February 7, 1908] 



SCIENCE. 



203 



supreme "effort" to compete successfully 

 with their fundamentally better equipped 

 rivals. 



The recent discoveries of Professor Jae- 

 kel have confirmed a prediction— sugges- 

 tion made nearly a score of years ago— 

 that we should yet find how manifold 

 were the forms of arthrodires during 

 their pei-iod of maximum prosperity. He 

 has found in Wildungen, in practically a 

 single spot and in a narrow fossiliferous 

 seam, usually but a few inches thick, no 

 less than twelve genera and fifty species of 

 arthrodires, nearly all of which are new ! 

 And paleontologists may well look forward 

 to the publication of his extraordinary re- 

 sults. Here, then, in the uppermost De- 

 vonian, close before their extinction, these 

 forms assumed the most varied characters, 

 even bilateral nan-owness, in which the 

 form of a swift-swimming teleost was 

 paralleled. I hope, however, that my col- 

 leagues will leave this case to parallelism 

 and will not make this creature, the latest 

 and most specialized of the placoderms, the 

 progenitor of teleosts ! Dr. Jaekel has, up 

 to the present moment, unhappily, pub- 

 lished only an abstract of his results. But 

 they show clearly enough that the forms 

 described are typical arthrodires ; and they 

 yield, I believe, no good evidence as to the 

 kinship of these forms to true fishes.- Dr. 

 Eastman's recent and careful elaboration 

 of the view (of a score of years ago) that 



= Thus there is no new light on the presence of 

 fins and girdles : what he regards as the " un- 

 doubtedly demonstrable " hip girdle in Coccosteus 

 is to certain morphologists, at least, a very doubt- 

 ful structure; whatever it be, it is rudimentary 

 in Jaekel's new forms. Nor does his explanation 

 carry conviction as to the under jaw of PhoUdos- 

 teus with angular and articular elements. The 

 former is, I believe, the " interlateral " plate well 

 known in the ventral armoring of coccosteids, the 

 latter probably an articular (detached) process 

 of the central plate. The views of Jaekel as to 

 the position of the Arthrodira can not, however, 

 be criticized in detail in the present paper. 



arthrodires are of lung-fish derivation and 

 that a primitive form of the Australian 

 lung-fish (Ceratodus) was their progenitor 

 I have already commented upon in Sci- 

 ence. On the evidence especially of the 

 flattened dental plates of a late form of 

 arthrodire, Mylostoma, he is convinced of 

 a kinship to the modern forms, and he 

 explains the absence of ceratodonts in the 

 pre-Mesozoic on the ground of the frag- 

 mentary nature of paleontological evidence. 

 He neglects, however, it seems to me, to 

 take into account what we do know of the 

 Paleozoic lung-fishes, and these documents 

 are both numerous and important, as DoUo, 

 for example, has pointed out. And he has 

 not evaded the morphological pitfall (it 

 seems so to me at least) of attempting to 

 establish homologies between more or less 

 terminal forms of widely different descent. 

 Indeed it is clear that if arthrodires are 

 descended from a primitive eeratodont, 

 their puzzling allies, bothriolepids and 

 cephalaspids must also be closely related to 

 the same ancestor, but this is difScult even 

 to imagine : for who can fancy, as early at 

 least as in the upper silurian, in which all 

 these forms occur, that even then they 

 could be traced back to lung-fishes essen- 

 tially eeratodont? Furthermore, if I mis- 

 take not. Dr. Eastman believes that sharks 

 are the ancestors of lung-fishes, and in this 

 event, how far back into the Urzeit would 

 our paleontological fancy project to find 

 the origin of our modern fishes? The- 

 paleontological record is scored, seamed 

 and scanty, we painfully admit, but I am 

 confident that it is not as bad as all this : 

 if the lung-fishes, arthrodira and their 

 anomalous allies came from a eeratodont 

 ancestor we shoitld certainly have found a 

 trace of it somewhere in the stupendous- 

 ly long interval between the upper Silu- 

 rian and the Mesozoie. We find, on the 

 contrary, that not merely is this creature 

 absent, conspicuously absent, but that its 



