234 ANNUAL REPORT SMITHSONIAN INSTITUTION, 193 4 



The form of the posterior part of the trunk and the tail are not as 

 well known, as these parts have been found completely preserved 

 only in one single species from Spitsbergen. The posterior part of 

 the trunk and the tail were probably considerably shorter and thicker 

 than in Tremataspis^ and covered by very solid, thick scales. The 

 structure of the cephalothoracic armature varies considerably in the 

 dilferent forms. In some it is more oval and rather flat (fig. 6, A, B) ,, 

 in others more elongate and almost cylindrical with a circular cross- 

 section. 



How were these forms able to move through the water? They 

 certainly had no paired fins nor any unpaired ones and they also 

 lacked any distinct spines. The only movable part of the animal^ 

 the posterior trunk and tail, were covered by exceptionally solid 

 and thick scales, which prevented the possibility of any strong and 

 lively movement. Thus we see that the primitive Cyathaspids had 

 neither any organ of equilibrium nor steering organs and that fur- 

 thermore their organ of propulsion, the hind part of the body and 

 the tail, were also less flexible and movable than in tadpoles or the 

 oldest Osteostraci. From this we can conclude that the Cyathaspids 

 were very poor swimmers and spent their time chiefly on the bottom, 

 perhaps partly buried in sand and mud. 



The next step in the adaption of the Heterostraci is found in the 

 large group of the true Pteraspids (fig. 7). They are all flatter, more 

 conspicuously compressed dorso-ventrally. The dorsal shield, which 

 was undivided in the Cyathaspids, has been split into several separate 

 plates. Of greater interest, however, is the development of the 

 different spines and projections. 



To begin with, we find distinct lateral projections on each side 

 behind the gill-opening (fig. 7, A 3). These projections vary 

 greatly in different forms, as one may see from the figure. In some 

 it assumed almost fantastic dimensions (fig. 7, D, E), in others it 

 was almost completely obliterated (fig. 7, C). As in the case of 

 the similar projections in Cephalaspids it is evident that these 

 spinelike parts in the Pteraspids served primarily as organs of 

 equilibrium, but secondarily also, and then particularly in the forms 

 with very large spines, as a gliding organ. Apart from these two 

 symmetrical lateral spines we find also a median dorsal spine de- 

 veloped in most of the Pteraspids. This dorsal spine is located at 

 the posterior end of the dorsal shield and is directed obhquely up- 

 ward and backward (fig. 7, A, B, C, D). This spine corresponds 

 entirely to the comblike development of the posterior part of the 

 cephalic shield in the Cephalaspids and must assuredly have served 

 as an organ of equilibrium. 



