September 24, 19 14] 



NATURE 



103 



part archaic characters, I propose to sketch before 

 sou \-er>- briefly the points of interest which the more 

 notable of these archaic types present. Some justifica- 

 tion may be found for my doing so because nearly all 

 of them have been submitted to detailed study in my 

 laboratorv- in Glasgow, and much of the work has 

 been done upon material supplied to me by your own 

 botanists. I take this opportunity of offering to them 

 collectively my hearty thanks. 



The tenure by Dr. Treub of the office of director 

 >f the Botanic Gardens of Buitenzorg, was rendered 

 famous by his personal investigations, and chiefly by 

 his classical researches on the Lycopods. These were 

 followed up by other workers, and notably by Bruch- 

 rnann ; so that we now possess a reasonable basis for 

 comparison of the different types of the familv as 

 regards the prothallus and embrvology, as well as of 

 the sporophyte plant ; and all these characters must 

 be brought together as a basis for a sound conclusion 

 as to their phyletic seriation. The most peculiar 

 living Lycopods are certainly Isoetes and Phyllo- 

 glossum, both of which are found in Australia. The 

 former need not be specially discussed here, as it is a 

 practically world-wide genus. It must suffice to say 

 that it is probably the nearest living thing to the fossils 

 Lepidodendron and Sigillaria, and may be described as 

 consisting of an abbreviated and partially differentiated 

 Lepidostrobus seated upon a contracted stigmarian 

 base. 



But Phylloglossum, which is peculiar to the Austra- 

 lasian region, naturally claims special attention. The 

 plant is well known to botanists as regards its external 

 features, its annual storage tuber, its leafy shoot with 

 protophylls and roots, and its simple shaft bearing the 

 short strobilus of characteristic Lycopod type. But its 

 prothallus has never been properly delineated, though 

 it was verbally described by Dr. A. P. W. Thomas in 

 1901 (Proc. Roy. Soc, vol. Ixi.x., p. 285). Perhaps 

 the completed statement may have been reserved as a 

 pleasant surprise for this meeting. But the descrip- 

 tion of thirteen years ago clearly shows its similarity 

 to the type of Lycopodiinn cernutim. The sporophyte 

 compares rather with L. intindatum. Both of these 

 are species which, though probably not the most primi- 

 tive of the genus, are far from being the most ad- 

 vanced. As all botanists know, the question of the 

 position of Phylloglossum chiefly turns upon the view 

 we take of the annual tuber and its protophylls. 

 Treub, finding similar conditions in certain embryos 

 of Lycopods, called it a "protocorm," and believed 

 that he recognised in it an organ of archaic nature, 

 which had played an important part in the early estab- 

 lishment of the sporophyte in the soil, physiologically 

 " idependent of the prothallus. I must not trouble you 

 liere with the whole argument in regard to this view. 

 ?acts which profoundly affect the conclusion are those 

 ^howing the inconstancy of occurrence of the organ. 

 Holloway has recently described it as of unusual 

 ize in your native L. laterale, as it is also in L. cer- 

 juittn. But it is virtually absent in those species 

 which have a large intraprothallial foot, such as L. 

 clavatum, as well as in the genus Selaginella and in 

 Isoetes. In L. selago, which on other grounds ap- 

 Bars to be primitive, there is no "protocorm." Such 

 |[acts appear to me to indicate caution. They suggest 

 'lat the "protocorm" is an opportunist local swelling 

 »f inconstant occurrence, which, though biologically 



iportant in some cases, is not really primitive. 



If this is the comparative conclusion, then our view 

 rill be that Phylloglossum is a type of Lycopod which 



IS assumed, perhaps relatively recently, a very prac- 

 ical mode of annual growth. Related, as it appears 



be on other pomts, with the L. inundatutn group of 



;cies, it has bettered their mode of life. L. inun- 

 ^atum dies off each year to the verv tip of its shoot, 



NO. 2343, VOL. 94] 



so that only the bud remains to the following season. 

 It is notable that Goebel has described long ago how 

 the young adventitious buds of this species start with 

 small " protocorms," quite like those of Phylloglossum 

 itself, or like the embryo of L. cernuum. And so we 

 may conclude that in Phylloglossum a tuberous 

 development, containing a store to start the plant in 

 the spring, has been added to what is already seen 

 normally each year in L. inundatutn. And this mode 

 of life of Phylloglossum begins, as Thomas has shown, 

 with its embryo. This appears to me to be a rational 

 explanation of the '* protocorm " of Phylloglossum ; 

 but it robs the plant 01 much of its theoretical interest 

 as an archaic form. 



The phylum of the Sphenophyllales was originally 

 based on certain slender straggling plants of the genus 

 Sphenophjilum found in the Palaeozoic Rocks ; but 

 they apparently died out in the Permian period. Your 

 native genera, Tmesipteris and Psilatum, were ranked 

 by earlier botanists with the Lycopods, but a better 

 acquaintance with their details, and especially the 

 examination of numerous specimens on the spot, indi- 

 cated a nearer affinity for them with the Spheno- 

 phyllales. It was Prof. Thomas who in 1902 first 

 suggested that the Psilotaceae might be included with 

 the Sphenophyllae in the phylum of the Spheno- 

 phyllales, and 1 personally agree with him. Dr. Scott, 

 however, dissents, on the ground that the leaves are 

 persistently whorled in the sphenophylls, while they 

 are alternate in the Psilotaceae ; and while the former 

 branch monopodially the latter dichotomise. But since 

 both these characters are seen to be variable within 

 the not far distant genus Lycopodium, the differences 

 do not seem to me to be a sufficient ground for keep- 

 ing them apart as the separate phyla of Spheno- 

 phyllales and Psilotales. Whatever degree of actual 

 relation we trace, such plants as Tmesipteris and Psi- 

 lotum are certainly the nearest living representatives 

 of the Sphenophylleae, a fact which gives them a 

 special distinction. The Psilotaceae also stand alone 

 in the fact that they are the only family of the Pterido- 

 phytes in which the gametophyte is still unknown. 

 They produce spores freely, but there the storj- stops. 

 Anv voung Australian who hits upon the way to 

 induce these recalcitrant spores to germinate, and to 

 produce prothalli and embryos, or who found their 

 prothalli and embr\os in the open, would have before 

 him a piece of work as sensational as anything that 

 could be suggested. Further, I am told that Tme- 

 sipteris grows here on the matted stumps of Todea 

 Barbara. I shall be alluding shortly to the fossil 

 Osmundaceae. May we not venture to fancy the pos- 

 sibility of some fossil Osmunda being found which 

 has embalmed for us among its roots a Mesozoic or 

 even a Tertiary Sphenophyll? And thus a link might 

 be found between the Palaeozoic tj-pes and the modern 

 Psilotaceae, not only in time, but even in character. 



We pass now to the last phylum of the Pteridophyta, 

 the Filicales. I am bound to say that for me its 

 interest far outweighs that of the others, and for this 

 reason : that it is represented by far the largest 

 number of genera and species at the present day, 

 while there is a sufficiently continuous and rich suc- 

 cession of fossil forms to serve as an efficient check 

 upon our comparative conclusions. 



Since 1890 it has been generally accepted that the 

 Eusporangiate ferns (those with more bulky sporangia) 

 were phyletically the more primitive t\-pes, and the 

 Leptosporangiate (those with more delicate sporangia) 

 the derivative, and in point of time later. The fossil 

 evidence clearly upholds this conclusion. But, further, 

 it has been shown that the character of the spor- 

 angium is merely an indicator of the general constitu- 

 tion of the plants in question. Where it is large and 

 complex, as in the Eusporangiates, all the apical seg- 



