Xlviii INTERRELATIONS OF GENERA. 



III. Interrelations of the genera of Megachiroptera. 



It has been considered desirable to give here a summary of the 

 mutual affinities and probable phylogen)' of the genera, " sections," 

 and subfamilies of living Fruit-bats, omitting as far as possible all 

 discussion of details. For these latter the reader is referred to 

 the paragraph " Affinities " under each genus, in the systematic 

 part of this Catalogue. 



Rousettus (in its typical form, as represented by the species 

 of the subgenus l^ousdUn-) is one of the least specialized genera of 

 living Megachiroptera. The rostrum is moderate in length and iu 

 no respect peculiarly modified, the premaxillse in simple contact 

 with each other in front (not ankylosed together, nor spaced), the 

 lateral margins of the postdental palate forming straight lines 

 converging backward, the postorbital processes are short, the post- 

 orbital foramina present, and the brain-case perfectly unmodified 

 in general shajje. On the other hand, that all the most primitive 

 cranial characters found in any living Fruit-bat should be united 

 in the skull of this or any other living genus, is hardly to be 

 expected; it is not difficult, therefore, to point out a few single 

 skull characters in which Eouseitus stands a little higher than one 

 or another genus of Megachiroptera : as in the large majority of 

 Fruit-bats the infraorbital canal is quite short (considerably longer 

 in MeJonijcteris, Nesonycteris, and Noto^^tens), and the ascending 

 branches of the premaxilla3 as narrow at their upper extremities 

 as near the alveolar margin (much less narrowed above in the three 

 genera just mentioned); the palate is perhaps rather too broad to 

 be considered quite unmodified in form, the tympanic bones, 

 though typically annular, are rather broader than in most other 

 genera, and the facial portion of the skull is distinctly, though not 

 very strongly, deflected against the basicranial axis (more so than 

 in the majority of Epomophorine and Cynopterine genera ; but it 

 should be noticed that the deflection of the facial axis often varies 

 to some extent in undoubtedly closely interrelated genera, and 

 sometimes even in species of one genus : see anted, p, xxiii). The 

 dentalformula is unmodified Megachiropteran (^ — ^ incisors, j. cheek- 

 teeth), the teeth in no way peculiar, except in so far as the inner 

 cusp of p' and p^ is completely fused with the outer, and that of 

 p^ uearly so, characters shared with a large number of other genera. 

 The palate-ridges are arranged in regular and almost equidistant 

 curves over the whole of the palate, some of the posterior ridges 

 always distinctly interrupted iu the middle; the tongue papilUe 

 essentially unmodified (there is a faint tendency to a lengthening 

 of the conical papillae at the extremity of the tongue : compare 

 Macroglossina). A short tail is present, the second finger clawed, 

 the wing-membranes inserted on the preaxial side of the foot, 

 the prevailing colours of the fur a dark tinge of hair-brown or 

 brown, the secondary sexual differentiation inconspicuous (males as 



