Arrangement of the Chiroptera. 353 



shortness of the middle finger, and the existence of a pair of 

 abdominal (not pubic) appendages (like those in Megaderma, 

 but smaller) connect this remarkable genus with the Nycteridse. 



Mystacina, represented by M. tuber culata (Gray), Tomes, 

 from New Zealand, connects the Emballonurinoz with the 

 Molossince. It possesses characters common to both subfamilies, 

 resembling the species of the first in the form of the ears and 

 in the short tail perforating the interfemorai membrane, the 

 second in dentition and in the general form of the body ; so 

 that its position is clearly among the Molossince. Unlike the 

 other genera of Emballonuridaj, the middle finger has three 

 phalanges, as in Thyroptera among the Vespertilionidje, and the 

 first phalanx is flexed downwards upon the metacarpal bone. 



Noctilio connects the Emballonuridse, especially the sub- 

 family Emballonurin<e, with the Phyllostomidae through the 

 Lobostomince. I have placed Noctilio provisionally in this 

 family ; for its affinities appear to me to be with Chilonycteris 

 and the genera allied thereto. These affinities are shown by 

 the presence of a small incisor on each side external to the 

 large central upper incisors, by the development of the mastoid 

 and paroccipital processes, by the short first phalanx of the 

 middle finger, which is not flexed upwards (in repose) on the 

 dorsal surface of the metacarpal bone (as in other species of 

 Emballonuridaj), and by the erect cutaneous ridges on the chin, 

 and especially by the form of the nostrils and ears and by the 

 microscopical characters of the fur. 



The Phyllostomidae are divisible into two subfamilies, 

 the Lobostomince and the Phyllostomince : — the first containing 

 the genera Chilonycteris, Pteronotus, and Mormops, united by 

 Prof. Peters into a group Mormopes ; the second the remaining 

 genera of the family. If the genus Noctilio were placed among 

 the Lobostomince, it would form a separate group distinguished 

 by the presence of two phalanges only in the middle finger. 



I have followed Prof. Peters's divisions of the genera of 

 Phyllostomince, retaining his names of subfamilies for what I 

 consider groups of allied genera*. Between these groups 

 transitional forms exist, which have not yet been found between 

 the Phyllostomince and the Lobostomince. Thus Rhinophylla 

 leads from the Vampyri to the Glossophagce ; and the close 

 connexion of the Vampyri with the Stenodermata is seen in 

 the similarity of the warts of the lower lip. Brachyphylla is 

 evidently so closely related to Desmodus that it appears diffi- 



* Prof. Peters has devoted so much attention to the Phyllostomidae, and 

 his knowledge of the species of this family is so profound, that I hesitate 

 to make any change in his system of grouping the genera. 



Ann. & Mag N. Hist. Ser. 4. Vol. xvi. 25 



