from the centre of the inner s 

 m. ;-. They are perhaps best 



These markings on the d 

 believe, best explained as 



They suggest the 

 sal surface of tr 



near the median anterior part of the head of 

 same tubercles were figured by A. Kunth 

 figures, the resemblance of the shields to t 



to understand how the usual type of vertebrate eye could be pre- 

 served in the form of these tubercles. If they are lateral eye 

 tubercles, they, like the compound eves of arthropods, must have 

 been enclosed by a firm dome-like covering, eontinuous with the 

 outer layers of the shell. I have never seen any specimen of 

 Pteraspis in which these so-called lateral eye notches were actu- 

 ally covered by an extension of the outer shell layer, although 



pis and Tremataspis. It is possible that there are two pairs of 

 such notches, or openings, in Pteraspis, Cyathaspis and Tolypele- 

 pis, one serving for the lateral eyes and the other for the attach- 

 ment of swimming appendages similar to those of Tremataspis. 



