September 21, 191 1] 



NATURE 



597 



Seem, perhaps, somewhat hazardous, considering the differ - 

 sting between the Psilotaceaj and Sphenophyllum ; 

 and the more cautious attitude of Seward, in setting up 

 a separate group for these forms, seems, on the whole, 

 more satisfactory than forcing these aberrant relatives of 

 ih Lycopods into the somewhat Procrustean bed of 

 Sphenophyllales, which necessitates the minimising of such 

 important differences as the dichotomous branching of the 

 axis and the alternate arrangement of their leaves, though 

 the latter character allows, it is true, of some bridging over. 

 But, even adopting this more cautious attitude, the study 

 of the Sphenophyllales has been of great help in coming 

 to a clearer understanding of certain morphological 

 peculiarities of the Psilotacea;, quite apart from the Hood 

 of light which this synthetic group of Sphenophyllales has 

 thrown upon the relationship of the Lycopodiales to the 

 Equisetales. 



More far-reaching in its bearing on the relationships of 

 existing plants has been the study of those interesting 

 fern-like plants which seem to show in their vegetative 

 organs a structure possessing both fern-like and Cycadian 

 affinities. Full of interest as these so-called Cycadofilices 

 were in their vegetative organisation, they were destined 

 to rivet on themselves the attention of all botanists by the 

 ili-mvfTv of their fructifications. No chapter in the recent 

 history of Palaeobotany is more thrilling than the discovery, 

 by the patient and thorough researches of Prof. Oliver, of 

 the connection between I.yginodendron and the well-known 

 palaeozoic seed, Lagenostoma. With Dr. Scott as sponsor, 

 this new and startling revelation met with ready accept- 

 ance, and, thanks to the indefatigable energies of Palaeo- 

 botanists, no fossil fern seemed at one time safe from 

 possible inclusion among the Pteridospermae. 



The infectious enthusiasm with which the discovery of 

 the seed-bearing habit of the Lyginodendrea? and the 

 Medullosa; was greeted carried all before it, and we in 

 England, particularly, have perhaps not looked carefully 

 enough into the foundations upon which rested the theory 

 that these groups form the " missing links " between the 

 Ferns and Cycads. A criticism against the wholesale 

 acceptance of this view has been put forward by Prof. 

 Chodat, 1 of Geneva, that distinguished and versatile 

 botanist whom we have on several occasions had the 

 pleasure of welcoming among us. Couched throughout 

 in friendly and courteous language, and full of admiration 

 for the work of those who were concerned in the establish- 

 ment of the group of Cycadofilices, now termed Pterido- 

 spermse, Prof. Chodat suggests that English Pakeobotanists 

 have not sufficiently appreciated the work of Bertrand and 

 Corneille ' on the fibro-vascular system of existing ferns, 

 and have not revised, in the light of the researches of 

 these French investigators, the interpretation given to the 

 arrangement of the primary vascular tissues of Lygino- 

 dendron. In Chodat's opinion the structure of the primary 

 groups of wood found in the stem and in the double leaf- 

 trace of this plant is not directly comparable with the 

 arrangement found in the petiole of existing Cycads. In 

 the latter the bulk of the metaxylem is centripetal, while 

 we have, in addition, a varying amount of small-celled 

 centrifugal wood towards the outside of the protoxylem, 

 and, though separated from it by a group of parenchy- 

 matous cells, the bundle may be conveniently described as 

 mesarch. In Lyginodendron, and the same applies to 

 Heterangium, the primary bundles of the stem appear at 

 first sight to be mesarch too, but in Chodat's opinion, if 

 I understand him correctly, the metaxylem is exclusively 

 centrifugal in its development, but, widening out and 

 bending inwards again, in form of the Greek letter w, 

 the two extremities of the metaxylem are united on the 

 inside of the protoxylem, forming an arrangement 

 described by Bertrand and Corneille in the case of several 

 fern petioles under the name of " un divergeant ferme." 



Several details of structure, such as the type of pitting 

 of the metaxylem elements and the separation of the 

 protoxylem from the adaxial elements of metaxylem by 

 parenchymatous cells, confirm Chodat in his view that the 



1 Chodat, R. : " Les Pte'ropsides des temps paleozoiques," .1 



physiques ti naturetles, Geneve, tome \-wi, 190?.. 

 - Bertrand, (\ E., and Corneille, F. : " Elude sur quelques characteris- 

 tiqu.es de la structure des filicinees actuelles," Trnaaux 

 lie. tool. 



NO. 2l86, VOL. 87] 



primary bundles of Lyginodendron are not really mesarch, 

 and that the stem of Lyginodendron is essentially 

 Filicinean in nature. Chodat cites other characters, such 

 as the presence of sclerised elements in the pith, and the 

 absence of mui ilage duels, in support of his view of the 

 purely filicinean affinities of the Lyginodendrea?. The 

 presence of secondary thickening in Lyginodendron, he 

 regards not as indicative of Cycadian affinity, but merely 

 as another instance of secondary growth in an extinct 

 Cryptogam, taking up very much the position of William- 

 son in his earlier controversy with French botanists with 

 regard to the secondary thickening of Calamites and 

 Lepidodendreae. Chodat is also at variance with Kidston 

 and Miss Benson as to the nature of the microspores borne 

 on the fronds of Lyginodendron or Lyginopteris, as he 

 prefers to call this plant. He certainly figures some very 

 fern-like sporangia, attached to the fronds of Lyginoden- 

 dron, but anyone who has worked with the very frag- 

 mentary and somewhat disorganised material contained in 

 our nodules knows how difficult it is to be absolutely certain 

 of structural continuity. Nevertheless a re-investigation of 

 the whole question of the microsporangia of Lyginodendron 

 seems to me clearly called for by the publication of Chodat's 

 figures. 



As regards the seed-bearing habit of Lyginodendron. 

 Chodat adopts wholeheartedly Oliver's correlation of 

 Lagenostoma with the fronds of Lyginodendron, but would 

 regard the seed, apparently devoid of endosperm at the 

 time of pollination, as a somewhat specialised macrosporic 

 development, of more complex structure, but analogous in 

 its nature to the seed-like organ exhibited by Lepidocarpon 

 in another phylum of the Pteridophyta. "In any cast 

 he concludes, " the origin and the biology of this kind of 

 seed must have been very different from those of the seeds 

 of the Gymnosperms." 



This contention, based mainly on the tardy development 

 of the endosperm in Lagenostoma, is the least weighty part 

 of Chodat's criticism, for it has never been asserted that 

 the seeds were identical with those of existing Cycads. We 

 know that the seed-habit was adopted by various groups of 

 Vascular Cryptogams, and it is revealed in fossil plants in 

 various stages of evolution, so that it may be readily pre- 

 sented to us at a special stage of its evolution in Lygino- 

 dendron. Moreover, we must remember that in so highly 

 organised a Gvmnosperm as Pinus, the macrospore itself is 

 not fully developed at the time of pollination. Though not 

 suggesting this as a primitive feature in the case of the 

 pine, we ran well imagine how, by a gradual process of 

 " anticipation," the prothallus might become established 

 before pollination in any group of primitive seed-bearing 

 plants. There are other more specialised rather than 

 primitive features in the complex structure of Lagenostoma 

 which might with much more reason be invoked, to show 

 that the seed of Lyginodendron does not form a step in the 

 series of forms leading to the Cycadian ovule. 



But leaving this point out of consideration, Chodat brings 

 forward some strong reasons for his conclusions that the 

 Lyginodendrese were plants possessing stems of purely 

 fern-like structure, increasing in thickness by means of a 

 cambium, that their foliage was of filicinean structure, but 

 provided with two kinds of sporangia, microsporangia 

 similar to those of Leptosporangiate ferns, and macro- 

 sporangia of specialised type, containing a single macro- 

 spore; This group, therefore, Chodat regards as a highly 

 specialised group of ferns, which, he considers, shows no 

 particular connection with the Cycads, and may have 

 formed the end in a series of highly differentiated members 

 of the Filicinea;. 



Of the Medullosa;, on the other hand, Chodat takes a 

 verv different view. Both in the structure of their primary 

 and secondary growth, as well as in their polystely, he sees 

 close affinity' of these forms to the Cycads, borne out by 

 smaller secondary features, such as the presence of mucilag- 

 ducts and the simple form of pollen-chamber. Chodat 

 considers the agreement of the Medullosa; with th( 

 Cycadacea? to be so close that he regards them as 

 Protocycadea;, the fern-like habit being restricted to the 

 position of the sporangia on the vegetative fronds. 

 Medullosa, therefore, would be only one link in the chain 

 connecting the Cycads with the Filicales, and a link very 



