262 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXV. 



not a single specimen from the whole of the rainy season (June to 

 September). As it is, however, the series is sufficiently complete 

 to solve the principal problem, namely, what do these colour 

 " phases, " this unparalleled variability in colour mean? 



The first question we have to settle, before trying to unravel the 

 mystery of the colour phases, are naturally these : — When does this 

 bat change its coat ? Does it moult once or (as usual among 

 Mammals) twice a j^ear ? And what is the new coat like ? Does 

 it show the same " individual variation " in colour as exhibited 

 when glancing down the whole series of specimens befoi'e us ? 



Fortunately the material gives clear answers to these preliminary 

 questions. Bh. rouxi moults twice a year, in this part of India in 

 October and again in May, and in both cases the new fur is dark 

 brown above, grey on the underside of the body, — the ordinary bat 

 colour, or very much like the coloar of a House Mouse. 



Where then do the brilliantl}^ orange and the intermediate phases 

 come in ? To give the facts at once, before entering into details, 

 these phases are due tn a (jractual, and on the whole evidently rather 

 rapid, change of the colour of the full grown hair. 



In order to follow the changes of the coat and colours in detail 

 we must now examine the appearance of the specimens month by 

 month, so far as the material allows. 



There are twenty-eight skins dated October (1912), all from one 

 locality (Seringapatam, S. Mysore), and very nearly of the same 

 date (18th and 19th) ; all are females and all fully adult* (epiphyses 

 of metacarpals ossified) ; the actual ages of the individuals are as 

 follows : — Five are about six months old, nine about eighteen 

 months, five two and a half years, four three and a half years, 

 and five four and a half j^ears. For reasons explained below we 

 must put the youngest generation ("bats of the 3''ear," six months 

 old) on one side ; they require special description. The remain- 

 ing twenty-three specimens may conveniently be divided into 

 two groups, those in the process of moulting (fifteen), and those 

 which have just completed, or practically completed, their autumn 

 change of coat (eight). 



(1) Fifteen October individuals, aged from one and a half to 

 four and a half vears, showing^ various stages of moult from the old 

 orange to the new dark brown coat.f — Three specimens (1935, 

 1 936, 1 940) exhibit the initial stages of the change of fur. They are 

 nearly orange-rufous (II, 11, h to j ±) above, approximately mars- 



* That is, full grown ; which does not necessarily imply that they are all 

 sexually mature) this is certainly not the case with the youngest class of indivi- 

 duals) those only six months old). 



t Collector\s numbers. G.C.8. U»34-1941, 1944— 194f., 19.31, 19.59, 19(jl, 1964. 

 _ t All colours are determined from Ridg-way's " Color standards and Color 

 Nomenclature," Washington, 1912, and the references between parentheses are to 

 the plates of that book. 



