84S DR RAMSAY H. TRAQUAIR ON FOSSIL FISHES COLLECTED BY THE 



plates, which, small as they are, remind, us strongly of the polygonal areas so often 

 seen on the surface of the plates of Psammosteus. 



It need scarcely be added that the large median dorsal and ventral plates of Dre- 

 panaspis and the oblong boat-like shields of Psammosteus mutually remind us of each 

 other, and though the former never show any polygonal areas, the smaller plates with 

 which they are surrounded are eminently suggestive of these areas as they occur in 

 Psammosteus paradoxus, Agassiz, and Ps. Taylori, Traq. 



There are, however, two other forms of plates often found associated with the large 

 oblong ones above referred to, the structure and external sculpture of which lead us to 

 refer them also to Psammosteus, Of these we have first .certain flattened and somewhat 

 falciform pieces, pointed and often worn at one extremity, and having towards that 

 extremity the characteristic Psammosteus-orn&ment, which, however, covers more of 

 the surface on one side than on the other. A fragment of one of these bodies was figured 

 by Pander (xxii. PL VII. fig. 22) as an " Ichthyodorulite." From their want of 

 bilateral symmetry, they must have occupied a lateral position, and they were indeed, as 

 already remarked, figured and described by Trautschold as the paddles of a species of 

 Coccosteus. Eelying on the microscopic structure of these plates or " ichthyodorulites," 

 which, as described by Pander, consists of true dentine without any bone lacunas, I 

 referred them in 1890 (xxxiii. p. 134) to the category of " Selachian Appendages," noting 

 also the certain amount of resemblance which they bear to the peculiar carboniferous 

 species known as Oracanthus. A year afterwards they were noticed, by Smith Wood- 

 ward in the second volume of his Catalogue (xxxviii. p. 126), and referred by him to 

 Agassiz's Psammosteus meandrinus ; and again, in 1895, in his Problem of the 

 Primceval Sharks, he compares those bodies to Oracanthus, adding, as regards their 

 position on the fish, that "they may have been arranged along the lower margin of the 

 body, as in certain Acanthodian sharks (e.g., Climatius), or the animal may have had 

 only a single pair of these spines at the back of the bead, as described by Dr Traquair 

 in Oracanthus.'" The second supposition is more feasible than the first, and I rather 

 think that the "ichthyodorulites" in question were backwardly directed developni' 

 of plates corresponding to the postero-laterals of Drepanaspis. 



The third form of Psammosteus-\>\'&te, of which Pander (xxii. PI. VII. fig. 16) 

 figured an example as possibly a caudal scale or spine of Asterolepis, is comparatively 

 small in size, bilaterally symmetrical, oblong, bluntly pointed at one extremity, 

 and sculptured externally with the usual shagreen-like tuberculation. 



Mr Smith Woodward remarks (xxxvii. p. 39), concerning these bodies, that they 

 are shaped much like the rostrum of Pteraspis, but in my mind there is not the slightest 

 doubt that they are ridge-scales of the tail, similar to those which are to be found in situ 

 in Drepanaspis. 



It is now pretty clear that Psammosteus is closely allied to Drepanaspis, — so closely 

 that it may be a question as to whether there is any need for family distinction, 

 think, however, that it is better for the present to keep them in separate families until 



