560 



PROF. E. W. CLAYrOLE 01^ THE STRTJCTFKE 



rather encourage the belief that the 

 outer layer had an even edge. In this 

 case the row of tubercles forming the 

 serration probably indicates the points 

 of attachment of muscles. 



Tt is a little remarkable, and to me 

 unexpected, that in no specimen yet 

 found have I seen any indication of the 

 canal-system so clearly apparent in the 

 plates of Holaspis as represented in 

 Lankester's figures and also in the 

 shields of other Pteraspids. It is scarcely 

 likely that the canals were altogether 

 wanting, and the absence may be due to 

 the mode of fossilization. Neither on the 

 casts of the outer surface nor on the sur- 

 face itself when preserved, though show- 

 ing the most delicate striation, have I 

 found the least trace of such structure. 



Nor have I seen in any of the lateral 

 plates any sign of the (so-called) 

 branchial tubes or holes found in the 

 laterals of Pteraspis. These do not 

 occur in Cyathaspis, and this fact, no 

 less than the structure of the shield, 

 indicates a closer relationship with that 

 genus than with the former. 



In addition to the specimen herein 

 described I have among the material in 

 hand several very small shields much 

 more convex than those of Palceasjois^ 

 and not exceeding 1 inch in length. 

 They suggest very strongly by their out- 

 line a relation to Cyathaspis, but they 

 may be merely young specimens of the 

 larger genus. Some of the stones also 

 abound with coprolitic matter, which is 

 mostly of a lighter colour than the rock. 



A comparison of my restored figure 

 with that of Pteraspis as given by Mr. A. 

 Smith Woodward shows that no very 

 great modification of strncture is neces- 

 sary to bring it and Palceaspis into 

 close resemblance. The curve there 

 shown on the upper edge of the ventral 

 plate might easily be so increased or 

 developed as to result in two lines meet- 

 ing at an obtuse angle, as seen in 

 Palceaspis, and this change would con- 

 vert the one lateral plate of Pteraspis 

 into two by dividing it in the middle. 

 This might result either in forming 



rig. 8. — Attempted resto- 

 ration of P. americana. 



^ 



[Showing details ascertained 

 or rendered probable so 

 far as at present known.] 



