October i, 1896] 



NATURE 



539 



til L-flTcct :\ passage over the gulf, Ijiu the gulf is still unliridged. 

 I cannot see anywhere the slightest indication of anything like 

 an intermediate form between the spore-hearing plant of the 

 Pteridophyta and the spore-bearing fruit of the Bryophyla. The 

 plant of the Pteridophyta is sometimes small and simple, but the 

 smallest ami simplest seem just as unlike a br)ophytic sporo- 

 gonium as the largest and most complex. On the side of the 

 moss group, Anthoceros has been often cited as a form showing 

 a certain approach towards the Pteridophytes, and Prof, Camp- 

 liell in particular has developed this idea with remarkable 

 ingenuity. An unprejudiced comparison, however, seems to me 

 to show nothing more here than a very remote parallelism, not 

 suggestive of affinity. 



There is no reason to believe that the Bryophyta, as we know 

 them, were the precursors of the vascular Cryptogams at all. 

 There is a remarkable paucity of evidence for the geological 

 antiquity of Bryophyta, though man of the mosses at any rate 

 would seem likely to have been preserved if they existed, 

 Brongniart said, in 1849, "The rarity of fossil mosses, and their 

 complete absence up to now in the ancient strata, are among the 

 most singular facts in geological botany " (" Tableau des Genres 

 de \'egetau\ Fossiles, p, 13) ; and since that time it is wonder- 

 ful how little has been added. Things seem to point to both 

 Pteridophyta and Bryophyta having had their origin far back 

 among some unknown tribes of the Algs. If we accept the 

 homologous theory of alternation, we may fairly suppose that 

 the sporophyte of the earliest Pteridophyta always possessed 

 vegetative organs of some kind. The resemblance between the 

 joung sporo|ihyte and the prothallus in some lycopods indicates 

 that at some remote period the two generations may not have 

 been very dissimilar. At least some such idea gives more satis- 

 faction to my mind than the attempt to conceive of a fern-plant 

 as deriveil from a sterilised group of potential spores. 



The Bryophyta may have had from the first a more reduced 

 sporophyte, the first neutral generation having, in their 

 ancestors, become more exclusively adapted to spore-producing 

 functions. I must not omit to mention the idea that the 

 Bryophyta, or at any rate the true mosses, are degenerate 

 descendants of higher forms. The presence of typical stomata 

 <jn the capsule in some cases, and of somewhat reduced stomata 

 in others, has been urged in support of this view. It is 

 possible ; but if so, from what have these plants been reduced ? 



Few people, perhaps, fully realise how absolutely insoluble 

 such a problem as we have been discussing really is. I say 

 nothing as to the mosses, which may have arisen relatively late 

 in geological historj'. The Pteridophyta, at any rate, are known 

 to be of inconceivable antiquity. Not only did they exist in 

 greater development than at present in the far-off Devonian 

 period, but at that time they were already accompanied by 

 highly organised gymnospermous flowering-plants. Probably 

 «e are all .igreed that Gymnosperms arose somehow from the 

 vascular Cryptogams. Hence, in the Devonian epoch, there 

 had already been time not only for the Pteridophyta themselves 

 to attain their full development, but for certain among them to 

 become modified into complex Phanerogams. It would not be 

 a rash assuinption that the origin of the Pteridophyta took place 

 as long before the period represented by the plant-bearing 

 Devonian strata as that period is before our own day. Can we 

 hope that a mystery buried so far back in the dumb past will be 

 revealed. 



It will be understood that I do not wish to assume the role of 

 partisan for the homologous theory of alternation. Possibly the 

 whole question lies beyond human ken, and partisanship would 

 lie ridiculous. But I do wish to raise a protest against anything 

 like a dogmatic statement that alternation of generations must 

 have been the result of the interpolation of a new stage in the 

 life-history. Let us, in the presence of the greatest mystery in 

 the morphology of plants, at least keep an open mind, and not 

 tie ourselves down to assumptions, though we may use them as 

 working hypotheses. 



HiSTOl.Oi-.ICAL ClI.\RACTI!RS OF THE TWO GENERATIONS. 



There is one histological question upon which I must briefly 

 touch because it bears directly on the subject which we have 

 been considering, I .shall say very little, however, in view of 

 the forthcoming discussion. 



It is now well known that in animals and in the higher plants 

 a remarkable numerical change takes jilace in the constituents 

 of the nucleus shortly before the art of fertilisation. The 

 change consists in the halving of the number of chromosomes, 



NO. 1405, VOL. 54] 



those rod-like bodies which form the essential part of the 

 nucleus, and are regarded by Weismann and most biologists as 

 the bearers of hereditary qualities. Thus in the lily the number 

 of chromosomes in the nuclei of vegetative cells is twenty-four ; 

 in the sexual nuclei, those of the male generative cell and of 

 the ovum, the number is twelve. When the sexual act is 

 accomplished the two nuclei unite, and so the full number is 

 restored and persists throughout the vegetative life of the next 

 generation. The absolute figures are of course of no import- 

 ance ; the point is, the reduction to one half during the 

 maturation of the sexual cells, and the subsequent restoration of 

 the full number when their union takes place. I say nothing as 

 to the details or the significance of the process, points which 

 have been fully dealt with elsewhere, notably in an elaborate 

 recent paper by Miss E. Sargant. 



Now, in animals (so far as I am aware) and in angiospermous 

 plants the reduction of the chromosomes takes place very 

 shortly before the differentiation of the sexual cells. Thus in 

 a lily the reduction takes place on the male side immediately 

 ptior to the first division of the pollen mother-cell, so that four 

 cell-divisions in all intervene between the reduction and the 

 final differentiation of the male generative cells. On the female 

 side the reduction in the same plant takes place in the primary 

 nucleus of the embryo-sac, so that here there are three divisions 

 between the reduction and the formation of the ovum. I 

 believe these facts agree very closely with those observed in the 

 animal kingdom, and so far there is no particular difficulty, for 

 we can easily understand that if the number of chromosomes is 

 to be kept constant from one generation to another, then the 

 doubling involved in sexual fusion must necessarily be balanced 

 by a halving. 



There are, however, a certain number of observations on 

 Gymnosperms and archegoniate Ciyptogams which appear to 

 put the matter in a different light. Overton ("Annals of 

 Botany," vol. vii. p. 139), first showed that in a Cycad, 

 Ceratozamia, the nuclei of the prothallus or endosperm all have 

 the half-number of chromosomes. Here then the reduction 

 takes place in the embryo sac (or rather its mother-cell), but a 

 a great number of cell-generations intervene between the reduc- 

 tion and the maturation of the ovum. In fact the whole female 

 oophyte shows the reduced number, while the sporophyte has 

 the full number. The reduction takes place also in the pollen 

 mother-cell. Further observations have extended this conclusion 

 to some other Gymnosperms. 



In Osmunda among the ferns there is evidence to show that 

 reduction takes place in the spore mother-cell, and that the 

 sexual generation has the half-number throughout. Prof. 

 Farmer has found the same thing in various liverworts, and 

 shown that the reduction of chromosomes takes place in the 

 spore mother-cell ; and his observations of cell-division in the 

 two generations have afforded some direct evidence that the 

 oophyte has the half-number and the sporophyte the full number 

 throughout. Prof, Strasburger fully discussed this subject 

 before Section D at Oxford (see "Annals of Botany," vol, viii. 

 p. 281), and came to the conclusion that the difference in 

 number of chromosomes is a difference between the two genera- 

 tions as such, the sexual generation being characterised by the 

 half-number, the asexual by the full number. 



The importance of this conception for the morphologist is 

 that an actual histological difference appears to be established 

 between the two generations ; a fact which would appear to 

 militate against their homology. Some botanists even go so 

 far as to propose making the number of chromosomes the 

 criterion Ijy which the two generations are to be distinguished. 

 Considering that the whole theory rests at present on but few 

 observations, I venture to think this both premature and objec- 

 tionable ; for nothing can be worse for the true progress of 

 science than to rush hastily to deductive reasoning from im- 

 perfectly established premises. 



The facts are certainly very difficult to interpret. Those who 

 accept the antithetic theory of alternation suppose the sexual 

 generation to be the older, and that in Thallophytes the plant is 

 always an oophyte, whether "actual" or "potential." Hence 

 they believe that in Thallophytes the plant should show through- 

 out the reduced number of chromosomes, reduction hypothetic- 

 ally taking place immediately upon the germination of the 

 oospore. If this were true it would lend some support to the 

 idea of the intercalation of the sporophyte, but at present there 

 is not the slightest evidence for these assumptions. On the 

 contrary, in the only Thallophyte in which chromosome-counting 



