THE FAMILIES AND GENEKA OF BATS. 15 



teristic of families. In a group like the bats, where divergence from 

 ordinary mammals has been chiefly in the direction of changing the 

 front limbs from legs to wings, it seems necessary to give special 

 weight in classification to such characters as these. 



Pteropidw.— The very characteristic structure of the humerus in 

 this group shows no important variation in the numerous genera 

 that I have examined. The head is broadly rounded, slightly oval in 

 outline, well defined everywhere, except on the side next the trochiter, 

 into which it merges. Trochiter ill-defined, low and broad, scarcely 

 rising above base of head, to which it is joined by a level, smoothly 

 rounded surface. Trochin slightly higher and narrower than troch- 

 iter, and much better defined, a well-developed groove, about as wide 1 

 as its lateral diameter, separating it from head, and a similar though 

 even more strongly defined groove isolating it dorsally from the low, 

 broad, deltoid crest. Shaft somewhat triangular proximally, flat- 

 tened distally, its sigmoid flexure distinct. Capitellum moved out- 

 ward slightly beyond line of shaft and entirely obliterating epicon- 

 dyle. Epitrochlea large, its lateral diameter more than half that of 

 capitellum, its spinous process distinct, though short. 



Microchiroptera. — The least specialized type of humerus in the 

 Microchiroptera is found in the Khinopomidae and Emballonuridse. 

 In these families both trochin and trochiter have risen about to level 

 of upper surface of head of humerus, a development which causes 

 little change in the form of either tubercle as compared with its con- 

 dition in the Pteropidse. The alteration in form is most noticeable 

 in the trochiter, which is now well defined throughout, and separated 

 from head by a distinct groove over which the smooth surface of 

 the head is extended. This groove is narrower and better defined 

 than that dividing the head from the trochin. Though so distinctly 

 increased in size the trochiter is not large enough to reach the edge 

 of the glenoid fossa of the scapula, the shoulder joint remaining 

 therefore strictly single. Head globular in Bhinopoma, somewhat 

 compressed in the Emballonuridse. As in all other Microchiroptera 

 the deltoid crest is high and knifelike, very different from the low, 

 rounded form in the Megachiroptera. The shaft of the bone is 

 nearly terete throughout, except where distorted by the deltoid crest, 

 and a little flattened distally. It has a slight single curvature. 

 Capitellum essentially as in the Pteropidse. 



A slight modification of this type is found in the Noctilionidse. 

 Here the head of the humerus is greatly compressed and strikingly 

 ridgelike. The well-defined trochiter is very small and narrow, 

 though rising fully as high as in the related groups, while the trochin 

 is greatly increased in size, being nearly as large as the head. The 

 groove .separating the trochin from the well-developed though rather 

 small deltoid crest is ill-defined.- 



