ART. 3 COXCEFiNING THE ARMOR OP TURTLES HAY 9 



the space is filled by outward expansion of the humeral and abdomi- 

 nal scutes. In Terrapene the space is obsolete. 



It seems worth while to try to account for the extension of the 

 scutes beyond the bone on which they originate. Briefly expressed 

 the explanation is that they were originally associated each with an 

 epithecal bone which later ceased to support it, leaving it to wander 

 until it reached the obstructing border of its neighbors. Sometimes 

 parts of three or more bones are traversed to meet the boundary; 

 sometimes only two. By the superior growth of epithecal bones 

 along certain lines the keels of the early ancestors of turtles were 

 produced. In the course of time some of the bones of the keels be- 

 came enlarged at the expense of other bones and of the scutes over- 

 lying them. Along the middle of the back of Toxochelys we find 

 enlarged epithecals reposing on the neurals. We may suppose that 

 the most favorable position of an epithecal would be on or near a 

 suture between two neurals, since blood vessels and nerves could more 

 readily reach it. If now an epithecal of the size of those of Toxo- 

 chelys were lodged across each neural suture the neural bone itself 

 would tend to be suppressed; and among the early Thecophora the 

 neurals themselves were gaining the upper hand. Hence about alter- 

 nate epithecals were suppressed. Although the dominating epithecal 

 was itself later dispensed with, the horny scute associated with it 

 would expand forward to reach the scute situated the length of two 

 neural plates in front. The same explanation will apply to the 

 fore-and-aft width of the costal scutes, which may cover one costal 

 bone, a part of the one behind, and a part of the one in front. 



When we examine the marginal scutes we find each one covering 

 a portion of one peripheral and a larger portion of the next periph- 

 eral in front. It seems to the writer that the explanation is as 

 follows: The epithecals of this row were small and one for each 

 peripheral did not menace the development of the latter. Hence its 

 scute could spread only over a part of the next peripheral in front. 



8. DR. G. K. NOBLE'S OBSERVATIONS ON CHELYS 



In 1923 Dr. G. K. Noble reviewed my paper ^ of 1922 and gave an 

 interesting account of his observations made on a young matamata 

 of about three-eighths the size of the animal described in the present 

 article. In his specimen he was able to find no traces of the epi- 

 thecal bones. Considering this young animal in connection with the 

 adult in which the bones were absent Noble concluded that my 

 "hypothesis should not be accepted without additional materials." 

 The present paper describes the additional materials desired. 



•Anrer. Naturalist, vol. 57, pp. 377-379. 



