370 Dr. E-. H. Traquair on Palceozoic Fishes. 



aspt's angh'ca in the Museum of Practical Geology, Jermyn 

 Street. Here it is clearly shown that the postero-external 

 extension of the shield in this species is due to the presence 

 of a small plate (woodcut, ag.), which seems to be absent in 

 all the specimens of P. acadica which I have hitherto seen. 

 It will be noticed in the restored sketch (p. 369) of the 

 arrangement of the cranial shield-elements in P. anglica that 

 the line of suture which divides this plate, ag., from the rest 

 of the shield repeats pretty exactly the re-entering angle seen 

 at the same place on the margin in the Canadian species. 

 The inference is therefore unavoidable that the plate in 

 question was also present in P. acadica, but was prone to be 

 lost through not having been anchylosed. This plate, to 

 which from its position we may give the name of " angular" 

 is apparently absent in the cranial shield of Coccosteus, the 

 arrangement of the elements of which otherwise agrees in 

 general plan with that in Phlyctcenaspis. 



Acanthaspis priimensis, sp. n. (PI. IX. fig. 1.) 



Among some fossils obtained a few years ago for the 

 Edinburgh Museum from Mr. B. Stiirtz, of Bonn, are two 

 specimens, or, rather, impressions of the anterior ventro- 

 lateral plate, with attached spine, of a species of Acanthaspis 

 from the Lower Devonian of Prihn, in the Eifel. Of these 

 the smaller and more perfect is here figured. 



As indicated by the impression, the plate, somewhat quad- 

 rate in form, was ornamented externally by a minute, often 

 vermicularly confluent tuberculation. From the anterior two 

 thirds of its outer margin a pointed process passes outwards 

 and backwards, to the antero-external border of which is fixed 

 a long, tapering, slightly curved hollow spine. The direction 

 of this spine in the complete animal must have been obliquely 

 backwards and outwards ; that it was hollow is shown by a 

 small portion of the stony core of the interior, which still 

 adheres at one point near its middle. The impression also 

 shows that the spine was ornamented externally by fine 

 longitudinal ribs, showing a minute sharp tuberculation of the 

 same character as that in the North- American Acanthaspis 

 armata of Newberry. The entire length of the spine is rather 

 more than twice that of its supporting plate ; its greatest 

 width is contained over eight times in its length. Length of 

 the plate (anterior ventro-lateral) 1 T \ inch, of the spine 2A 

 inches. The other specimen, which is not so good, is some- 

 what larger, the spine having a length of 3 inches. 



There can be no doubt that the above-described fossil, to 



