1 98 Bulletin American Museum of Natural History. [Vol. XXII, 



and the antero-inferior convex. More rarely it is practically quadrate, 

 or the anterior portion may be rounded on the upper as well as on 

 the lower margin. It varies not less in proportions than in form: 

 in perhaps the larger number of cases the antero-posterior and vertical 

 diameters are nearly or exactly equal; in other cases the vertical 

 greatly exceeds the antero-posterior, and in other cases it may be 

 less. Six skulls selected to show variation in this respect have the 

 antero-posterior length and the vertical length of the lachrymal as 

 follows: 14.5 X n, 14 X 14, 15 X 13, 13.5 X n, 15 X 14, i4X 12 mm. 

 Figures 1-3 illustrate examples of purely individual variation. 



The number of free dorsal bands in the carapace, along the median 

 line, is 8 in 1 8 specimens out of 19, the other having 9, or 9 and 10 

 bands respectively in all. Four Texas specimens also have 8, while 

 one from Trinidad, B. W. I., 4 from Santa Marta, Colombia, and 3 

 from Chiriqui, Panama, have each 9 free bands, or 10 in all. 



In regard to the relationship of the nine-banded armadillos of 

 Texas and Mexico to those of Central America and South America, 

 little can here be said, owing to lack of proper material. These ani- 

 mals are subject everywhere to such a wide range of individual 

 variation that large series of specimens are necessary from many 

 localities in Central and South America for comparison with those now 

 in hand from Mexico before satisfactory conclusions can be reached. 

 It appears evident, however, that the Mexican form is subspecifically 

 separable from those occurring in Central America and southward, 

 but I fail to see any appreciable differences that are of any importance 

 between specimens from Texas * and western Mexico. 



4. Tayassu angulatum sonoriense (M earns). 

 (PLATE XX.) 



Twenty-four specimens, of which 14 are fully adult, and the others 

 more or less immature; both sexes are about equally represented. 

 They were all collected at or in the immediate vicinity of Escuinapa, 

 as follows: Dec. 24-26, Jan. 3-10, 19, Feb. 17-23, March 4, April 19 

 22, May 3, 23. 



As shown below, this series of specimens presents a wide range of 

 individual variation, both in coloration and cranial characters, but 

 the animal from southern Sinaloa seems better referable to sonoriense 

 than to either of the other described forms of the T. angulatum group. 

 None of the several recognized forms of this group seems to be at 



i Recently separated by Mr. Vernon Bailey as Tatu novemcinctum texanum. (N. Amer. Fauna 

 No. 25, p. 53, tigs. 5 and 6. Oct .. 1905.) 



