472 



Bulletin American Museum of Natural History. [Vol. XXXVII, 



above and along the sides of the body below, without white at the base of the pata- 

 gium. Wing membranes blackish, thick and leathery. 



The present group, whether considered as of generic or as only of sub- 

 generic value, is a natural group, much more sharply defined from all others 

 than is Cheer ephon from Nyctinomus, although these two groups have been 

 long currently given recognition as genera. In the present connection it is 

 given subgeneric rank. The four known species are here described as new, 

 although one of them may be the same as Nyctinomus thersites of Thomas, 

 which apparently belongs to the Allomops group. They are small to medium 

 sized species, one of them, Allomops nanulus, being the smallest known 

 African molossid. 



In working out this group, in connection with other African molossid 

 material, I have been impressed with the small value of the character most 

 commonly considered as an important index to the relation of species re- 

 ferred to Nyctinomus and Cheer ephon, namely, the emargination or non- 

 emargination of the front border of the premaxillse. In one instance in the 

 study of a series of 23 specimens collected the same day at the same locality, 

 and evidently beyond question conspecific, I took up the skulls first, with 

 skulls of other species, to determine their generic affinities. Later on 

 collating the skulls (marked only with their catalogue numbers) with the 

 skins to which they belonged, I found I had referred part to Nyctinomus 

 and part to Cheer ephon! Further study of the same series showed that in 

 this case at least emargination and non-emargination of the front border 

 of the premaxillse was largely dependent on the age of the specimen, the 

 vacuity behind the incisors becoming gradually closed by ossification with 

 the increased age of the specimen. 



In one of the species here referred to Allomops, represented by six adult 

 specimens, two, on the basis of the condition of the premaxillse, are typically 

 Nyctinomus and the other four typically Chcerephon. They were otherwise 

 indistinguishable, either on the basis of external and cranial characters or 

 by the age of the specimen. In another species here referred to Allomops, 

 represented by seven specimens, all of the skulls had the front palatal border 

 emarginate, but as all of the other characters, both external and cranial, 

 were typical of the Allomops group, it seemed proper to give preference to 

 the combined weight of the Allomops characters, despite the Nyctinomus- 

 like condition of the premaxillse. 1 



1 After writing the above I first became aware of Oldfield Thomas's paper, entitled ' On a remarkable 

 new Free-tailed Bat from Southern Bombay' (Journ. Bombay Soc. Nat. Hist., XXII, pp. 87-91, 

 April, 1912), in which he states that he had reached the same conclusion regarding the emargination 

 or non-emargination of the front border of the premaxillae in the Old World molossid bats. He says: 

 " I may note also that all sorts of intergradations are found in the premaxillae and that it is often almost 

 impossible to decide whether a given specimen should be referred to one genus or the other [Nyctinomus 

 or Chserephon] by this character of the premaxillae." 



