SACCOMYIDiE— PEHOGNATIlIDINiE— P. PENICILLATUS. 507 



upon analogy, by the peculiar character of the long penicillate tail, already 

 fully described. 



I find it difficult to describe the color of the upper parts satisfactorily. 

 It was called by Baird "a light dull yellowish-brown or tawney, lined with 

 dark-brown". The prevailing tone is decidedly "gray'' in comparison with 

 the strong mixed yellowish and blackish of P. fasciatus — somewhat the shade 

 of dilute coffee and milk. The lighter hairs are dull yellowish-gray, with 

 darker lining of hair brown. There is a general tinge of the palest possible 

 cinnamon. All the bases of the hairs are of the color of fresh-cut lead. 

 There is no appreciable lateral stripe of different color from the upper parts, 

 although, just along the line of junction with the white, the color lightens a 

 little for lack of the darker brown lining. The extreme tip of the snout, the 

 cheeks, the whole fore leg, the inner aspect of the hind leg, the feet, under 

 side of tail, and under parts generally, are white; the line of demarcation 

 runs straight from nose to hind leg, where it drops down to the heel. The 

 upper surface of the tail and the brush at the end are hair-brown. Some of 

 the whiskers are black; others, like the claws, are colorless. 



The external sexual characters of this and other species of the family 

 are readily recognizable. In consequence of the development of the os penis, 

 the preputial sheath is a resisting prominence immediately in advance of the 

 anus. The same part in the female shows a conical flap depending from the 

 front of the vulva. 



Having only about a dozen adult specimens of this species to examine, 

 the full range of variation can only be given with approximate accuracy. 

 Still the series is a very good one as far as it goes, and the number of alco- 

 holics afford reliable measurements. Dr. Woodhouse's type is much laded by 

 twenty years' exposure to the light. The others agree perfectly in colora- 

 tion, and the differences in size are slight. I may here remark that, in Pro- 

 fessor Baird's table of measurements the dimension (nose to tail) given from 

 the dried specimens there enumerated is certainly over the mark. Thus, 

 Woodhouse's type is given as 4.12 long, whereas the original description of 

 the same specimen says 3.50. 



