Vol. II, Pt. I] VAN DENBURGH— GALAPAGOS TORTOli,ES 247 



it very difficult to determine positively from it which race his 

 specimen represented. This being true, the fate of Harlan's 

 specimen becomes of much interest, since only from it can we 

 obtain the desired data. Unfortunately, there is little doubt 

 that this specimen no longer exists. Baur,^ it is true, men- 

 tions examining Harlan's original specimen in the Philadelphia 

 Academy of Natural Sciences, but the specimen to which he 

 refers does not agree with the description or measurements 

 given by Harlan, and, indeed, has never been regarded as 

 Harlan's type by the authorities of the Academy. At my re- 

 quest. Dr. Arthur E. Brown has been kind enough to look the 

 matter up, and, while the Philadelphia Academy has no com- 

 plete records of its museum in those early days, he has found 

 in an early volume of the Journal, in the list of donations in 

 February, 1827, mention of a "Testudo elephant opus from 

 Richard Harlan, M. D." As this was only five months after 

 Harlan's paper was read, it seems fair to presume that the 

 specimen presented was the one which had served as the basis 

 of his description. With Mr. Witmer Stone, Dr. Brown then 

 "made a careful search through a lot of odds and ends of old 

 material, with the result that we found the cleaned leg bones 

 of one side, and a part of the legs of the other side with dried 

 skin still on them, of a Testudo about the size of Harlan's type, 

 with an index number (366) making it almost certain that it 

 came from Harlan." Dr. Brown says, "In the opinion of both 

 Mr. Stone and myself, these fragments are probably all that is 

 left of the type of T. elephant opus, which had apparently been 

 mounted, but long ago became dismembered, leaving only these 

 scraps which do not bear any of the specific characters." 



It therefore seems fairly certain that no one ever will know 

 from the specimen itself what Harlan's Testudo elephantopus 

 really was, and that any opinion must be based upon the meager 

 data to be derived from Harlan's original description and plate. 

 I have already stated that I believe these to be scarcely adequate. 

 The points of value in this connection are : that it was a young 

 individual, probably a female, with elevated central areas and 

 concentric ridges on the plates, pectoral plates meeting exten- 

 sively on the median line — therefore not from Chatham Island ; 

 breadth over curve (22.6) greater than length over curve 



BAm. Naturalist, Dec. 1889, p. 1043. 



