8 5 8 



THE AMERICAN NA TURALIST. [Vol. XXXVII. 



tendency to topple over and thus bring the ocular surface under- 

 neath. In this position, however, owing to the convexity of the 

 surface, now underneath, any forward movement would tend to 

 elevate the head and thus counteract the effect of its greater 

 weight and volume. 



When the Cephalaspida:, therefore, left the bottom they 

 probably turned over on to the ocular surface and made their 

 swimming excursions in much the same manner as the above 

 mentioned arthropods. 



Whether the locomotion of the Cephalaspidae was aided by 

 the movements of branchial appendages concealed beneath the 

 arching under surface of the head, can only be determined by 

 the study of more perfect remains than have been as yet dis- 

 covered. Certainly the very strongly concave under surface 

 of the head indicates the presence there of some organs freely 

 movable and of considerable size, projecting ventrally and later- 

 ally from the middle region of the head. 



What we have said in regard to mode of life of the Cepha- 

 laspidae will apply, with slight modifications to other members 

 of the ostracoderms. 



The Asterolepidae, with their powerful, but badly constructed 

 and impractical appendages and large centrally placed eyes, rep- 

 resent the extreme development of the free swimming type. 

 It is quite certain that in some genera the attachment of the 

 slender swimming appendages close to the ventral side, and the 

 very highly arched dorsal surface, must have made it extremely 

 difficult, if not impossible, for these animals to swim with the 

 ocular, or dorsal surface, uppermost. 



The methods of locomotion in the ostracoderms and the 

 arthropods thus have a double value for they show us that ani- 

 mals like Limulus, the Phyllopoda and probably the Trilobita and 

 the Merostommata, adopted when swimming a vertebrate posi- 

 tion with the neural side uppermost, and that the ostracoderms 

 must have frequently assumed the typical invertebrate position. 



In Pteraspis, Cyathaspis and Tolypelepis, the lateral eyes are 

 probably absent, or were covered with such thick layers of the 

 shell as to render them nearly useless. The trioculate median 

 eye, however, is well developed and is represented in Pteraspis. 



