January 25, 1900] 



NA TURE 



309 



pidian affinities of which he insisted in 1896. The sole repre- 

 sentative of this family, Drepanaspis, from the Lower Devonian 

 of Gmiinden in West Germany, was only described by Schliiiter 

 in 1887. We append a figure of it, from which it will be seen that 

 while in respect to the possession of central (^. ), lateral {p. I) and 

 rostral (r.) plates it " points forwards " to the Pteraspidian type, 

 in the possession of tesserie it " points backwards " to the Coelo- 

 lepid ; and by appeal to the condition of the Psammosteidae, 

 which the author holds to be near related forms, the * ' stellate 

 tubercles " of which he has already come to regard as " shagreen- 

 L^ranules which have coalesced," he builds up an argument for 

 the origin of the plates of the higher Pteraspidians by fusion of 

 shagreen-granules in linear series. 



Having thus with much justification strengthened his con- 

 viction that the Pteraspidian fishes have had a common origin 

 with the primitive Elasmobranchs, and that the Ccelolepidre, 

 Psammosteidae and Diepanaspid^e arc in order representative 

 of the ascending series which lead up to the Pteraspidians 

 proper, the author returns to the consideration of the Cephal- 



FiG. 5. — Drepanaspis Gmundeiunsis, restored outline, dorsial aspect, 

 omitting surface ornament, x = orbit. 



aspidians, and, recalling the points of structural community 

 between the aforenamed Ateleaspis and the Coelolepidae, he 

 concludes that between the Pteraspid and Cephalaspid groups 

 there is, " after all, an actual connection." 



As might have been expected, he incidentally rejects Clay- 

 pole's alleged discovery of paired fins in Palaeaspis. 



It will be remarked that the central point oq which the con- 

 clusion as to the common origin of the primitive Elasmobranchs 

 and the Ostracodermi (using this term in the author's sense) 

 turns, is the homologising of the plates of the higher Pteras- 

 pidians with those of the less fully developed Drepanaspid and 

 Psammosteid types. So far as the central and rostra! elements 

 are concerned well and good ; but while admitting the assumed 

 homology of the coronal plate of Pteraspis with the postero-lateral 

 of Drepanaspis, we do not feel so sure about regarding these as 

 the results of encasement of the lateral fin-flaps of Thelodus and 

 Lanarkia, which are rendered thereby " utterly functionless as 

 fins by being enclosed in unyielding bony plates." In dealing with 



NO. 1578, VOL. 61] 



these facts, Dr. Traquair has been led into a digression, savour- 

 ing of leniency towards the lateral fold theory of the origin of the 

 vertebrate limbs, to which indeed he has already expressed him- 

 self favourable in his recent memoir on the Selachian Cladodus 

 Neilsoni. Following Powrie, he draws attention to the re- 

 semblance between the supposed fin-fold of Thelodus and that 

 of the living Rays. If in this he is right, the characteristic 

 feature of both is the forward extension into the head-region ; 

 and when it is remembered that ontogenetically the pectoral fin 

 bud arises post-branchially, this extension can only be the result 

 of forward rotation, and therefore an index of extreme 

 specialisation. If this be so, the author's implication that 

 " we have here a very interesting point in connection with the 

 much discussed question of the morphology of the paired limbs 

 in vertebrates — a new and important corroboration of the 

 lateral fold theory," can hardly be taken in the sense which 

 would seem intended. There is, however, an alternative read- 

 ing, to which he has himself pointed the way, which we are most 

 readily inclined to adopt, viz. that already alluded to (t/. antea) 

 as involving a comparison between the posterior pectoral 

 lobe of Thelodus (^c.' Fig. 1) and the cornual flap of Cephalaspis 

 which, with the author, we believe Lankester to have been 

 right in regarding as " the equivalent of a pectoral fin." To put 

 the case otherwise, we do not see the proof that the presumed for- 

 ward extension of the fin-fold {pc") is fin-like in structure, and we 

 incline to the cor^lusion, especially in view of the perforation of 

 its supposed investing plate by the so-called respiratory aperture 

 in Pteraspis ( 1 ) that it is this alone which has been converted 

 into thecornual plate of that genus and the postero-lateral plate 

 of Drepanaspis; (2) that this is in all probability represented by 

 the cornua of the Cephalaspid ise, which may therefore well be 

 accessory branchial organs probably enclosing a central passage 

 — "atria," in fact, if not actual opercula ! wherefore it would be 

 interesting to ascertain whether they enclose a central canal or 

 passage and are lined by shagreen-granules or tesserce, as might 

 well be if this interpretation is correct. And we further suggest 

 as the natural sequence to this (3) that the posterior lobe of the 

 supposed fin-fold of the Coelolepidae {/(•.' Fig. i) alone represents 

 the pectoral member of the true fishes, and is in turn represented 

 by the cornual flaps of the Cephalaspidians ; and (4) that the 

 pectoral member of the Pteraspidians has yet to be sought. 



We put forward this view with all reserve, and we submit that 

 while it is not opposed to the facts, it still further justifies the 

 belief in a connection between the Cephalaspidian and Pteras- 

 pidian forms which Dr. Traquair has revived. All known facts 

 of morphology justify the conclusion that the paired limbs of the 

 vertebrate have been wholly evolved within that phylum, and 

 there are not a few which suggest that the pectoral and pelvic 

 members have been acquired independently in anterio-posterior. 

 succession. If so, may not the Coelolepidae and Cephalaspidiae 

 be now regarded (probably with the Pteraspidiae as very distant 

 allies) as the representatives of an apodal stage in evolution, at 

 which the pelvic member had not yet come into existence. 

 Their extreme structural simplicity and entire lack of jaws, 

 teeth, and apparent endoskeleton capable of preservation in the 

 fossil state, are certainly not at variance with this view, and 

 under it the old belief in their extremely specialised nature and 

 their presumed degeneration, towards which even Traquair him- 

 self Inclines, with that in their afiinity with the Marsipobranchs, 

 largely disappears. 



The question is one for the palaeontologist ; and while con- 

 gratulating the officers of the Geological Survey and the Edin- 

 burgh Museum upon the addition to their already matchless 

 collection of these wonderful remains, and the author upon the 

 masterly manner in which, as a true morphologist, and with 

 " soul," he has worked them out, we look to him and his 

 friends in the field to furnish the next link in the chain. The 

 present one is a triumph for all concerned, worthy the author of 

 "The Palaeoniscidae and interpreter of Palaeospondylus, and 

 as marking progress it is equal to anything achieved in the 

 palaeontology of the last two decades. G. B. M. 



UNIVERSITY AND EDUCATIONAL 

 INTELLIGENCE, 



Oxford. — The following are among the lectures announced 

 for the present term : — Prof Clifton, acoustics ; Mr. Walker, 

 physical optics ; Mr. Hilton, elementary mechanics and physics ; 

 Mr. Baynes, elementary heit and light ; Mr. JervisSmith, 

 dynamo and motor machinery and electrical testing ; Prof. 



