104 



NATURE 



[September 24, 1914 



mentations are, as a rule, complex, and the construc- 

 tion of the whole plant relatively bulky. Where the 

 sporangium is delicate and relatively simple all the 

 apical segmentations follow suit, and the construction 

 of the plant is on a less bulky model. On this basis 

 we may range ihe ferns roughly as a sequence, start- 

 ing from relatively bulky types of the distant past, 

 and progressing to the more delicate types of the 

 present day. The large majority of the living species 

 belong naturally to the latter. But the former are 

 still represented by a few genera and species which, 

 like other survivals from a distant past, are frequently 

 of very restricted distribution. 



An interesting feature of the Australasian flora is 

 that a considerable number of these relatively ancient 

 forms are included in it. Thus the Marattiaceae are 

 represented by one species of Marattia and one of 

 Angiopteris, Though in themselves interesting, they 

 will be passed over without special remark, as they 

 are very widely spread tropical forms. 



All the three genera of Ophioglossaceae are included, 

 there being two species of Ophioglossum and two of 

 Botrychium, while Heiminthostachys is recorded from 

 Rockingham Bay. This family is coming more than 

 ever to the front in our comparisons, owing to their 

 similarity in various asp;;. Is to the ancient Botry- 

 opterideaj. Though the Ophioglossaceae have no 

 secure or consecutive fossil history, still they may now 

 be accepted as being very primitive but curiouslv 

 specialised ferns. Perhaps the most interesting point 

 recently detected in them is the suspensor found by 

 Dr. Lyon in Botrychium ohliquum, and by Dr. Lang 

 in Helrninthostachys. This provides a point for their 

 comparison with the similar embryonic condition in 

 Danoea, as demonstrated by Prof'. Campbell. The 

 existence of a filamentous initial stage of the embryo 

 is thus shown^ for three of the most primitive of living 

 ferns. Its existence in all of the Bryophytes, and in 

 most of the Lycopods, as well as in the seed-plants, 

 is a very significant fact. Dr. Lang suggests that 

 "the suspensor represents the last trace of the fila- 

 mentous juvenile stage in the development of the plant, 

 and may have persisted in the seed-plants from their 

 filicineous ancestrv." Such a possibilitv wqvild fit 

 singularly well with the theory of encapsulation of 

 the sporophyte in the venter of the archegonium. 



The representation of the ancient family of the 

 Osmundaceae in the Australasian flora is very fine, 

 though limited to five living species, while Osmunda 

 itself is absent. It is, however, interesting that the 

 family dates back locally to early fossil times. It was 

 upon two specimens of 'Osmundites from the Jurassic 

 Rocks in the Otago district of New Zealand that the 

 series of remarkable papers on the fossil Osmundaceae 

 by Kidston and Gwynne-Vaughan was initiated. It is 

 no exaggeration to say that these papers have done 

 more than any other recent researches to promote a 

 true understanding, not only of the Osmundaceae- 

 themselves, but of fern-anatomv as a whole. They 

 have placed the stellar theory in ferns for the first 

 time upon a basis of comparison, checked bv reference 

 to stratigraphical sequence. It would be leading us 

 too far for me to attempt here to summarise the 

 important results which have sprung from the study 

 of those fossils, so generouslv placed bv Mr. Dunlop 

 in the hands of those exceptionally able to turn them 

 to account. It must suffice to sav that it is now 

 possible to trace as a fairly continuous story the steps 

 leading from the protostelic state to the complex 

 condition of the modern Osmunda. These facts and 

 conclusions are to be put in relation with the 

 anatomical data fast accumulating from the Ophio- 

 glossaceae in the hands of Prof. Lang and others. 

 From such comparisons a rational explanation of the 

 evolutionary steps leading (o the complex stelar state 



NO. 2343, VOL. 94] 



in ferns at large begins to emerge. This is no mere 

 tissue of surmises, for the conclusions are based on 

 detailed comparison of types occurring in lower hori- 

 zons with those of the present day. 



1 must pass over with merely nominal mention your 

 interesting representation of the ancient families of 

 Schizaeaceae, Gleicheniaceae, and Hymenophyllaceae, all 

 of which touch the very foundations of any phyletic 

 system of ferns. Also the magnificent array of Dick- 

 sonieae and Cyatheae, and of the important genus 

 Lindsaya — ferns which take a rather higher position 

 in point of view of descent. But I am bound to devote 

 a few moments to one of your most remarkable ferns, 

 endemic in New Zealand — the monotypic Loxsoma. 



This species has peculiar characters which justify 

 its being regarded systematically as the sole repre- 

 sentative of a distinct tribe. It is also restricted 

 geographically to the North Island of New Zealand. 

 These facts at once suggest that it is an ancient sur- 

 vival, a conclusion with which its solenostelic axis, 

 its sorus and sporangium, and its prothallus readily 

 accord. I have lately shown that the Leptosporangiate 

 ferns fall into two distinct series, those in which the 

 origin of the sorus is constantly superficial, and those 

 in which it is a^ constantly marginal. Loxsoma is 

 one of the " Marginales." It shares this position with 

 the Schizaeaceae, Thyrsopterideae, Hymenophyllaceae, 

 and Dicksonieae, and the derivatives Davallieae and 

 Oleandreae. Its nearest living relative is probably 

 Thyrsopteris, which is again a monotypic species 

 endemic in the island of Juan Fernandez. There is 

 also a probable relation to the genus Loxsomopsis, 

 represented by one species from Costa Rica, and a 

 second lately discovered in Bolivia. Such a wide and 

 isolated distribution of types, which by their char- 

 acters are certainly archaic, suggests that we see in 

 them the relics of a filicineous state once widely spread, 

 which probably sprang from a Schizaeaceous source, 

 and with them represent the forerunners of the whole 

 marginal series. If we look for further enlightenment 

 from the fossils, it is to the Secondary rocks that we 

 should turn. It is then specially interesting that Mr. 

 Hamshaw Thomas has lately described a new Jurassic 

 fern, Stachypteris halli, which has marginal sori, and 

 is probably referable to a position like that of Lox- 

 soma and Thyrsopteris, between the Schizaeaceae and 

 the Dicksonieae. In fact, the gaps in the evolutionary 

 series of the Marginales are filling up. We may 

 await with confidence fresh evidence from the Jurassic 

 period, upon which Prof. Seward is directing an inten- 

 sive interest. 



I should be ungrateful indeed if I did not mention 

 your very full representation of Blechnoid ferns : for 

 developmental material of several of these has been 

 sent to me by Dr. Cockayne, and others from New- 

 Zealand. A wide comparative study of the genus has 

 led me to somewhat unexpected results in regard to 

 the plasticity of the sorus, its phyletic fusions and 

 disruptions. The consequent derivative forms are seen 

 in Woodwardia and Doodya on one hand, and on 

 the other in Scolopendrium and Asplenum. These 

 ferns together constitute a coherent phvlum springing 

 ultimately from a Cyatheoid source. The details upon 

 which this conclusion is based I hope to describe in a 

 separate communication to the section. 



And, lastly, the Hydropterideae deserve brief 

 mention. Represented in your flora by two species 

 of Azolla, and one each of Marsilea and Pilularia, they 

 typify a condition which must theoretically have 

 existed among ferns in very early times, viz., the 

 heterosporous state. But hitherto, notwithstanding the 

 existence of our living Hydropterideae, no fossil fern 

 with microscopic structure preserved had been detected 

 J from the primary rocks, showing this intermediate 

 / condition between the homosporous type and that of 



