354 Mr. G. E. Dobson on a Natural 



cult to regard these genera, though differing so widely in denti- 

 tion, as belonging to different groups; yet Desmodus has been 

 considered the type of a primary division of the Chiroptera. 



Megachiroptera. 



Pteropidae. 



Analysis of Natural Affinities. 



The natural affinities of these bats with any of the families 

 of Microchiroptera is not easily traced. Some zoologists con- 

 sider that they form but another family of Chiroptera and can- 

 not be separated into a distinct suborder, and support their 

 opinion by enumerating the many points of agreement in struc- 

 ture between Macroglossus and the species of the group Olos- 

 sophagce of Phyllostomidre. But, in the form of the wings, 

 and even in the microscopic characters of the fur, the Pteropida? 

 are also connected with the families of the Vespertilionine 

 alliance ; while they differ altogether (as described in the defi- 

 nition of the suborder) from the Microchiroptera in the general 

 form of the ear-conch, of the teeth, and of the bony palate — 

 also in the tail being inferior to the interfemoral membrane, 

 not contained in it or appearing on its upper surface, as in all 

 other families of Chiroptera. These very different natural 

 characters lead me to trace the descent of the Pteropidae from 

 a group of Palseochiroptera, distinct from that from which 

 the Vespertilionine and Emballonurine alliances have sprung, 

 but with affinities to that section of the latter group from which 

 the Emballonuridse are derived. Thus the connexion of the 

 Pteropidae with the Emballonurine alliance has left traces in 

 the index finger of Rhinopoma with two distinct bony phalanges 

 (found in no other genus of Microchiroptera), in the very large 

 and peculiarly shaped feet of Noctilio and in the form of its 

 wings, in the well-developed premaxillary bones and upper 

 incisors of Phyllostomidse generally, and particularly in the 

 long tongue and muzzle and feeble molars of the Glossophaga? 7 

 and in the frugivorous or semi-frugivorous habits of some 

 species of this family. 



It is not, however, in the agreement of certain characters 

 (which may have resulted from similar adaptative causes) that 

 we should seek for proof of special affinity, but in the general 

 agreement of all the natural characters considered together ; 

 and it would be, in my opinion, as unphilosophical to consider 

 the Heaths and Campanulas (to take an example from the vege- 

 table kingdom) united in one order because they agree in the 

 insertion of the stamens, or, conversely, to distinguish them 

 merely by the different mode of dehiscence of the anthers. 



