Revieivs — Dr. Traquair on Silurian Fishes. 71 



As might be anticipated, Dr. Traquair also discusses the propriety 

 of admitting the existence of the sub-class Ostracodermi, to which 

 the Heterostraci and Osteostraci have been referred by Cope and the 

 British Museum Catalogue. He concludes that no more satisfactory 

 arrangement can be suggested at present. He accordingly accepts 

 it, and adds the new order Anaspida to include the Birkeniidse. 



The most disappointing feature of the memoir is, that it provides 

 no new facts which will be universally regarded as advancing our 

 knowledge of the affinities of the Ostracodermi. Dr. Traquair is 

 very emphatic in his refusal to believe that they are " Agnatha." 

 Very justifiably, he declares that he has no faith in negative 

 evidence. He must have some positive reason before he can admit 

 that these organisms were destitute of a mandibular arch and paired 

 limbs. On the other hand, he thinks his new discoveries justify 

 a belief that the Ostracodermi are the modified descendants of an 

 early group of Elasmobranchii. He " prefers to consider " the 

 Ccelolepidaj, or lowest Ostracoderms, "as having definitely split 

 oif from the [Elasmobranchii], from which they doubtless originally 

 came" (p. 844). 



When, however, the arguments for this new interpretation are 

 closely examined, I maintain that they are found to depend just 

 as much on negative evidence as the hypothesis of Cope. So far as 

 I understand them, they are based on two propositions — (i) that 

 the shagreen of the Co3lolepida3 is identical with that of an 

 Elasmobranch, and (ii) that the lappet-like postero-lateral angles 

 of the anterior broad part of the body in the Coelolepidse, 

 Drepanaspidse, and Cephalaspidse are pectoral fins. 



As to the first proposition, I think there is justification for 

 asserting that, in the present state of knowledge, it affords no sure 

 proof of affinity between the Ostracoderms and the Elasmobranchs. 

 We cannot conceive of the dei'mis of a chordate animal beginning to 

 acquire superficial calcified hard parts by any other process than by 

 the deposition of mineral matter round papillae. At the remote period 

 when the Chordata were first beginning to develop an exoskeleton, 

 it is extremely probable that these animals were already well 

 differentiated ; at least, we cannot explain their diversity and high 

 grade in the early Devonian period unless we admit this probability. 

 When dealing with Silurian fish-like organisms, therefore, a form 

 armoured with tubercles of dentine does not necessarily belong to 

 the Elasmobranchii; it may be a Chordate animal of any kind just 

 beginning to assume a protective covering. The admitted fact that 

 since early Carboniferous times an exoskeleton exclusively of dentinal 

 tissues has persisted only in Elasmobranchii and Chimseroidei, is no 

 proof that any organism similarly armoured in the strange archaic 

 Devonian and Silurian faunas is intimately related to either of 

 these sub-classes. 



In short, if Dr. Traquair has discovered that the Coelolepidaa are 

 closely related to the Heterostraci proper and to the Osteostraci, as 

 seems almost certain, I maintain that he has not helped us towards 

 forming any more plausible theory of their origin. He has merely 



