EAST AFRICAN MAMMALS IN NATIONAL MUSEUM. 43 



CROCIDURA NYANSiE KIJAB.^ Allen. 



1900. Crocidura kijabse Allen, Bull. Amer. Mus. Nat. Hist., vol. 26, p. 173. 



Marcli 19. (Kijabe, British East Africa; type in Amer. Mus. Nat. Hist.) 

 1910. Crocidura nyansse Roosevelt, African Game Trails, Amer. ed., pp. 480 and 



487; London ed., pp. 491 and 498. 



Specimens. — Twenty-five, from localities as follows: 



British East Africa: Aberdaro Mountains, 11,000 feet, 1 (Hel- 

 ler); Laikipia, 1 (Heller); Mount Kenia, 3, including 2 in alcohol 

 (Mearns, Loring, Heller) ; Mount Umengo, 1 (Heller) ; Naivaslia 

 Station, 17 (Loring); Nakutichu River, Naivasha Plains, 1 (Heller); 

 Nyeri, 1 (Loring). 



Considerable individual variation in color is sliown in this series 

 of skins. The upperparts range from clear, rich, reddish brown to 

 almost blackish, and the bellies from light gray or buff to a shade 

 almost as dark as the back. The type-specimen of Jcijahse, kindly 

 lent me by the American Museum, through Dr. J. A. .^Ulen, is a skin 

 of the darkest style, with the body almost unicolor. In the Naivasha 

 series, however, are specimens of both extremes, and it is evident that 

 color subspecies of this large shrew should not be recognized unless 

 based on long series of specimens. Dollman, who probably had 

 access to more skins of typical nyansse than I have seen, recognized 

 this race on the generally darker color of the series when compared 

 with the average color in a series of nyansse. On the basis of our 

 collection alone I should not have recognized the subspecies, although 

 there are several skins considerably darker than any specimen in our 

 small series of nyansse proper. 



The single specimen from the Taita Hills (Mount Umengo) appears 

 inseparable from examples in the Naivasha series, although its locaUty 

 is much nearer to the type region of a related species not represented 

 in our collection, Crocidura martiensseni Neumann, than to the near- 

 est point from which we have skins of C. n. hijahx. Just what is the 

 relationship between Mjahse and martiensseni seems uncertain, but 

 the KUimanjaro form is certainly a considerably larger shrew than 

 either Mjahse or typical nyansse. Heller has made the following 

 notes on the type specimen of martiensseni in Berlin : 



Type 9) alcoholic. No. 8909; Loc. Magrosso; specimen not marked type; skull 

 extracted and cleaned. Skin, in alcohol, uniform mummy-brown, belly same as 

 back; hind foot, 22 millimeters. Skull: Condyloincisive length, 33.5; breadth of 

 braincase, 13.5; upper tooth row, 15; mandible, including incisor, 21.3. 



Roosevelt, Heller, and Loring make the following remarks on the 

 specimens of C. n. hijatse collected by the Smithsonian African Expe- 

 dition:^ 



Chiefly in the high country, near water courses; found round the edge of the forest, 

 at Kenia and Kijabe. A fierce, carnivorous creature, preying on small rodents as 



'African Game Trails, pp. 480 and 4S7. 1910. 



