5c6 



NA TURE 



[September 20. 1900 



"with smooth ganoine. But now we have the ventrals as well as 

 the pectorals acutely lobate in shape, and presumably archi- 

 pterygial in structure ; the top of the head is covered with 

 many small plates, there is no longer a dentigerous maxilla, the 

 skull is autostylic, and the palatopterygoids and the mandibular 

 splenial are like those of Ceratodus and bear each a tooth-plate 

 with radiating ridges. 



Now, comparing Dipterus with the recent Ceratodus and 

 Protoptertis, the first conclusion we are likely to draw is, that 

 the older Dipnoan is a very specialised form, that its heterocercal 

 tail and separate dorsals and anal are due to specialisation from 

 'the continuous diphycercal dorso-ano-caudal arrangement in the 

 recent for-nis, that the Holoptychiidoe were developed from it by 

 shortening up of the ventral archipterygium, as well as by the 

 ■changes in cranial structure, and that the Rhizodpntidse and 

 •Osteolepid^ are a still more specialised series in which the pec- 

 toral archipterygium has also shared the fate of the ventral in 

 /becoming shortened up and uniserial. 



Five years ago, however, M. Dollo, of the Natural History 

 Museum at Brussels, the well-known describer of the fossil rep- 

 tiles of Bernissart, proposed a new view to the effect that the 

 'process of evolution had gone exactly in the opposite direction ; ^ 

 and after 4ong consideration of the subject I find it difficult to 

 -escape from the conclusion that this view is more in accordance 

 with the facts of the case, though, as we shall see, it also has its 

 •own difficulties. 



I have already indicated above that we are, on account of the 

 more specialised structure of the teeth, justified in considering 

 'the Holoptychians, with their acutely lobate pectorals, a newer 

 type than the Rhieodonts, even though they did not survive so 

 ■long in geological time. What, then, of the question of 

 autostyly? 



We do not know the suspensorium of Holoptychius, but that 

 of the Rhieodontidae was certainly hyostylic, as in the recent 

 Polypterus. Now as there can be no doubt that the autostylic 

 ■condition of skull is a specialisation on the hyostylic form, as 

 seen also in the Chimseroids and in the Amphibia, to suppose 

 •that the hyostylic Crossopterygii were evolved from the autostylic 

 iDipnoi is, to say the least, highly improbable ; in my own 

 ■opinion, as well as in that of M. Dollo, it will not stand. 

 And if we assume a genetic connection between the two 

 groups it is in accordance with all analogy to look on the 

 Dipnoi a« the children and not as the parents of the 

 •Crossopterygii. 



M. Dollo adopts the opinion of Messrs. Balfour and Parker 

 that the apparently primitive diphycercal form of tail of the 

 recent Dipnoi is secondary, and caused by the abortion of the 

 termination of the vertebral axis as in various "Teleostei," so 

 that no argument can be based on the supposition that it repre- 

 sents the original " protocercal " or preheterocercal stage. Very 

 likely that is so, but it is not of so much importance for the 

 present inquiry, as both in the Osteolepidze and Rhizodontidse 

 we find among otherwise closely allied genera some which are 

 heterocercal, others more or less diphycercal. Diplopteriis, for 

 example, differs from Thursius only by its diphycercal tail, and 

 in like manner among the Rhizodontidse Tristichopterus is 

 heterocercal, EustJienopteron is nearly diphycercal, and there 

 •can be no doubt that, in spite of this, their caudal fins are 

 perfectly homologous structures. 



But of special interest is the question of the primitive or non- 

 primitive nature of the continuity of the median fins in the recent 

 Dipnoi. Like others I was inclined to believe it primitive, and 

 that the broken-up condition of these fins in Dipterus was a 

 subsequent specialisation, and in fact gave the series Phanero' 

 J>leuroit, Scaumenacia, Dipterus macropterus and D. Valettcien- 

 nesii as illustrating this process of differentiation. This view of 

 course draws on the imperfection of the geological record in 

 assuming the existence of ancient pre-Dipterian Dipnoi with 

 continuous median fins, which have never yet been discovered. 

 But Dollo, using the very same series of forms, showed good 

 reason for reading it in exactly the opposite direction. 

 The series is as follows : — 



(i) Dipterus Valenciennesii, Sedgw. and Murch., from the 

 Orcadian Old Red, and the oldest Dipnoan with whose shape 

 we are acquainted, has two dorsal fins with short bases, a 

 heterocercal caudal, and one short-based anal. 



(2) Dipterus macropterus, Traq., from a somewhat higher 

 •horizon in the Orcadian series, has the base of the second dorsal 

 much extended, the other fins remaining as before. 



1 "Sur laPhylogeni« des Dipneustes," Bulletin Soc. beige geol. paleont. 

 ^hydr., vol. ix. 1895. 



NO. 1612. VOL. 62] 



(3) In Scaumenacia curta (Whiteaves), from the Upper 

 Devonian of Canada, the first dorsal has advanced considerably 

 towards the head, and it^ base has now become elongated, while 

 the second has become still larger and more extended, though 

 still distinct from the caudal posteriorly. 



(4) In Phaneropleuron Andersoni, Huxley, from the Upper 

 Old Red of Fifeshire, the two dorsal fins are now fused with 

 each other and with the caudal, forming a long continuous fin 

 along the dorsal margin, while the tail has become nearly 

 diphycercal, with elongation of the base of the lower division of 

 the fin. But the anal still remains separate, narrow, and short- 

 based. 



(5) In the Carboniferous Uronemus lobatus, Ag., the anal is 

 now also absorbed in the lower division of the caudal, forming 

 now, likewise on the haemal aspect, a continuous median fin be- 

 hind the ventrals. There is also a last and feeble remnant of a 

 tendency to an upward direction of the extremity of the vertebral 

 axis. 



(6) In the recent Ceratodus Forsteri, Krefft, the tail is diphy- 

 cercal (secondary diphycercy), the median fins are continuous, 

 the pectorals and ventrals retain the biserial archipterygium, but 

 the cranial roof-bones have become few. 



(7) In Protopte7-us antiectens, Owen, the body is more eel-like, 

 and the paired fins have lost the lanceolate leaf-like appearance 

 which they show in Ceratodus and the older Dipnoi They are 

 like slender filaments in shape, with a fringe on one side of 

 minute dermal rays ; internally they retain the central jointed 

 axis of the " archipterygium," but according to Wiedersheim the 

 radials are gone, except it may be one pair at the very base of 

 the filament. 



(8) Finally, in Lepidosiren paradoxa, Fitz., the paired fins are 

 still more reduced, having become very small and short, with 

 only the axis remaining. 



From this point of view, then, Dipterus, instead of being the 

 most specialised Dipnoan, is the most archaic, and the modern 

 Ceratodus, Protoplerus and Lepidosiren are degenerate forms, and 

 instead of the Crossopterygii being the offspring of Dipterus-\We 

 forms, it is exactly the other way, the Dipnoi owing their origin to 

 Holoptychiidce, which again are a specialisation on the Rhizo- 

 dontidse, though they did not survive so long as these in 

 geological time. Consequently the Ceratodus limb, with its long 

 median segmented axis and biserial arrangement of radials, is not 

 an archipterygium in the literal sense of the word, but a deri- 

 vative form traceable to the short uniserial type in the Rhizo- 

 donts. But from what form of fin that was derived is a question 

 to which palaeontology gives us no answer, for the progenitors 

 of the Crossopterygii are as yet unknown to us. 



Plausible and attractive as this theory undoubtedly is, and 

 though it relieves the palaeontologist from many difficulties 

 which force themselves upon his mind if he tries to abide by the 

 belief that the Dipnoan form of limb had a selachian origin, and 

 was in turn handed on by them to the Crossopterygii, yet it is 

 not without its own stumbling-blocks. 



First, as to the dentition, on which, however, M. Dollo does 

 not seem to put much stress, it is impossible to derive Dipterus 

 directly from the Holoptychiidse, unless it suddenly acquired, as 

 so many of us have to do as we grow older, a new set of teeth. 

 The dendrodont dentition of Holoptychius could not in any way 

 be transformed into the ctenodont or ceratodont one of Dipterus : 

 both are highly specialised conditions, but in different directions. 

 Semon has recently shown that the tooth- plates of the recent 

 Ceratodus arise from the concrescence of numerous small simple 

 conical teeth, at first separate from each other. ^ Now this stage 

 in the embryo of the recent form represents to some extent the 

 condition in the Uronemidfe of the Carboniferous and Lower 

 Permian, which stand quite in the middle of DoUo's series. 



Again, the idea of the origin of the Dipnoi from the Crosso- 

 pterygii in the manner sketched above cuts off every thought of a 

 genetic connection between the biserial archipterygium in them 

 and in the Pleuracanthidae, so that we should have to believe , 

 that this very peculiar type of limb arose independently in the 

 Selachii as a parallel development. It may be asked. Why not ? 

 We may feel perfectly assured that the autostylic condition of the 

 skull in the Holocephali arose independently of that in the 

 Dipnoi, as did likewise a certain amount of resemblance in their 

 dentition. But those who from embryological grounds oppose 

 any notion of the origin of the Dipnoi from "Ganoids" might 

 here say, if they chose. If so, why should not also the same form 

 of limb have been independently evolved in Crossopterygii ? 



Accordingly, while philosophic palaeontology is much indebted 

 1 " Die Z&hneniwickelung des Ceratadus Forsteri." (Jena, 1899.) 



