1904.] Allen, The Tamandua Anteaters. 39 1 



animal in the flesh is 957 mm. against 1155 in the Chiriqui 

 form. 



(3) The distinguishing feature of the Vera Cruz skulls is 

 the narrowness of the whole preparietal portion of the skull 

 in comparison with any of the other forms, the braincase 

 being of about the usual width but flat, with the interorbital 

 and preorbital portions of the skull, especially the latter, 

 slender and tapering. It shares in a measure the back- 

 ward position of the occipital condyles with the Chapada 

 form. 



(4) The Chiriqui skulls are the largest of all. The brain- 

 case is relatively narrow, in comparison with the anteorbital 

 region, and also very convex antero-posteriorly in contrast 

 with the depressed and very flat frontal region. The occipital 

 condyles are anterior in position, in strong contrast with 

 the Chapada and Vera Cruz forms. The great length of 

 the rostrum distinguishes it also from the Chapada and 

 Santa Marta types, from which latter it further differs in 

 its large size. 



With these four series of skulls before one for comparison, 

 the differences inter se are impressive and important (see 

 Eigs. 1-4), though hard to express adequately in descriptions. 

 They certainly indicate that the Tamandua group of Ant- 

 eaters is subject to a degree of local variation, at least in the 

 skulls, entitled to recognition. Although some of the forms, 

 and possibly all, are subject to a wide range of individual 

 variation in color, there are still other features by which they, 

 and doubtless other forms, can be separated. 



In view of the apparent continuity of the distribution of 

 the group, it seems best to treat the local forms above in- 

 dicated as merely subspecies. As a starting-point, it seems 

 proper to take Guiana as the type locality of the Linnaean 

 Myrmecophaga tetradactyla, since it was based on the accounts 

 of early writers whose material or observations were made 

 either in the Guianas or in "Brazil." It is at least following 

 the precedents established in similar cases to accept Guiana 

 as the type locality for tetradactyla. 



