l6 JOHAN RIÆR. M.-N. Kl. 



to fonn Ihis <;cnus into a new laniily, Arcja^jnildc, half-way between 

 Aieleasjnddd and ('ephalabpidacK The new species will be named J. 

 rohuslKs on account of the robust structure of the body. 



The smaller and rarer form is most nearly related to the typical 

 ('('ltlHil.(i,s])i(/(i(\ bui has retained some [)rimitivc characters, that the tj'pi- 

 cal forms of tiie genus (Jcp/inlnsjns have lost. This must also be given 

 as a new genus. The dorsal shield in this new form is very short and 

 the body slender, and I therefore call it Mirjanph gracilis. For the 

 present it seems most natural to place this new genus into the family 

 Cephalaspidae. 



Both A<:('r(i><j)i.s and Micra.spin show two distinct dorsal fins. This 

 is surprising, as the previously known forms of (,'eplialas]i/i/(jniorp]ii are 

 said to have one only. I believe however, that this difference is only 

 apparent. The dorsal fin previously stated in the Ceplialaspidae clearly 

 corresponds with the hinder dorsal fin in the new forms. I now pre- 

 sume, that the foremost dorsal fin has been drawn into the dorsal shield 

 and transformed into the characteristic dorsal spine in such forms as 

 C. Lyelli Ag. with a strongly developed dorsal shield. In other forms 

 without a dorsal spine e. g. C Murchisoni Eg., this process of differen- 

 tiation has not yet taken place, and therefore in the event of one finding 

 better preserved specimens than have hitherto been forthcoming, we shall 

 probably be able to assign two dorsal fins to the latter forms also. 



These two new forms are of great interest, both from a palaeon- 

 tological and a phylogenetical point of view. They prove that Traqu- 

 air's theory as to the derivation of the Cephalaspidomorplii from the 

 highly primitive Upper Silurian Coelolepidae is the correct one. This 

 series of evolutions, especially from Ateleaspns to CepJtalasjns, will be 

 one of the most complete amongst palaeozoic fishes. 



Both forms have been found in my collection in extremely well 

 preserved specimens, and it is to be hoped that it will be possible to 

 reconstruct them completely, even as to the minutest details. 



Aceraspis robustus no v. gen. et sp. 



A medium sized, but strongly built form having a total length of 

 16.5 — 18 cm. from the snout to the end of the tail; the head shield 

 takes up not quite one third of the length. 



This form represents an intermediate link between Atelaspidae and 

 Cephalaspidae. The head-shield is complete, but short, without a post- 

 erior median spine. Its posterior corners are truncated without cornua and 



^ Goodrich, A Treatise in Zoology 1909. Fishes, Page 203. 



