83G DR RAMSAY H. TRAQUAIR ON FOSSIL FISHES COLLECTED BY THE 



origin. Part of the original osseous matter of the scales still adheres to the matrix, so 

 that their sculpture (PI. IV. fig. 12, magnified three diameters) is not properly seen ; 

 however, as far as one can judge, it is similar to that of the scales represented in fig. 11, 

 though the ridges seem finer and more numerous. 



Fig. 8 of the same plate is another fragment, which is evidently the upper lobe of a 

 forked caudal fin, shaped like that of Thelodus or of Lanarkia. The scales on the 

 body-prolongation are mostly rhombic, though sometimes nearly square in form, and 

 show traces of an ornamentation resembling that in the two former specimens. On the 

 part corresponding to the fin-membrane the scales become very small, and assume a 

 more or less linear arrangement. 



Microscopic Structure. — In a vertical section of scales from Birkenhead Burn which 

 are evidently referable to this species, I distinguish three layers : — an upper, consisting 

 of sections of the superficial tubercles and ridges ; a middle, or cancellated layer ; and 

 a lower, or laminated one. The ridges and tubercles which make up the outer layer are 

 usually solid in section, though sometimes hollow at the base, and are permeated, but 

 not very closely so, by branching tubules, which pass in a radiating manner from the 

 base to the periphery. In most cases, and especially towards the base of the tubercle 

 or ridge, these tubules are seen to be provided with irregular dilatations which also give 

 off" other branches, and which cannot in fact be removed from the category of lacunae 

 or " bone-cells," however much they may differ from the orthodox lacunas in their 

 frequently irregular shape. In a few instances the tubules seem towards the base of 

 the tubercle to be merely very coarse, without the presence of distinctly differentiated 

 lacunas, but that is rare. There is no surface layer of ganoine. The middle layer con- 

 sists of a cancellated or spongy tissue in which distinct structure is hard to discern ; 

 the lower layer seems to be composed of thin laminas superimposed upon each other, 

 with obscure, elongated specks between, which I have no doubt are also lacunas. I 

 hope to figure these interesting details on a future occasion. 



Observations. — In general form Ateleaspis resembles the Coslolepidas, but on the 

 head the shagreen-bodies have coalesced into small polygonal plates, — behind, into flat 

 rhombic scales ; the thickness of these plates and scales being added to by ossification 

 in a deeper layer of the skin. The tessellated aspect of the head-covering reminds us 

 both of Cephalaspis and Psammosteus, but the position of the crescentic markings, 

 which apparently indicate the outer margins of the orbits, shows a greater affinity with 

 the former genus, which is confirmed by the microscopic examination of the scale, 

 which discloses presence of undoubted bone-lacunas with branching processes. There 

 are, however, neither lateral cornua, nor post-orbital valley, nor pre-orbital fossas ; but 

 as we seem to have here an approximation to the formation of a shield like that of 

 Cephalaspis, I propose for the genus the name of Ateleaspis, or " imperfect shield " 

 (a.Te\r)<; and acrin<;). Further consideration of the zoological bearings of this remarkable 

 form I reserve for the second part of this paper.* 



* The paragraphs above designated as "Microscopic Structure" and "Observations" have been rewritten since the 



