COLOUR PHASES OF THE RUFOUS HORSESHOE-RATS. 26r> 



iVom the tact that the auburn and brown areas are about equal in 

 extent, be only about halt-way throug-h the nioiilt, and in that case 

 we should find somewhere in the fur evidence of new hair coming 

 up, shorter than the rest; but there is no evidence whatever of this 

 kind, all the hairs are full grown, all of their proper length, there is 

 on the whole surface of the animal not a single spot showing a new 

 crop of hair. Thirdly, the moult, as we have already seen (Octo- 

 ber specimens, group one, above), proceeds along definite lines, 

 the new fur of the upperside first appearing on the back of the 

 neck, spreading outward to the sides of the neck and backward 

 along the median line of the dorsum, so that the last portion of the 

 old fur to disappear is an orange stripe on each of the back and 

 rump along the lateral membrane (see figs. 1, 2, and 3). A glance 

 at fig. i will show how entirely difierently our specimen behaves, 

 both above and beneath. For these reasons there can be no doubt 

 that our specimen w^as killed while changing the colour of its new 

 full grown coat. This result is further supported by an examination 

 of the four other individuals of this groiip, all of which are of the 

 same age. The change of colour is in all four carried a aood deal 

 further than in No. 1949, is in fact nearlij completed. The upper- 

 side is approximately auburn (II, 11, m) above, this colour gradu- 

 ally lightening to cinnamon (XXIX, 15") toward the base of the 

 hair, while the underside is ochraceous-tawnj^ (XV, 15', i). But 

 the change is only " nearly " completed ; there is still, in all four 

 examples, a conspicuous amount of greyish colour on the throat 

 (this disappears, as we shall see, in the follo\A ing month), and in one 

 specimen (1948) there is a small patch of grey left on the belly. 



To the conclusions derived from groups one and two of the Octo- 

 ber individuals (see above) we can now add those obtained by our 

 examination of group three, viz., (Ij Bats of the year undergo an 

 autiimn moult, like the older individuals; (2) this new coat is quite 

 similar to the new coat of older individuals, dark brown above, 

 mouse-grey beneath ; (3) but this new coat (at least in the indivi- 

 duals we have seen thus far) when fully developed changes its colour 

 into auburn above, ochraceous-tawn}^ beneath ; (4) this colour 

 change probably takes place comparativeh^ rapidly. 



Even if we had never seen the five October specimens of " group 

 three " we shoiild have been forced to the conclusion that shortly 

 after the autumn moult the colour of the new coat "tnust change. 

 We have seen that when orange specimens moult in October they 

 put on a mouse-coloured coat. But we shall find that already in 

 November (and then throughout the " winter ") plenty of aiiljurn 

 and orange specimens occur. How could this fact be explained, 

 otherwise than by a change of colour of the fuUgrown hair ? 



We are now evidently well on the way to a better understanding 

 of the colour phases in this bat, andean proceed to examine more 



14 



