222 



N'ow, in his remarks, Dr. Andersen emphasizes the value of 

 skull-characters which in the "Catalogue of Chiroptera" were only 

 made use of (in connection with species here discussed) for subspecific 

 differentiation. Distinctions based on skull chai^acters are not to be 

 lightly attacked, but I think in that case less importance should be 

 attached to the ear dimension. 



If we accept both brachyotis and angulatus as sub-species of 

 hracJiyotis we have two forms of one species occupying the same 

 localities : which is contrary to generally accepted opinion, for while 

 nearly all zoologists recognize the inter-gradation of geographical 

 races the great majority are not prepared to i"ecognize overlapping 

 or intei^mixture ; that is to say you cannot have two sub-species of a 

 species li\T[ng side by side. 



If, however, you are prepared to admit that two forms of a species 

 do occur together there is no reason why three (or more) should not 

 be accepted, so this alone is not good cause for not regaining 

 tittheechtilus as a brachyotis too, though that course might certainly 

 be less convenient. 



Our suggestion was somewhat superficial, but at the moment we 

 simply had in view the Cynoptertis bats of the Malay Peninsula and 

 Dr. Andersen's main dividing line of the section under discussion. To 

 transfer angulatus across that line would not improve matters for then 

 we should have two bats of the same species {sphinx), titthsecheilus and 

 angulatus again living side by side. To overcome this difficulty we 

 have three alternatives to choose from. Either angulatus has no 

 real existence, the material forming it being part sphinx and part 

 brachyotis — not very probable; or it is a very variable form of 

 brachyotis of which the typical race is non-existent from Sumatra to 

 Annam and Assam; or it is an independent species. In any case the 

 character of the ears does not seem a very good one to rely upon for 

 the separation of the primary forms — providing our measurements 

 and those given for the type series are correct. C. Boden Kloss. 



