded by the Radish (Raphanus) and Cabbage (Bras- sica). Fig. 133, A, shows a seed of the Radish,
UNEQUAL COTYLEDONS
117
and, as shown in figs. B-E, the last of which repre- sents a young seedling, the cotyledons are applied to one another face to face, and then folded along the middle.
Unequal Cotyledons.
I now turn to those species in which the two cotyledons are unequal in size.
Flo. 133. — Baphanus saiivus. A, outline of seed, x 4 : in, micropyle ; h, hilum. B, embryo extracted from seed, x 4. C, embryo, x 4 ; vertical section. D, embryo, seen from the side, x 4 : o.c, outer cotyledon; I'.c, inner cotyledon; r, radicle; ^, testa. E, germinating seedling, x 2, showing the cotyledons still folded.
Several of these cases have been discussed by Dar- win,' who attributed the inequality to the fact of ' a
' Movement g of Plants, p. 94.
118 ON SEEDLINGS
store of nutriment being laid up in some other part, as in the hypocotyl, or one of the two cotyledons, or one of the secondary radicles.' I differ with the greatest hesitation from so high an authority; but do not see the connection between the store of food being partly laid up in some other part of the plant and the in- equality between the cotyledons. Why should it affect one more than the other ? I venture to suggest that the difference is due rather to the position of the em- bryo in the ssed, which in some cases favours one cotyledon more than the other. For instance, in many cases the cotyledons are what is called ' incumbent,' that is to say, the radicle is folded upon one of the cotyledons, and in such species the outer cotyledon is often rather larger than the other, as, for instance, in Hesperis matronalis (see ante, p. 32).
The Hemp (Cannabis) and Gaylusea present us with cases more or less resembling that of Hesperis.
In the Mustard (fig. 11), Cabbage, Radish (i2ap/iam<s), and some other Crucifers, the difference is more marked, and is due to a different cause. The cotyledons, as just mentioned, are applied to one another face to face (fig. 133, B-D), and then doubled longitudinally one inside the other. The outer one, therefore, having more space, becomes larger and the petiole is longer. In Cereus (fig. 134) the embryo is much curved, and the cotyledons being thick and fleshy, the inner one is naturally smaller than the outer.
In Nyctagineae the cotyledons were always unequal
ABRONIA
119
in the seeds observed, but in some cases one was nearly aborted. The anthocarp of Mirabilis longiflora is obovoid, closed by constriction above the true fruit, hardened or crustaceous and sculptured on the outer surface. The wall of the fruit itself is thin and mem- branous as is the testa which adheres to it. The embryo is doubled upon itself with unequal-sized, con- cave cotyledons occupying the periphery of the seed and
in oc ic
Fig. 134. — Cereus Napoleonis. Section of seed, x 9 : oe, outer coat of testa ; ic, inner coat of testa.
Fig. 135. — A, embryo of Abronia are- naria, x 6 : ic, smaller cotyledon. B, embryo of A. umhellata, x (5 : ic, right and inner cotyledon ; Ic, left cotyledon ; r, radicle.
enclosing the farinaceous endosperm. The radicle is long, terete, stout and occupies the other side of the seed, and together with the outer cotyledon determines the smaller size of the inner cotyledon by restricting its growth.
In Abronia umbellata (fig. 135, B) the embryo is large, much curved or doubled on itself, and lids outside the endosperm, the two edges of one cotyledon reaching almost to the radicle. The second cotyledon (ic) is minute, generally not more than one-seventh the length
120 ON SEEDLINGS
of the outer one, though in one seed it attained nearly hall' the length of the other. In another species of the same genus, A. arenaria, the smaller cotyledon is re- duced to a mere knob (fig. 135, A, ic).
The cotyledons remain very unequal for a consider- able time after germination. Thus the oblong lamina of the larger one eight days after germination was 8-8'5 mm. long, 4-5-6 mm. wide, and twenty-eight days after germination 14-14*5 mm. long and 9 mm. wide; while that of the smaller one eight days after germination was obovate, 1*25 mm. long, 1 mm. wide, and twenty-eight days after germination obovate, tapering to the base, 14-18 mm. long and 8'75-10 mm. wide above the middle. The petiole of the larger one eight days after germination was 5-12 mm. long, and twenty-eight days after germi- nation 24-28 mm. long ; that of the smaller one eight days after germination very small or almost absent, but twenty-eight days after germination 25-30 mm. long.
The larger cotyledon eight days after germination exhibits a distinct midrib and two lateral nerves near the base, but twenty-eight days after germination the cotyledon becomes rather fleshy, and has the midrib only discernible ; the smaller cotyledf)n at first exhibits a midrib only, and that very indistinctly, but twenty- eight days after germination it is obovate, tapered to the base and trinerved, about the same size as the originally large cotyledon or larger, with a longer or shorter petiole. In all these characters it seems to
EUDIxMENTARY COTYLEDONS 121
behave like a true leaf, but maintains the position of a cotyledon.
Fifty-six days after germination the originally large cotyledon measured as follows: — Petiole 27 mm. long, with the lamina 18-5 mm. long, 11 mm. wide; the originally small one now the largest, with a lamina 23 mm, long, 12 mm. wide, and its petiole 49 mm. long, faintly trinerved.
Cases in which one of the cotyledons is rudimentary also occur in the genera Ranunculus ' and Carum.'^ Cy- clamen also has only one cotyledon, which soon becomes foliaceous and of large size resembling a true leaf. The cotyledon of Cyclamen persicum is cordate and more or less crenate, greatly resembling the first true leaf. In a batch of seedlings, individuals may frequently be found having the cotyledons divided into two reniform , sessile, or stalked pieces resembling a compound leaf with a pair of leaflets. The apparently compound cha- racter is further strengthened by the complete suppres- sion of what should be the midrib of the lamina of the originally and normally simple cotyledon.
In Petiveria octandra (Phytolaccaceae) (fig. 136, A) the cotyledons are very interesting. The one is about 3 cm. long, 1^ broad, oblong, tapering at both ends, and entire. The other is shorter and broader, 2 cm. long by 1^ broad, subcordate, with a large terminal lobe, and one,
' See Irmisch, Beitr. zur rergl. MorpTiol. der Pflamen (Hallo, 18.54).
* Hegelmaier, \ergl. Untersuchungen (Stuttgart, 1878).
122
ON SEEDLINGS
more or less pronounced, on each side. At first I thought this curious want of symmetry must be accidental. It is, however, normal, and is explained by the peculiar form and arrangement of the embryo (fig. 136, B-D) in the seed. The fruit is an achene of peculiar form ; it is
Fio. 136. — Petiveria octandra. A, seedling, half nat. size. B, embryo pai'tly unfolded, x 6. C, outer and shorter cotyledon. D, inner and longer cotyledon.
oblong- linear, subcuneate, and laterally compressed, bifid at the apex, crowned with 2-6 unequal, closely reflexed, acute bristles, and more or less hairy, one-celled, one- seeded, and indehiscent. The seed is oblong-linear, and tapers somewhat, conforming to the interior of the fruit. The embryo is large ; the inner cotyledon is
COEEOPSIS
123
doubled on itself; the outer one is also turned over at the end, and wraps round the narrower one/ not reach- ing, however, to the narrow end of the seed : the terminal lobe of the shorter and broader cotyledon is the part which is folded over, and the lateral lobes, which are much smaller in the embryo, are also due to the fold, as shown in figs. B-D.
Fig. 137. — Coreopsis AtJcinsoniana, x 10. A, lonKitwdinal section of achene. B, transverse section of achene. C, embryo. Pc, pericarp; T, testa; C, cotjledon; -K, radicle.
In Coreopsis Atkinsoniana the seeds are obovate, curved longitudinally, and compressed dorso-ventrally, conforming to the interior of the fruit. The embryo again is slightly bent, following the direction of the seed. Consequently the one cotyledon occupies the inner, the other the outer side of the curve ; and, as shown in fig. 138, the outer one is distinctly larger than the other.
' The narrow cotyledon is sometimes absent.
124
ON SEEDLINGS
In Thunbergia reticulata the cotyledons are unequal, ovate, obtuse, slightly emarginate, and cordate at the base. The larger one is slightly denticulated, and has a curious embossed area in the centre ; the smaller
one is, on the contrary, smooth and entire, or nearly so. The seed is orbicular or oblong ; ex- albuminous, 3-4 mm. in diameter, and compressed, with a cavity on the inner side. The embryo is slightly Fig. 138.— Coreopsis Atkimoniana. curved ; and the cotyledons
Seedling, x 10.
lie with their faces towards the hilum, which is very prominent ; the inner cotyledon is turned up at the edges, and wraps, to a certain extent, round the outer one. The raised or embossed patch in the centre of this cotyledon is due to the inward curva- ture of the testa.
UnSYMMETRICAL COT"kTLEDONS.
In other cases, as in the Geraniums (fig. 140), Schinus (the False Pepper) (fig. 145), Clitoria(fig. 144), Laburnum (fig. 101) Lupines, and others, there is in- equality, not between the two cotyledons, but between the two halves of each cotyledon. In the Geraniums this is due to the manner in which the cotyledons are folded. In the Cabbage and Mustard we have seen that one
GERANIUM
125
cotyledon is folded inside the other ; in the Geranium they are convolute (fig. 139), one half of each being folded inside one half of the other — the two inner halves being the smaller, the two outer the larger ones.
In the last case the seed is oblong or subcylindrical, finely tuberculated, about equally thick at both ends, and fixed to the placenta s> little above its middle. The chalaza is at the lower end close to the base of the carpel. Endosperm is absent, and the curved embryo, which is green, at least while fresh, oc- cupies the whole of the seed.
Fig. 139. — Section through embryo
of Geranium, showing the mode Fig. 140. — Geranium Wallichianu?n. of folding of the cotyledons. Nat. size.
Taking G. Wallichianum (fig. 140) as a type, we find that the radicle is the only part that is curved ; and assuming that the embryo starts from near the micro- pyle above the middle of the seed, then it would grow until it reached the upper end of the seed, when the apex curves downwards until it reaches the base, and here we always find the apex of the cotyledons pointing to the
126 ON SEEDLINGS
base of the carpels or ovary. There still being room for the cotyledons to develop laterally, they commence to coil longitudinally round the sides of the seed in the direction of the sun when the seed is held in its natural position.
The result is that one longitudinal half of each cotyledon grows faster than the other, and grows over, enclosing the smaller half of the other ; and when the embryo is full grown, a transverse section shows each of the cotyledons coiled in the shape of the letter S, but the small coil of each is inserted in the large coil of the other, as in the figure. The coils are, however, closely drawn together so as to fill the seed compactly, and the radicle is incumbent on the back of the cotyledons, but on the ventral face of the seed; there is a small notch or cavity in the centre of the bundle or coil at its lower end into which the rather prominent chalaza fits. When flat- tened out the cotyledons are now transversely oblong or reniform, slightly sinuate at the apex, with the mid- rib excurrent, forming a small tooth ; the base is deeply cordate. The two halves are unequal, the smaller one occupying the interior of the coil. The cotyledons are even already petiolate, and examination shows that this is necessary in order to allow of their becoming con- volute. The inner or upper faces are always applied to one another, but the coils or smaller enfolded half of each prevents the two midribs from coming in contact as they would do if the cotyledons were flat. Hence the necessity for petioles while yet in the seed.
BUCKWHEAT
127
The seed of Buckwheat (Fagopyrum esculoituni) (fig. 141) is large, ovoid, trigonous, rarely with the angles nearly obliterated and rounded. The embryo is relatively very large, variously folded and convolute longitudinally, similar to what occurs in Geranium, except that endosperm is here present and is nearly
Fig. 141. — Fagopyrum esculentum,
A, transverse section of seed, ^ 4.
B, embryo, x 8.
Fig. 142. — Fagopyrum tataricum. Half nat. size.
divided in half by the cotyledons. In transverse section the latter are seen to be twisted in the form of the letter S, with endosperm occupying the sinus. They are deeply but unequally auricled at the base, and the two halves of the lamina are unequal, and the whole therefore unsymmetrical for the same reason as in
128
ON SEEDLINGS
octc
Geranium. The larger halves are outermost and hug the testa, closely following two out of the three angles of the seed and enclosing the smaller halves. The latter retain their smaller size after germination (see also F. tataricum, fig. 142). The radicle is large, fusiform, and completely enclosed by the folds of the coty- ledons.
The embryo in the young state becoming too wide for the interior of the seed, becomes first folded together
longitudinally, and growth continuing the cotyledons become spirally convolute, forming a large open spiral with all the intermediate spaces filled with a white floury mass of endosperm. The cotyledons at length get outside the endosperm and direct their course into two out of the three angles of the seed, thus obtaining the greatest possible width. Looking from the apex of the seed — that is, from the radicle end of the embryo to the lower end of the seed — the cotyledons aro after the first folding all twisted in the same direction, from right to left, or in the opposite direction to the course of the sun, or this direction may be reversed. In con- sequence of this one half of each cotyledon is enclosed by the other half, so that the outer half of each has more
Fro. 143. — Seed of Laburnum vulgare, x 6 : OC, outer coat ; IC, inner coat.
UNEQUAL COTYLEDONS
129
room to develop, and is both larger and has larger auricles.
In the Laburnum (fig. 101) and in Clitoria (fig. 144), on the other hand, the inequality in the two sides of the cotyledon is due to the inequality between the two sides of the seed (fig. 143).
Fig. 144. — Seedling of Clitoria ternatea. Fig. 145. — Seedling of Schinvs Half nat. size. terebinthifolius. Two-thirds
nat. size.
In Heritiera macrophylla (Sterculiaceae) the cotyle- dons fill the seed, which conforms to the shape of the carpels, and the fact that these are somewhat unequal- sided renders the seed and consequently the cotyledons so likewise.
In the Lupines the seeds are obliquely oblong, com-*
130 ON SEEDLIXGS
pressed laterally, and without endosperm, the embryo being large, fleshy, yellowish, and occupying the whole seed. It is doubled on itself, and the cotyledons are folded along the radicle, which nearly equals them in length, with the smaller halves turned towards the radicle, and in such a manner that they and the radicle together occupy one half of the seed, and about equal the larger halves of the cotyledons, which fill the othor.
In Triphasia the inequality is due, partly at any rate, to a different cause. The oval seeds are somewhat flattened, especially on the ventral aspect. The embryo is large, and occupies the whole seed. The cotyle- dons are very unequal in size, and the smaller one is more or less enclosed in the larger. But, in addition to this, there are often, indeed generally, two and sometimes three embryos in each seed ; these differ in size, and the smaller ones often intrude more or less on one of the cotyledons belonging to the lai'ger ones.
A similar condition prevails in the seed of the Orange (Cihms Aurantium). Each contains from two to four embryos sufficiently large to germinate and grow into plants, besides a similar or even greater number of very small ones that are unable to germinate. The large ones are fleshy, colourless, and generally or always have very unequal cotyledons of various shapes owing to mutual compression. There is usually one large cotyledon in each seed, belonging probably to the
UNEQUAL COTYLEDONS
131
true embryo, while all the rest are smaller, and packed on to the face of the large one, completely filling the seed. The large cotyledon is often deeply concave with the others packed into the cavity. Strasburger states that the supernumerary embryos are developed by
Fig. 146. — Schinus Molle, n 3. A, germinating seed. B, seedling four days after germination.
prolification from the tissue of the nucellus bordering on the embryo-sac.
In Schinus (figs. 145, 146) the shallow sinus on one side owes its origin to the indentation of the lower side of the horizontal seed by a thickening or elevation of the receptacle in that region.
K 2
132 ON SEEDLINGS
Crenate Cotyledons.
The vast majority of plants have the edges of the
cotyledons entire. There are some few, however, in
which they are more or less crenate,
as, for instance, in Cordia subcordata
(fig. 106).
In this species the embryo occu- pies the whole of the ovoid-conical seed. There is no endosperm, and
Fio. 147.— Embryo of '^ '
Cordia subcordata, the cotyledons, in Order to occupy
X 2 : r, radicle. ^ ^ ^ , . .
the whole space, are longitudinally folded (fig. 147), thus giving rise to the crenations on the margin.
LoBED Cotyledons.
The great majority of cotyledons are entire, but some are more or less lobed. For instance, those of the Mallow (fig. 108) are broadly ovate, minutely eraarginate, cordate at the base, and three-lobed or -angled towards the apex, with three veins, each running into one of the lobes.
Those of Lavatera and Althaea are similar. The embryo is green, curved, and in a young state occupies a great part of the seed. The cotyledons are applied face to face ; then, as growth continues, the tip becomes curved, and depressed into a median longitudinal furrow, the fold of the one lying in that of the other.
LOBED COTYLEDONS
133
The embryo is of the form shown in fig. 148 : the horn or process r is the radicle ; the rest is the cotyledon, of which the free end / is folded on itself and turned downwards. In this way the embryo fills the seed, leaving a small space between the cotyledons and also between / and r which is occupied by endosperm. Perhaps it may make the arrangement clearer to take
r /
Fig. 148. — Embi} o of Mallow. Enlarged,
Fio. 149
Fig. 150.
Figs. 149 axid 150. — Piece of paper prepared to show the mode and effect of the folding of the embryo.
a piece of note-paper, cut it into the form of an egg (fig. 149), turn the broad portion a h downwards, so that the parts a and b have their under faces turned to one another, and then press down the line e /, and bring the points c and d together, so that c e f and d ef have their upper surfaces together, and the apex / pointing downwards. We shall then have an object shaped as
131 , ox SEEDLINGS
in fig. 150, with a sliai'p point at c, which would not conform to the rounded shape of the seed. If, now, to make it do so, we cut off the point c along the dotted line and then unfold the paper, we shall find that it has the form of the cotyledon of Malva (fig. 108) with a bay or notch on each side. In Erodium the arrangement is somewhat similar, and it seems clear that the lobes are due to the manner in which the embryo is folded.
In Oenothera and some allied species the cotyledons also present a terminal lobe. This, however, is not due to folding. The terminal lobe is the original cotyledon, and the basal portion an altogether subsequent growth, which, moreover, to some extent, assumes the character of the true leaf. I shall presently refer to this interest- ing group in more detail.
The case of Petiveria octandra (fig. 136) has been already described.
Emarginate Cotyledons.
In a great many species the cotyledons are. emarginate, and even in some more or less deeply bifid. No explanation of this has, so far as I know, yet been offered. It is, in fact, as I shall hope to show, by no means always due to the same cause.
One of the simplest cases is that of the Oak, where the thick fleshy embryo occupies the whole of the seed. The chalaza is situated at the centre of one end, at the extremity of the cotyledons, and the walls of the seed
1
EMARGIJSATE COTYLEDONS
135
being at that point somewhat thickened, the cotyledons are slightly pressed in. The same explanation applies to various other species, as, for instance, to the Impatieua (figs. 87 and 151), Poterium (figs. 114 and 152), Cuphea (figs. 153 and 191), and Nettle (Urtica) (fig. 154).
In Helianthus Cncumis the seed itself is slightly
notched at the point where it articulates with the
A
Fig. 151. — Longitudinal, A, and trans- verse, B, sections of seed of Impatiens parviflora, x 10 : PI, plumule ; C, coty- ledons ; r, radicle ; Ch, chalaza.
Fig. 152. — Longitudinal section of Poterium Sanguisorba, X 9 : a, aehene ; re, receptacle ; ^, testa; c/t, clialaza; r,radicle.
receptacle ; and the cotyledons, which, with the rest cf the embryo, eventually occupy the whole interior of the seed, conform to this notch.
In such cases as the Mustard, Cabbage, Radish, and many other Crucifers, the emargination is due to a totally different cause. The seed (fig. 133, A) is oblong, thick, and slightly narrower at one end than
136
ON SEEDLINGS
the other. There is no endosperm, so that the embryo occupies the whole seed, and as this is somewhat deep, the cotyledons, in order to occupy the whole space, are
Fig. 153. — Longitudinal and transverse sections of seed of Cuphea silenoides, x 10 : oc, outer coat ; ic, inner coat.
Fio. 154. — ^Achene of Nettle {Urtica dioica), x 80 : ^, pericarp; E, en- dosperm ; T, testa ; Ch, chalaza.
Fig. 155. — Zilla mya- gr aides. Nat. size.
folded and arranged one over the other, like two sheets of note-paper, as shown in fig. 133, B-E, the radicle being folded along the edge. Fig. 1 33, D, represents
BIGNONIACEiE 137
the embryo a little opened out ; and fig. 1 33, C, a section showing the radicle and the outer and inner cotyledons. To this folding the emargination is due. If a piece of paper be taken, folded on itself, cut into the form shown in fig. A, with the fold along the edge from m to A, and then unfolded, the reason for the form of the cotyledon becomes clear at once. Zilla myagroides, another Crucifer, affords a similar case (fig. 155).
But it may be said that in the Wallflower the seed has a similar outline, and yet (fig. 156) the cotyledons are not emarginate. The reason of this is that in the Wallflower {Cheiranthus Cheiri) (figs. 28 and 29) the seed is more compressed than in the Mustard and Radish, and the cotyledons are not folded ; so that the whole, not the half, of each cotyledon corresponds to the form of the seed.
In the Bignoniacese, again, a large number of species have emarginate cotyledons ; and this would appear also, as for instance in Pithecoctenium Aubletii, to be due to the chalaza, though in a different manner. The seeds themselves are transversely oblong, much com- pressed dorsally, and surrounded on all sides except the base by an extremely thin, transparent, membranous wing, which is traversed by nerves radiating from the central part of the testa, and is uneven at the margin. The raphe is ventral, extending from the hilura to the centre of the embryo. The chalaza is attached to the embryo-sac a little above the middle of the embryo. The radicle is very small, distant from, but pointing to,
188
ON SEEDLINGS.
the hilum. The embryo is straight and flat; the cotyledons grow until they come to tlie point of attach- ment of the chalaza, when they extend forwards on each side, forming two lobes.
In Oroxylum indicum the general structure of the seed is very similar, but the growth of the two lobes of
Flo. 156. — Seedling of Cheiranthus Cheiri. Two-thirds nat. size.
Fig. 157. — Eccremocarpua scaler, x 2
the cotyledons is even more luxuriant, so that they actu'ally overlap. A structure more or less similar occurs in other genera of this family. Compare, for instance, Eccremocarpus scaber (fig. 157).
The emargination is very much deeper in other groups, and due to other conditions, for instance in the
CONVOLVULUS
139
Convolvulacefe. In Convolvulus Soldanella (fig. 158) the embryo, which is eventually very large, lies at first straight in the seed embedded in a clear jelly-like endosperm, and rests on a solid ovate, grooved, white ridge (fig. 158, B and C, a), which rises from close to the micropyle. This tongue-like ridge grows with the embryo. At the opposite end of the seed the raphe and chalaza form a somewhat prominent ridge (?>) pro- jecting into the endosperm. The cotyledons in this
Fig. 158. — Convolvulus Soldanella, x 2. A, embryo. B, section of seed after removal of the dorsal surface, embryo, and endosperm. C, side view of ditto.
stage are plano-convex, applied face to face, orbicular, entire, green, with distinct petioles, 5-nerved, with two lateral, subopposite branches from the midrib some distance below the apex. The plumule and radicle are small. The cotyledons gradually increase in size and grow over the process (a) in which the radicle lies, extending to the ^ apex of the seed, doubling over and abutting against the ridge formed by the raphe and chalaza (h), and thus becoming more and more emarginate at the apex. The notch is therefore due to
140
ON SEEDLINGS
their continuing to grow at the sides after their apex has reached this ridge.
In Ipomoea purpurea (fig. '159, A-C), where the seed is constructed generally on the same model as in Convolvulus Soldanella, the ridge formed by the raphe and chalaza (fig. 159, B and C, h) is more prominent, and consequently the notch of the cotyledon is deeper. Lastly, in Ipomoea dasysperma (fig. 160, A-C) the projecting ridge of the chalaza is still more developed
Fig. 159. — Ipomoea purpurea, x 2. A, embr3-o. B, section of seed after removal of dorsal surface, embryo, and endosperm. C, side view of •ditto.
and reaches nearly to the process which supports the radicle ; the cotyledons are thus precluded from grow- ing in length, and in consequence send out two long wings, so that they are divided almost to the base (fig. 160, A).
In Shorea (Dipterocarpeae), again, the division of the cotyledons is perhaps due to an internal process of the seed. I have not, however, had an opportunity of examining a specimen.
In Eucalyptus (fig. Ill) we have a different case. The
EUCALYPTUS
141
embryo is (with the exception of the petioles) straight or nearly so, fleshy, white, occupying the whole of the seed, and conforming to it in general outline; the
Fig. ICO. — Ipomcea dasysperma, x 2. A, embryo. B, section of seed after removal of dorsal surface, embryo, and endosperm. C, side view of ditto.
cotyledons are deflexed and convolute round the radicle, which the lobes equal in length, while half of one cotyledon lies over half the other ; one half of each consequently lying against the testa. The radicle is stout, fleshy, truncate at the end where it lies against the testa, but otherwise en- tirely enclosed by the folded cotyledons. The true length of the cotyledon is deter- mined by the distance be- tween the end of the petiole and the opposite pole of the seed. The side of the cotyledons, however, being folded back, that part which lies beyond the petioles is enabled by folding round the radicle to widen.^ and this conse-
P'
a.
Fig. 161. — Embryo of Eucalyptus Globidua, one cotyledon being cut away ; p, petiole ; p', cut end of petiole; i.l, inner lobe of coty- ledon ; r, radicle, x 4.
142
ON SEEDLINGS
quently gives the cetyledons their more or less pro- nounced hour-glass shape.
Moreover, in speaking of emarginate cotyledons, we must distinguish between two very opposite cases, of
which I may take Galium Aparine and (Enothera Lind- leyana as illustrations. In the former the cotyledons commence with an entire end (fig. 162), and subsequently,, not as a rule till they have left the seed, become emar- ginate (fig. 163); in CEno- thera Lindleyana, on the con- trary, the cotyledons are at first emarginate, but ulti- mately entire. The em- bryo gradually appropriates all the endosperm, but the supply being largest at the wider end of the seed, this is the last part to be absorbed. In neither of these cases does the emargination appear to be directly due to the struc- ture of the seed, nor to be in itself of any advantage to the plant. It seems rather to depend on the conditions of growth. In Galium Aparine the cotyledon terminates in a peculiar gland,' which would appear in this and
' This gland has already been mentioned by Gravis in his work on Urtica dioica (Brussels, 188fi), p. 13'J. lie olsjrves that about a
1
Pig. 162.— Yoving seedling of Oalium Aparine, x 2.
CRUCIFEKvE
143
other cases, after the emergence of the cotyledon, to stop its growth at that point, and thus to produce the emargination.
In a number of Cruciferge the notch or sinus is not due to any peculiar conformation or structure of the seed, as the cotyledons are, as far as can be seen before germination, entire ; but to the subsequent greater growth of the sides as compared with the organic apex.
Pig 163. — Young seedling of Galium Aparhie, a few days older, x 2.
Such emarginate cotyledons occur in Nasturtium syl- vestre, Arabis Turczaninowii, Cardamine hirsuta, C. grseca, Lunaria biennis, Cochlearia glastifolia. Sisym- brium officinale (fig. 164), and others. The entire (A) and emarginate states CB) are shown in the figures of Sisymbrium officinale.
dozen small stomata occur on it, while they are entirely absent on the rest of the upper surface. He regards it as a water-gland, • un organe destine 4 rem^dier ^ I'exc^s de tension dans I'appareil aquif^re.'
14A
OX SEEDLINGS
Fig. IQi.—Sisijmhrium officinale, x 2.
SENECIO — BKYONIA 145
A few terminal pores or water-stomafa are found at the apex of the midrib.
In Senecio, again, the majority of species have entire cotyledons. In some, however, as in S. eruca?- folius, they are emarginate. Even here, however, they are at first entire (fig. 165), and the emargination does not make its appearance until after germination, when the cotyledons gradually become much widened (fig. 166). Ill fact, S. squalidus, S. viscosus, S. vulgaris, and others, have the cotyledons narrow and entire ; while in S. eructefolius and S. cruentus, where they grow
Fig. 165. — Young seedling of Fig. 166. — Ditto, a few days
Senecio eruccefoHus. older.
more in width than in length, they become emarginate. Among other cases where the cotyledons are at first entire, but after germination become emarginate, may be mentioned some species of Lithospermum.
Bryonia laciniosa (fig. 203) also has the cotyledons emarginate, while in B. dioica (fig. 204) they are entire. They are, however, originally entire in both cases, and the emargination in B. laciniosa is due to the fact that in that species the cotyledons grow much more than in B. dioica. There is no great difference in size between the seeds, those of B. laciniosa being perhaps one-tenth larger. On the other hand, the cotyledons of B
L
146
ON SEEDLINGS
laciniosa attain a length three times greater than those of B. dioica, as shown in the figures. In the genus Tacsonia, the cotyledons are entire in T. Van-Volxemii (fig. 167) and T. Lescheuaultii, and emarginate in T.
Fig. 167. — Tacsonia Van Volxeviii. Half iiat. size.
Fig. 168. — Tacsotiia ignea. Nat. size.
ignea (fig. 168). Here also, however, they are at first entire (fig. 169), and only become emarginate after leaving the seed.
In Berberis Aquifolium the embryo is sun-ounded on all sides by endosperm, being only three-quarters as
TACSONIA— BAREERRY
147
-Jia,
Fig. 169. — Tacsonia ignea, x 12. A, longitudinal section of seed : M it II, position of micropyle andhilum; P, endosperm; Ba, raphe; Ch, chalaza. B, trans- verse section of seed : C, coty- ledons.
B
Jta^
Fig. 170. — Berberis AquifoUum. A, longitudinal section of seed, x 10 : Ch, chalaza ; R, radicle ; T, testa ; t, tegmeii. B, transverse section of seed, x 10 : C, cotyledon ; T, testa ; t, tegmen ; Ba, raphe.
I. 2
148
ON SEEDLINGS
long as the seed, but has nevertheless emarginate cotyledons. The notch is exceptionally large in this
instance, but the struc- ture of the seed dix's not give any explana- tion of its presence. It is moreover often alto- gether absent ; and when present, is apparently due to the slower growth of the organic apex. The seed itself is bluntly triangular in transverse section, and shows the cotyledons lying in the broader plane of the seed, but not always strictly in apposition, a circumstance possibly due to the squeezing produced by the endosperm, while the cavity is a* little wider than the embryo it con- tains. The shape of the seed is due to the mutual pressure of several seeds in a berry or fruit. Either the dorsal aspect of the seed, or the sides, may be the broader according to the position of the same on the placenta. The general
Fio. 171. — Berberis Aquifolium, Nat. size.
DIVIDED COTYLEDONS
149
sLape of seeds in tlie Order is obovoid, oblong, or globose, and they are generally immersed in pulp with- out being subject to much pressure.
In the seedling (fig^ 171) the oblong cotyledons are obtusely pointed.
Divided Cotyledons.
The genus Pterocarya has very curious cotyledons (fig. 180), due to a cause entirely different from any of those we have considered hitherto.
They are bipartite, each primary division narrowing to a cuneate base, and being again deeply divided, so
Fig. 172. — Pterocarya caucasica. A, transverse section of nut, x (S; showing the four hollows c, c, c, c, which are occupied by the four prolongations of the seed. B, seed, x 6.
as to make in all four ultimate, linear-oblong, obtuse, entire segments. In this case the endocarp is thickened and bony, and its cavity is divided at the base into four cells (fig. 172, A, cccc) by the thickening and consequent intrusion of the dorsal and ventral walls.
150
ON SEEDLINGS
^
The seed (fig. 172, B) is conical above, deeply four- lobed below (fig. 172, B, H Z I), one lobe passing into each of the cells of the endocarp. The embryo again follows suit, each cotyledon sending a lobe into each of the four seed-lobes, and thus assuming the peculiar form characteristic of the species.
Pterocarya caucasica flowers with us early in May. The pistil is syncarpous and inferior ; the ovary of two carpels, one-celled, one-ovuled ; the ovule basal, erect, and orthotropons.
Pig. 173.— Fruit of Pterocarya, x 2 : P, perianth ; W, W, wings.
Fig. 174. — Longitudinal section through the flower of Pterocarya caucasica, x.6: B, bract ; B', B', bracteoles ; Ov, ovule ; P, peri- anth ; St, style : S, S, stigmas. (May 23.)
Fig. 174 is a section through the young flower at the end of May, showing a bract B at the base, two bracteoles B' B' at the sides, the ovule Ov^ perianth P, and 88 the two large, spreading, papillose stigmas. The cavity of the ovary is small and nearly filled by the ovule.
By about the middle of June the young fruit has grown considerably in thickness, though not much in length. The ovary and ovule are longer, and at the
PTEROCARYA
151
base of the former the tissue has in two places (fig. 175, Co, Co) become almost colourless from the removal of the protoplasm.
By the end of June the fruit has still further increased in length as weU as in breadth. The growth in length has especially taken place between the base and the uppermost point of attachment of the bracteoles,
|
.Pc |
^-. |
|
..-0 |
|
|
\..-0v |
W^ |
|
\.-Co |
Co-- |
Fig. 175. — Pterocarya caucasica. Longitudinal section of fruit, x 6. (June 25.) Co, Co, two places where the tissue has become colourless ; Pc, pericarp. (Other letters as in fig. 174.)
Fig. 176. — Pterocarya caucasica. Transverse section of fruit, x 8 : W, W, wings ; Pc, pericarp ; Co, Co, Co, Co, four spaces of altered tissue ; Ov, ovule.
which therefore seem to have been carried up. The bracteoles have also increased in size, while the perianth remains unaltered. The two masses of colourless tissue as seen in longitudinal section at the base of the ovary are still solid or unbroken.
The fruit continues to grow rapidly, especially at the base, so that by the end of July the posterior half of the bracteoles seems to be carried still further uj),
152
ON SEEDLINGS
I
distinctly more so than the anterior. The neck of the fruit, on the contrary, has increased considerably in thickness, but scarcely at all in length. Fig. 176 re- presents a transverse section, and on each side of the ovule (Co, Co, Co, Co) are the four approximately cir- cular patches of colourless tissue, which in a longitudinal section appear more elongated. In them the tissue is commencing to disintegrate, while round them, on the
Fig. 177. — Pterocarya cavcasica, X 3 : Va, Va, vascular tissue ; Co', Co', masses of solid cortical tissue; Co, Co, cavities.
Fig. 178. — Pterocarya caucasica. Longitudinal section of more advanced fruit, x 3. (Sept. 1.) T, testa ; En, endosperm ; Em, embi-yo ; P, perianth ; Va, vas- cular tissue ; Co, small cavity not yet filled by the seed.
contrary, it is becoming distinctly sclerenchymatous. From the development of the lower part of the fruit, especially on the posteiior side, the posterior portion of each bracteole appears to be nearly on the summit of the fruit, the anterior portion leing rather lower down. Fig. 177 is a longitudinal section taken on August 8. It passes through two of the masses of loose tissue mentioned above, which now form cavities ; while, on the other hand, the surrounding tissues have become
PTEROCARYA
153
much denser, leaving, however, oval spaces of cortical tissue shown in section at Co', Co'.
The ovary has not materially altered, and the ovule is still very small.
A few days later, however, the ovule has grown considerably and nearly fills the cavities. Fig. 178 shows a longitudinal section taken on September 1. T is the testa, showing that the seed has now assumed its four-lobed form, though it has not yet quite filled
Fig. 179. — Pterocarya caucasica. Transverse section through a more advanced fruit near the base, x 6. (Sept. 21.) T, testa ; C, C, C, C, folds of cotyledons ; B', B', bracteoles or wings ; Ca, small mass of cortical tissue.
the cavities in the fruit. The greater part, however, is occupied by endosperm, the embryo (Em) being still comparatively small.
Fig. 179 represents a transverse section near the base of the fruit, taken on September 21, when the fruit and embryo had attained nearly their full size,
154 ON SEEDLINGS
but had not yet reached maturity. Neither the placenta nor the original true cavity of the ovary is shown in this section, because they were situated at a higher level. The ovule from the first was basal, and the seed, even at maturity, may be looked upon as lying astride the basal placenta, with its four lobes projecting into as many cavities excavated from the originally solid base of the fruit. The testa is shown at T, lining the interior of the cavities and enclosing the variously folded lobes of the cotyledons (C, C, C, G). The walls surrounding the cavities are thick and sclerenchymatous, with exception of the thin outer rind and its appen- dages, the bracteoles or wings, shown at B', B'. The cotyledons of the embryo diverge, one to each side of the fruit, and their lobes pass in pairs into .each of its four cavities. As growth proceeds and the short lobes become too wide for the cavities, they become con- duplicate in order to accommodate themselves to the restricted space and at the same time to fill it. The secondary fission seems intended to facilitate folding, and was probably originally brought about by excessive plication. If the two lobes had been in one piece, the latter would have had to be twice conduplicate longi- tudinally, which would have been difficult to accomplish. The folding is not always on the same plan, as may be seen by reference to the figure.
The Walnut (Juglans). — The fruit of the Walnut differs from that of Pterocarya in several remarkable particulars, and while the cotyledons of Pterocarya are
^
WALNUT 155
leaf-like and aerial in germination, those of the Walnut never emerge from the seed.
Chabraeus long ago remarked on the wonderful richness of nature as displayed in the Walnut, ' pras- sertim miranda figurae luxuria naturam in hoc fructu lusisse certum est.' The Walnut, from its fancied resemblance to a head, the outer woody covering being compared to the skull, and the folds of the cotyledons to the convolutions of the brain, was formerly supposed to be especially efficacious in brain-disease.
In the Walnut (Juglans regia) the ovary is one-celled or imperfectly four-celled, and one-ovuled ; the ovule is erect and orthotropous, with the micropyle superior. The fruit is drupaceous, oblong-globose, crowned with a small point consisting of a three- to five- toothed involucre formed by the union of the bract and bract- eoles, and by the remains of the four-toothed perianth and of the style. The smooth rind is rather fleshy, and beset with submerged glandular dots ; it bursts irregularly when mature. The corrugated endocarp is hard, bony, or brittle, with a spongy lining or inner coat which forms large irregular corrugations internally, and is apparently excavated into four large cavities at the base ; the excavations are continued to the top of the main cavity of the ovary, hollowing out the sides of the endocarp so as to furnish a larger amount of space for the seed than is originally provided for it. In Bentham and Hooker's ' Genera Plantarum ' the base of the endocarp is said to be intruded, imperfectly dividing
156 ON SEEDLINGS
the fruit into two or four cells. The endocarp further consists of two valves or halves, which, however, cannot be separated without force.
The seed is large, strongly and irregularly corrugated, seated on the central and originally basal placenta, which in the mature fruit is about one-third above the base of the cavity of the endocarp ; it is deeply four-lobed at the base, filling the four cavities. The pale brown testa is thin, closely applied to the corrugations of the endocarp externally before the seed becomes dried up, and internally to the lobes of the embryo.
In the young state the endosperm fills the interior of the seed with a clear jelly-like mass, in the apex of which is the small embryo, with the radicle upwards. Gradually, however, the cotyledons grow and eventually absorb the whole of the endosperm, thus filling the interior of the seed, except, of course, the small portion occupied by the plumule and radicle.
We have seen that in the fruit of Pterocarya foiir hollow spaces gradually form themselves in the originally solid fruit, and that into these spaces the seed sends four prolongations, into which again the cotyledons subseq.uently grow. Now in the Walnut a very similar process takes place, only the hollow spaces are much larger and confluent with the ovarian cavity, so that instead of a solid wall with hollow spaces filled by the se(;d, it gives the impression as if the seed were thrown into folds occupied by the wall of the fruit. To occupy these spaces fully, the cotyledons themselves
WALNUT AND PTEROCAEYA
157
were thrown into folds as we now see them. The fruit of Pterocarya is much smaller than that of the Walnut, the ancestors of which doubtless had a smaller fruit. As it increased, the cotyledons became fleshier and fleshier, and found it more and more difficult to make their exit from the seed, until at last they have given up any attempt to do so, Heuce the curious folds, with which we are so familiar, are due to the efforts made by the original leafy cotyledons to fill the interior of the nut.
Comparison of the Fruits of Pterocarya and Juglans. — Thus, then, while essentially similar, the fruits of
Fig. 180. — Pterocarya caucasica. Nat. size.
Fig. 181. — Juglans regia. One- tenth uat. size, n, nut.
Pterocarya and of the Walnut offer several remarkable differences. They recall in some respects the relations between the fruits of the Hornbeam and of the Beech. The fruit of Pterocarya, like that of the Hornbeam, is winged, which is not the case with the Walnut or the Beech j it is in the two former smaller, and a great deal
158 ON SEEDLINGS
harder tlian in the two latter. Again, the cotyledons of Pterocarya are aerial (fig. 180), while those of the Walnut (fig. 181) no longer perform the functions of leaves and never quit the seed.
In the Walnut, as in some other trees, it is an ad- vantage that the seeds should be large rather than numerous. In this way they are able to contain a large supply of nutriment, which suflSces rapidly to carry the young plant above the grasses and other low herbage. These seeds form the food of squirrels and other animals, which accordingly serve to disperse them, and thus perhaps they are enabled to dispense with any other means of transport. Moreover, for such large fruits wings would perhaps be scarcely adequate.
In Pterocarya, on the contrary, the fruits are much smaller, and wings therefore more suitable. Possessing in themselves the means of dispersal, they have no need of offering any attraction to animals. In fact every one which is eaten is so much pure loss. Hence, while the shell of the Walnut is sufficiently hard to protect the seed from the severity of the weather, and from the attacks of small species which would not help in their dispersal, it offers no obstacle to larger animals. That of Pterocarj^a, on the contrary, is very hard and strong, and even the interior portion (the walls and pillars surrounding the four hollows) are of the same charac- ter, while in the W^alnut they are, comparatively, quite soft.
One reason why the similarity of construction in
1
WALNUT AND PTEEOCARYA 159
the two seeds does not at first strike the observer, is that in Pterocarya the lobes of the seed evidently enter the pericarp ; in Juglans, on the contrary, the lobes are so much larger that it rather seems as if the pericarp sent projections into the seed.
That the present condition of the Walnut seedling is not original, we have interesting evidence in the pre- sence of small leaves reduced to minute scales, as in the Oak and many other plants with subterranean cptyledons.
These scales evidently indicate the former presence of actual leaves which are now no longer required. 'J'he curious lobings and foldings of the seed in the Walnut also remind us of the time when the cotyledons were variously lobed and folded so as to occupy the whole space in the gradually enlarging seed. At pre- sent they seem to fulfil no useful function, but are an interesting indication of the past history of the plant.
In Eschscholtzia (fig. 112) the cotyledons are deeply bifid, resembling a hay-fork with two long prongs. In this case we find no such structure of the fruit or seed to account for the peculiarity. My first idea was that such cases might possibly be due to some difference in the endosperm, as occurs in certain Umbellifers, Delphinium, &c., which might have permitted growth more readily in certain directions than in others. Thin sections, however, showed no such differences. Moreover, Schizopetalon Walkeri (one of the Cruciferse) (fig. 183) has the cotyledons as deeply divided as in Eschscholtzia; and as there is no endosperm, but
160
ON SEEDLINGS
the embryo occupies the whole seed, within which the long lobes of the cotyledons wind about more or less irregularly, the division cannot be due to differences in
the endosperm. There are, moreover, other plants, such as the Sycamore (fig. 84) and Hop, where the cotyledons are also narrow, winding, and oc- cupy the whole seed, but are not divided. We must there- fore seek some other explana- tion, and I will suggest the following.
In most of the species which I have examined, when the cotyledons remain in the seed they do not leave the ground. In some cases, however, as in Anona (fig. 182), the hypocotyl is long, stout, and rises in the form of a loop during germi- nation, while the cotyledons, which at first are very small, gradually increase almost to the length and breadth of the seed, throw themselves into undulations, and, it being per- haps on this account impossible
Fig. 182. — Germinating seed- . ..it .i i /•
ling of A7wna. Nat. size. to Withdraw themselvcs from
NARROW COTYLEDONS 161
the seed, are eventually torn from the hypocotyl. In Bignonia insignis, again, the cotyledons, though flat and leaf-like, are unable to emerge from, or at any rate do not emerge from, the seed. This may possibly give us a clue to such cases as Eschscholtzia and Schizopetalon, which, I would venture to suggest, may have reference to the manner in which the cotyledons free themselves from the seed. If this is delayed, the young plant suffers considerably, and indeed often perishes. That the pro- cess is not, however, so simple as might be imagined, may be seen from the interesting case afforded by the Dipsaceae (see ante, p. 50) or Cucurbitaceae, where, in Mr. Darwin's words,' ' the seed-coats are ruptured by a curious contrivance, described by M. Flahaut. A keel or peg is developed on one side of the S'lmmit of the radicle or base of the hypocx)tyl ; and this holds down the lower half of the seed-coats (the radicle beiug fixed into the ground), whilst the continued growth of the arched hypo- cotyl forces upwards the upper half, and tears asunder the seed-coats at one end, and the cotyledons are then easily withdrawn.'
May not the narrowness of the cotyledons in Eschscholtzia and their deep fission be due to a similar cause ? The seed is slightly pyriform, and the radicle emerges from the narrower end. It bursts through the soil in an arch, and instead of leaving the seed-coats in
' Morementt of Plants, p. 102. Bower has pointed out that ia Welwitschia a corresponding process serves to absorb the endosperm, acting in fact as a feeder to the young plant {Quart. Journ. Micr. Sci. vol. xxi).
M
162
ON SEEDLINGS
the earth as usiial, carries them up with it. Then the two arms of the cotyledons separate, widen the orifice, and draw themselves out.
This suggestion seems to be confirmed by the] evidence of Schizopetalon (fig. 183), one of the otherJ
A D
Fig. 183. — Stages in the growth of Beedling of Schizopetalon Walkeri, x 2^.
few cases where the cotyledons are bipartite. Here, also, the radicle emerges through a comparatively small orifice, and the seed-coats, from which the cotyledons seem to have some difiiculty in freeing
OPUNTIA— PINUS -C£ESS
163
themselves, are carried up by the hypocotyl, while eventually the lobes of the cotyledons draw them- selves out one by one. In Opuntia basilaris, again, which differs from O. Labourtiana in having narrow cotyledons, the seed-coats are similarly carried up, and the cotyledons free themselves by divergence. In this species, also, it is interesting that one or both cotyledons
Pig. 185. — Section of seed of Lepi- dium graminifoliian, x 15.
Fig. 184. — Seedling of Pinus rigida, x2.
Fig. 186.— Section of seed of L. sativum, xl5.
are often bifid. Is it possible that the multiplicity of the cotyledons in Conifers (fig. 18-4) can be due to the same cause ?
In Ephedra there is a special membrane, which re- mains attached to the root, and thus prevents the testa from being carried up on the tips of the cotyledons.
The common Cress (Lepidium sativum) (fig, 187), to which I have already referred, is a very interesting
M 2
164
ON SEEDLINGS
case, for while in the other species, at any rate in the other English species, of the genus the cotyledons are entire (see fig. 188), in Lepidium sativum, on the con- trary, each possesses two long, narrow lateral lobes.
Fig. 185 represents a section through the seed of] L. graminifolium, which may be taken as representing the ordinary arrangement in the genus. The seeds, conforming to the shape of the capsule, are somewhat triangular, with the radicle in the narrow end. The embryo occupies the whole of the seed, there being no endosperm. In L. sativum (fig. 18G) the seed is of the same form, but nearly twice as large, and if, therefore, the cotyledons were to occupy the whole additional space, they would become extremel}^ thick. In en- dospermic seeds this would pre- sent no difficulty, as the additional space would be simply filled by endosperm. In Lepidium, however, this device cannot be resorted to ;' but the two lobes just fill up the vacancy.
In the Lime (Tilia) (fig. 189) we have another very interesting case.
The cotyledons are broad, foliaceous, rhomboid- subtriangular, and 5-lobed. The lobes are oblong and obtuse, with a strong nerve running into each, tha
Fig. 187. — Lepidium sativum, x 3.
LIME
165
outer ones always largest and sometimes ovate ; the middle pair always the smallest, and oblong or subu- late.
Fig. 188. — Le^idimn gramini folium , x '2.
The fruit is an ovoid or subglobose woody nut, with five obtuse angles, tomentose with somewhat rufous hairs, one-celled by the rapture of the septa, one-
166
ON SEEDLINGS
seeded, indehiscent, tipped with the persistent base of the style, and attached to a large deciduous bract which serves to disseminate it by the aid of the wind.
The seed is ascending or erect, obovoid or subglobose, deep brown, and smooth, with a firm or crustaceous testa.
Fig. 189. — TUia vulgaris. Seedling. Nat. size.
The endosperm in the mature seed is copious, firm, pale yellow, and homogeneous. There is nothing in any way analogous to the causes which have led to the existence of the lobes in the species previously described.
LIxME
167
The embryo is at first straight ; the radicle is stout and obtuse ; the cotyledons ovate, obtuse, plano-convex, fleshy, pale green, and applied face to face. The latter, however, grow considerably ; and when (fig. 190, A) they meet the wall of the seed bend back on themselves, and then curve round, following the general outline of the seed (fig. 190, B). If anyone will take a common tea-cup and try to place in it a sheet of paper, the
A B
Ch. C
Fig. 190. — Tilia. A, section of fruit, x 4. Pc, pericarp; ^0, aborted seed; OT, testa of perfect seed; jTT, tegmen; P, endosperm ; C, cotyledon. B, embryo, x 8.
paper will of course be thrown into ridges. If these ridges be removed and so much left as will lie smoothly inside the cup, it will be found that the paper has been cut into lobes more or less resembling those of the coty- ledons of Tilia. Or if, conversely, a piece of paper be cut out into lobes resembling those of the cotyledons, it will be found that the paper will fit the concavity of the cup. The case is almost like that of our own hand, which can
168 ON SEEDLINGS
be opened and closed conveniently owing to the division
of the five fingers.
It may be said that the seed of the Sycamore (Acer) is not very dissimilar in form to that of the Lime ; and yet the cotyledons are long, narrow, and strap- shaped, while those of the Lime are rhomboid and iive-lobed ; bat it must be remembered that in the Syca- more the embryo occupies the whole seed, while in the Lime it is embedded in endosperm.
The peculiar lobed form of the cotyledons of Tilia thus enables them, I would suggest, to lie conveniently in the globose seed.
AuRiCLED Cotyledons.
Some cotyledons are markedly auricled. As illus- trations I give Poterium (fig. 114) and Hakea (fig. 96). This form is, I am disposed to suggest, a provision to fill up vacant space in the seed. In the seed of Hippophae (fig. 121) the form of the cotyledon leaves at each side of its base two spaces (2.9), which are occupied by endosperm. In Cuphea (fig. 153), Ruellia (fig. 126), and Poterium (fig. 152), on the other hand, there is no endosperm, and it is consequently an advantage that the cotyledon should develope auricles in order to fill up the space.
If this is the explanation of the auricles, we should expect to find them developed principally in families where the endosperm is deficient. Now in the species 1 have examined, auricled cotyledons occur in 35 genera,
AUEICLED COTYLEDONS
169
belonging to 22 families, of whicli 13 have no endo- sperm, while in 6 of the 9 others it is reduced almost to a film.
Fig. 191. — Three istages in the growth of the seedhng of Ciqjhea silenoidcs.
Fig. 192. — Leuaena glauca. Nat. size.
The argument in the case of Cuphea is further strengthened by the peculiar conformation of the radicle
170
ox SEEDLINGS
(fig. 153), which is three-lobed, the reason being, I would suggest, that the radicle co-operates with the cotyledons in the endeavour to fill up the vacant spaces.
I
Fig. 193. — Acacia Burkittii. Nat. size.
In support of this view I would also observe that the auricles seem to be of little use to the young
SUCCULENT COTYLEDONS 171
plant. For instance, the embiyo of Cuphea, wliile in the seed (fig. 153) has very large auricles, which in the seedling (fig. 191) soon disappear. In Ruellia, again (fig. 126), we have a similar case.
Auricles are well developed in the leguminous species Leucasna glauca (fig. 192) and Acacia Burkittii (fig. 193).
Succulent Cotyledons.
In the family Crassulacege the cotyledons observed are all succulent like the plants themselves. Like the leaves, the cotyledons of Crassula quadrifida (fig. 194) are covered with glands secreting a white substance. They persist for a long time, attain considerable size, and are generally very unequal.
In the Cactus family and the closely allied FicoideaB the succulent, fleshy nature of the adult plant is generally foreshadowed in the cotyledons.
Among the species of Mesembryanthemum they are broadly or narrowly oblong, rounded at the apex, sessile and connate, or perfoliate at the base, and so succulent that no venation is discernible, except sometimes a faint indication of a midrib.
Mesembryanthemum tricolorum (fig. 195) .presents a short and comparatively broad type. The leaves are linear or semiterete, and succulent with the first pair developed close to the cotyledons, or if the seedlings are crowded at some distance from them. The leaves of M. capitatum are also semiterete. The cotyledons of
172
ON SEEDLINGS
1
M. serratum and M. echinatum (fig, 196) closely conform to those of M. tricolorum ; and the primary leaves are subulate, semiterete or obtusely trigonous and succulent.
Pig. 104. — Crassida qiiadnfida. Nat. size.
A modification of the above type is exhibited in the cotyledons of M. cordifolium, which are broadly oblong, foliaceous, and flat but succulent. It is a significant
CACTACE^
173
fact that the leaves are also flat and ovate, or sub- cordate, from the first pair onwards.
Other instances are afforded by Phyllocactus cteno- petalus (fig. 197) and Opuntia occidentalis (fig. 198). In these the leaves are represented by spines, while the leaf-functions are performed by the deep-green fleshy stem.
Fig. 195. Mesemhryanthermitn tricolorum. Nat. size.
Fig. 196. Meseriibryanthenium echinatum. Nat. size. C, cotyledon.
In many succulent plants the hypocotyl becomes much swollen, and the cotyledons more or less reduced in size ; in some cases they may be considered as absent, or at all events functionless, although morphologically represented. The Cactus family will again supply some good examples. In the seeds of Mamillaria longimamma they appear as the edges of a three- cornered cleft, and in the seedling of M. Goodrichii they are no more conspicuous and soon become indis-
174
ON SEEDLINGS
cernible. Similar instances are furnished by Echino- cactus viridescens (fig. 199) and E. Orcutii. The seed- ling of the first-named eleven months after sowing of the seed appears as a small obovoid body represent- ing a short fleshy turbinate hypocotyl bearing minute
Fig. 197. — Phijllocactua ctenopetalus. Nat. size.
tooth-like cotyledons, above which the short stem is even stouter, leafless, and covered with small spines in tufts arranged in five or six longitudinal -i-ows. Seven months after the sowing of the seeds, E. Orcutii is similar to the last, but smaller.
ECHINOCACTUS— CEEEUS
175
The cotyledons of Echinocactus Wislizeni attain some size in the seed, but after germination they become merged in the succitlent axis by the swelling of the latter, so that seven months after sowing they appear as small triangular teeth about 1 mm. long, and of the same breadth, projecting from the sides of the seedling. About four months after sowing, the cotyledons of Cereus tilophorus borne by the clavate
— col.
|
Fig. 198. |
Fig. 199. |
Fig. 200. |
|
•.ntia occidentalis. |
Echinocactua |
Cereus Etnoriji. |
|
Nat. size. |
viridescens. Nat. size. |
Nat. size. |
succulent hypocotyl were only 1-25 mm. long, and
2 mm. broad.
In C. Emoryi (fig. 200) they are short, very fleshy, amalgamated with the hypocotyl and stem, and about
3 mm. long.
Similar forms occur in some succulent members of the family Asclepiadeae.
In Sarcostemma brevistigma (fig. 201) the cotyle-
176
ON SEEDLINGS
dons are oblong-ovate, obtuse, entire, subfleshy and glabrous. They soon drop off. The leaves are all reduced to ovate or subulate, acuminate scales so small as to give the stems a leafless appearance.
In Stapelia variegata (fig. 202) the club-shaped liypocotyl is succulent and compressed the opposite
I
(
Fig. 201. — Sarcostemma brevistigma. Nat. size.
Fio. 202. — Stapelia variegata. Nat. size. C, cotyledon.
way to the cotyledons, which are ovate, obtuse, twisted obliquely and each facing an opposite way, broad, sessile and persistent. The succulent stem is smooth, and made bluntly four-angled by the decurrent bases of the leaves. The simple leaves are small, succulent, and coloured like the stem and hypocotyl.
PETIOLES
177
Petioles.
The cotyledons are sometimes sessile, as in Acer (fig. 84), Hippophae (fig. 89), Hakea (fig. 96), Clitoria (fig. 144) ; sometimes supported on petioles, which in
Fig. 208. — Seedling of Bryonia laciniosa. Half nat. size.
many cases, as in Microloma (fig. 102), attain a con- siderable length.
Occasionally we meet with both sessile and petioled cotyledons even within the limits of the same genus. For instance, Delphinium Staphysagria (fig. 104) has the
N
178
ON SEEDLINGS
cotyledons sessile, while those of D. elatum (6g. 103) have petioles. In Biyonia laciniosa (fig. 203) the coty- ledons are nearly sessile, while those of B. dioica (fig. 204) have petioles.
There is nothing, so far as I know, in the structure of the seed to account for this difference. It is ob- servable, however, that while the cotyledons of Bryonia laciniosa (fig, 203) and those of D. Staphysagria (fig. 104) are raised by the hypocotyl somev,'hat above the level of the ground, those of Bryonia dioica (fig. 204) and of D.elatura (fig. 103) are attached close to the ground. In fact the cotyle- dons are carried up in both cases, but in B. lacinio?a and D. Staphysagria by the hypocotyl- edonary portion of the axis or stem of the plant, in B. dioica and in D. elatum, on the other hand, by their own petioles. I may also refer to Vitis hypoglauca (fig. 205) and V. cebennensis (fig. 206), the former with a short hypo- cotyl and long petioles, the latter with a long hypocotyl and short petioles.
Three species of Eucalyptus form an interesting series in this connection. The cotyledons are about on the same level in each, but in one (E. Globulus) (fig. 207) are shortly stalked, in the second (E. calophylla) (fig. 209) are
Fig. 204. — Seedling of Bryonia dioica. Nat. size.
VITXS -EUCALYPTUS
Fig. 206. — Vitis cehennensls. Half nat. size.
Fjg. ^01— Eucalyptus Globulus. Half nat. size.
ir2
180
ON SEEDLINGS
borne on stalks of moderate length, while in the third (E. marginata) (fig, 209) they have long stalks. The length of the hypocotyl varies inversely as that of the stalks.
In short we may say that the cotyledons are, as a general rale, sessile when they are raised by the growth of the hypocotyl, while they are petiolate when they take
Fig. 208. — Eucalyptus calophylla. Nat. size.
their origin close to the ground. There are no doubt some exceptions ; for instance, in some species of Hedysarum, e.g. H. coronarium (fig. 210), the cotyledons are radical and yet sessile. I have, however, in Algeria often seen seedlings of this group in hot, exposed situations, where they ' held the field ' alone, and being
I
STALKED COTYLEDONS
181
sure of ample heat and light, did not reouire to be raised above the surface.
The opposite exception is perhaps more common, namely, when the cotyledons, though raised, are still petiolate. Here, however, they are probably petiolate
Fig. 209. — Eucalyptus marginata. Nat. size.
for the same reason as the leaves — namely, when the foliage is large, leaf-stalks are an advantage in carrying the lower leaves out of the shadow of those immediately above them. Pentapetes phcenicea (fig. 211) is a good example.
In another species of Delphinium (D. nudicaule)
182
ON SEEDLINGS
n
Pericarp and testa.
Fig. 211. — Seedling of Pentapetes phanicea. Nat. size.
Fia. 210. — Hedysarum coronariiim,
Nat. size, c, cotyledons; r,
primary root.
I
CONNATE COTYLEDONS
183
(fig. 105) the cotyledons are raised well above the surface of the ground on a stem consisting of their own two stalks or petioles, which are connate, though readily
Fig. 212. -Seedling of
Smy7-niu7n perfoliatum.
Half nat. size. C, cotyledons.
Fig. 213. — Podophyllum Emodi. Nat. size. C, cotyledons.
separable from one another. Connate cotyledons also occur in Phlomis tuberosa, Smyrnium perfoliatum (fig; 212), Polygonum Bistorta. Podophyllum Emodi (fig-
184
ON SEEDLINGS
i
213), &c. Gray' observes that the economy of this arrangement is not apparent. Assuming, however, that the elevation of the cotyledons is an advantage, perhaps, as I have suggested, from carrying them above the surrounding herbage, the combination of the two
petioles — reversing the old fable of the bundle of sticks — would, with the same amount of material, give a con- siderable addition of strength.
For instance, in Smyrnium perfoliatum (fig. 212) the hypocotyl is undeveloped, the coty- ledons are oblong-ellip- tic, generally unequal- sided, with long petioles connate into one terete piece for 5-5-'"-5 centi-
-Germinating seedling of
Fig. 214.
Geranium hohemician, x 6.
metres of their length, split a little way at the
base to allow the plumule to emerge, and free in the
upper part.
In this case it is obvious that if the petioles had
been separate, they would have been far too weak to
stand upright, and their length therefore would have
been comparatively useless.
^ Structural Botany, p. 21.
USE OF PETIOLE ^ 185
In Polygonum polystacliyum, again, the petioles are connate, and form a hollow tube through which the leaves pass ; so that the seedling has the appearance of possessing an erect hypocotyl with nearly sessile coty- ledons.
In other cases, however, the existence of petioles apparently has reference to the arrangement of the embryo in the seed.
In the Geraniums, for instance, as has been already mentioned, the cotyledons are folded on themselves, one half of each lying svithin
the other. Fig. 214 repre- /^■bi^ '
sents an embryo partially unfolded, and it will be seen that in the position assumed
by the cotyledons the peti- Fio. 215.— 'Embryo oi Eucalyptus
*^ ^ Globulus, one cotyledon beinj;
oles are necessarily as long cut away : p, petiole ; p', cut end
of petiole ; i.l, inner lobe of
as half their breadth. In cotyledon, x 4. Eucalyptus Globulus, again (fig. 215), the mode of fold- ing of the cotyledons would be impossible but for the petioles.
Lastly, in cases where the cotyledons do not leave the seed, the petioles leave room for the free growth of the plumule, as in Sapindus (fig. 168)
Stipulate Cotyledons.
Stipules are generally absent, but occur in Genipa and Psychotria, two genera of Rubiaceae, a stipulate family par excellence.
%:
Fig. 216.~Genit)a clusiifolia. Nat. size
STIPULATE COTYLEDONS
187
In Grenipa clusiifolla (fig. 216) one stipule of each cotyledon unites with the adjacent stipule of the oppo- site cotyledon, forming one ovate, acute piece (S) which is colourless and subscarious.
I
y I
Pig. 2n.—Psychotria sp. Nat. size.
188 ON SEEDLINGS
Similarly in Psychotria (fig. 217) the stipules are united so as to form one ovate, acuminate piece (»S^ between each petiole, soon becoming dry, brown, and scarious.
Size of Cotyleoons.
It is hardly necessary to say that the size of the plant does not determine that of the cotyledons. Winkler has pointed out that the largest of our Nettles has the smallest cotyledons.' It is, on the other hand, natural that large seeds, as a general rule, produce large cotyledons. This is, however, by no means a complete explanation. There are many cases in which the cotyledons grow considerably after quitting the seed. In the wonderful genus Welwitschia it was at one time supposed that the two great leaves were persist- ent cotyledons. This view is now abandoned. In many of the ^lonocotyledons, however, the cotyledons acquire a considerable length. I have already had occasion to allude to cases among the Dicotyledons in which the cotyledon continues to grow for some time after quitting the seed.
i The cotyledons of a species of Capparis (fig. 218) from Melbourne were large, foliaceous and evergreen, and persisted for more than a year on the plant.
A very curious case presents itself in the different species of Streptocarpus (Gesneracese). The cotyledons
• ' Ueber die Keimblatter der deutschen Dicotylen,' Verh. Bot. Ver. Braiidenhurg, 1874, p. 11.
SIZE OF COTYLEDONS
189
are quite normal for some time after germination. They are small or even minute, orbicular, entire, sessile, and thinly glandular-hairy, or they develop a
Fia. 218.— CZoppam. Nat. size.
190
ON SEEDLINGS
very short petiole and become roundly ovate or trian- gular, showing a midrib, but no other venation. The
Fig. 219. — F!i'rnptocarpus Bexii, var. parrijiorua. A, B, C, different stHge» of heeJling, x 6. D, older seedling, nut. size : F.fj, first leaf; S.L, second leaf ; C, persistent cotylfdon. E, seedling with three cotyledons, from one of which the first leaf is developed, x 5.
different stages of growth after germination are well shown by S. Rexii, rm-. parviflorus (fig. 219). There are
STREPTOCARPUS l&l
two or not infrequently three cotyledons. The first leaf is developed from one of them by intercalary growth which sets in at the base of the lamina and continues for a considerable time. The original true lamina of the cotyledon is carried up on the apex of the leaf, and the tip is therefore the oldest part. Finally it withers and dies away, and the apex of the leaf dies away from above downwards in the same fashion. In the early stage the leaf as a whole becomes obovate, then broadly ovate and obtuse. It retains this shape for some time ; but ultimately becomes oblong or strap-shaped with a more or less cordate base, attaining a length of six to ten inches or more. This variety develops a rosette of leaves, the second one of which seems to arise from the short, thick, fleshy petiole of the first. S. Rexii, var. floribundus, is a strong- growing form of the species. In its early stages it behaves in the same way as the last, the first leaf being developed from one of the cotyledons by inter- calary growth at the base. A rosette of leaves is ultimately formed, and when the plants are strong the leaves vary from six to twelve inches or more in length. The cotyledons of S. Dunnii are also perfect and normal in the early stages, small, rotund and entire. The only leaf the plant produces is developed from one of the cotyledons and becomes oblong, narrowed toward the apex, lobulate and dentate, cinereous above from the presence of grey hairs, and rusty beneath with reddish ones ; it varies from twelve to thirty-six inches
192
ON SEEDLINGS
ia length, and nine to twenty inches in width at the broadest part.
Some of the Onagrarieae have seedlings with very curious cotyledons. For instance, I was greatly puzzled by the seedling of CEnothera bistorta, in which (fig. 231) the cotyledons were long and linear, suddenly widening
Fig. 220. — Very young seedling FiG. 221. — Eucharidium grandiflorum: of Eucharidimn grandi- 10 days after germination, x 3.
fiorum, X 3.
at the end into a large orbicular expansion, which gives them a very peculiar appearance.
In Eucharidium grandiflorum (fig. 222) the form of the cotyledon might not unnaturally be supposed to be a case similar to that of Malva. In reality, however,
EUCHARIDIUM 193
the explanation is very different. In Eucharidium the lobes have nothing to do with the arrangement of the embryo in the seed. The young plant, indeed, imme- diately after germination, presents no trace of them. The cotyledons, when they first emerge from the seed (fig. 220), are oblong-orbicular, sessile, cordate or auricled at the base, and emarginate at the apex, with a small purple tooth in the notch : they grow rather rapidly, become shortly petioled, and develop one or two lateral, incurved nerves on each side of the mid- rib.
In the next stage, about eight days after germina- tion, they exhibit a very slight constriction near the base of the cotyledons, with a small obtuse tooth. This basal portion increases much more rapidly, while the growth of the terminal portion (which is, in fact, the original cotyledon) becomes gradually arrested. The tooth becomes more marked, and by the tenth day (fig. 221) the new portion is obtusely bidentate or crenate, and nearly equals the original cotyledon in size.
In its final form (fig. 222) the new portion is both broader and longer than the true cotyledon, and differs from it not only in the crenations, but in the possession of a more conspicuous midrib and rather stiff hairs. Not only is this basal portion interesting in its mode of development, but also from its similarity to the sub- sequent leaves. In fact, as fig. 222 shows, it may be
194
ON SEEDLINGS
:rJ
I
Pig. 222. — Eucharidium grandiflorum : showing final form of cotyledons. Nat. size.
CLARKIA
195
said that we have a compound structure formed of a leaf at the base, terminated by the cotyledon.
If, indeed, this species stood alone, we might regard the resemblance as accidental ; btit we find a very similar growth in other allied species.
In Clarkia pulchella the cotyledons immediately after germination closely resemble those of Eucharidium grand iflorum (tig. 220). In a short time they become broadly ovate, emarginate, suddenly narrowed, and
Fig. 223. — Clarkia pulchella. Two-lhirds nat. size.
rounded at the base. In this case there is no great change of form ; but while the margin of the original cotyledon is glabrous, that of the new growth and of the true leaves (fig. 223) is finely ciliate.
o 2
196
ON SEEDLINGS
In Clarkia gauroides the cotyledons immediately after germination are oblong-orbicular, minutely emar- ginate, with a small tooth in the notch, slightly auricled
Fig. 224. — Clarkia gmiroicles, x 2.
at the base, and sessile, with a scarcely discernible midrib.
They then enlarge, and three days after germination (fig. 224, A) become orbicular or broadly obovate, sub- cuneate at the base, and shortly petiolate.
CLARKIA
197
A new growth now occurs at the base of the cotyle- dons, which is at first narrower and soon becomes con-
I
Fig. 225. — Clarkia gauroicles, > 2.
193
ON SEEDLINGS
spicuous by the presence of a small tootli on each side. (Fig. 224, B. Five days after germination.)
The new growth elongates, bearing two to four teeth on each side, and the whole cotyledon becomes oblong, with a broad emarginate upper half consisting of the true cotyledon, and a basal narrower half which is truly foliar with the characteristic white and reddish midrib of the leaves, and the marginal teeth. (Fig. 224, C. Ten days after germination.)
Finally the cotyledons become broadly ovate, cuneate at the base, petiolate, and in all or most cases more or less distinctly alternate ; the lower part or new growth has four to six small obtuse teeth on each side, and is broadly subelliptic in outline, and minutely ciliate at the margin. The upper part or true cotyledon is comparatively small, suborbicular, and emarginate, with a tooth in the notch. (Fig,
225. Eighteen days after ger- mination.)
In Clarkia rhomboidea (figs.
226, 227) the cotyledons are at first orbicular, entire or faintly
emarginate with a prominent apical tooth, sessile, and like the hypocotyl purplish, glabrous, but soon become minutely and papillosely pubescent.
Fig. 226 shows their appearance nine days after germination. The constriction below the middle denotes the line between the upper part or true cotyle-
FiG. 226. Clarkia rhomboidea, x 3
CLAEKIA
199
don, which is minutely pubescent, and the lower part or new growth, which is much more conspicuously pubescent with a pale pink or purplish midrib ending in a broader purple blotch on the base of the true cotyledon.
Fig. 227. — Clarkuz rhomboidea, x 2.
The lower foliar portion soon becomes broader atd longer than the upper cotyledonary part, which is also only very slightly pubescent.
Thirty-five days after germination (fig. 227) the cotyledons are oblong, distinctly petiolate, constricted jiear the apex at the union of the foliaceous and cotyle- donary part, emarginate with a minute tooth in the
200
ON SEEDLINGS
notch, cuneate at the base, and with a distinct midrib throughout, minutely pubescent all over the upper sur- face, but more sparingly on the under surface, which is reddish-purple.
Fig. 228. — Clarkia iniegripetala, x '2.
As seen from the figure, there is a strong resem- blance between the foliar portion of the cotyledons and the leaves which immediately succeed these.
In Clarkia integripetala the cotyledons imme- diately after germination are shortly oblong, emarginate
CLARKIA 201
with a minute tooth in the notch, shallowly auricled at the base, sessile and glabrous.
The day after germination they have become trans- versely oval, and a shallow semicircular depression and a grey midrib have appeared at the base of each.
Next day a tooth on each, side near the cuneate base denotes the point where the true cotyledon ends, and the foliar portion begins, and four days after germina- tion they are obovate and very shortly petiolate ; the upper part transversely oblong, emarginate, with the apical tooth now almost obsolete : the lower rotund- subcuneate with one to two small teeth on each side, shorter and narrower than the upper, and minutely and papillosely pubescent.
Eight days after germination the cotyledons are as a whole oblong. The lower part has become broadly oblong with three to four minute teeth on each side, and a distinct reddish or silvery midrib, and is suddenly narrowed into a distinct petiole.
Seventeen days after germination (fig. 228) the lower part is broadly ovate with three to five teeth on each side and one to two pairs of very faint somewhat curved nerves running towards them, and a silvery mid- rib terminated above by a transverse brown line. It is papillosely pubescent all over. The upper part is transversely oblong or oval, emarginate, with a tooth in the notch, and minutely and much less conspicuously pubescent than the lower.
In Oenothera stricta the cotyledons immediately
202
ON SEEDLINGS
after germination are oblong, obtuse, slightly auricled at the base, otherwise entire ; sessile, thinly glandular- pubescent on the upper surface, and ciliate. By con- tinued intercalary growth at the base they eventually
Fig. 229. — CEnothera stricta : 80 days after germination. Nat. size.
become (fig. 229) spathulate, obovate or oblong-obovate, and obtuse, with a tooth on each side, indicating the point of union of the original cotyledon and the new growth; the lower part has a distinct midrib, tapers
(ENOTHERA
203
much at the base, and is glabrous ; the puberulous- pubescent petioles are connate at the base.
The first leaves are alternate, lanceolate, obtuse, tapering to the petiole, obsoletely and distantly toothed
Fig. 2^0.—(Enothera bistort a, \ S.
at the margins, and, like the cotyledons, glabrous with pubescent petioles.
In Qj]nothera bistorta immediately after germina- tion the cotyledons are oblong, obtuse, entire and
204
ON SEEDLINGS
FxG. 231.—CEnothera historta.
sessile with a few long, scattered, glandular hairs, especially at the base.
Six days alter germination the base has become elongated petiole-like, and glandular-pubescent, sud- denly narrowed to a short petiole ; the upper half re- mains rotund and glabrous except at the base and possibly a few short hairs underneath.
Eight days after germi- nation (fig. 231) they are much longer; the upper true cotyledonary part is larger, but otherwise unaltered, while the foliar basal and narrow part has become linear, sometimes with a minute tooth on either side.
The lower portion elongates greatly, and in the ultimate stage of tho cotyledons (fig. 230) is linear, tapering at the base into a petiole. The very short upper part or true cotyledon is roundish or oval, entire, glabrous, and without an evident midrib. The lov^ er part terminates abruptly below the upper, and is linear, tapering gradually to the base, with one or two minute and distant teeth on each side. It is thinly hairy, and greyish-green in colour with an evident midrib sunk on the upper surface, and prominent beneath.
The first eight leaves are linear, obtuse, tapering at the base into tlie petiole, thinly silky with adpressed pubescence, minutely and distantly toothed at the
(ENOTHERA 205
margin, greyish-green, and more or less distinctly marked near the margin with black dots, with a very distinct, colourless midrib, flattened above and prominent beneath. They bear, in fact, a close resemblance to the lower portion of the cotyledons.
CEnothera macrantha (fig. 232) affords a similar instance.
Here, therefore, we have an interesting group in which at first the cotyledons are very similar, but by
Fig. 232. — (Enothera macrantha, x 3.
subsequent growth at the base develop into several dis- tinct types, in each case closely resembling the leaf characteristic of the species. We can therefore have little, if any, doubt that this growth is influenced by the form of the leaf.
A similar intercalary growth of the cotyledon occurs in Mimulus luteus (fig. 233) aii3 in Linaria. The cotyledons of the former are triangular and inci-
206 ON SEEDLINGS
piently trifid when full grown. The terminal tooth represents the original lamina of the cotyledon soon after germination. The seedling itself is dimorphic according to circumstances. If it has plenty of room
Pig. 233. — Mimulus luteus. Nat. size.
to develop in the seed-bed the primary leaves are triangular, while the internodes of the stem are hardly developed ; but if the seedlings are crowded, the primary internodes elongate considerably and the leaves become oval.
Relation between Cotyledons and Succeeding Leaves.
It is but rarely, however, that we can trace any con- nection between