ART. 1 MAMMALS FROM CHINA HOWELL 63 



to this race on the basis of their dusky metapodials. Another charac- 

 ter differentiating chihliensis from Fukien confucianus is the slender- 

 ness of the rostrum ; and the bulhie average a shade smaller. 



RATTUS CONFUCIANUS LUTICOLOR (Thomas) 



Mtis confucianus luticolor Thomas, Abst. Proc. Zool. Soc. Loudon, 1908, p. 45 

 (Yenanfu, Shensi, China). 



/Specimens. — Eighteen: Yenanfu, 7; 15 miles south of Sianfu, 

 Shensi, 3; 50 miles northwest of Tai3'uanfu. 2; 20 miles east of 

 Tai3'uanf u, 5 ; and 30 miles west of Kweihuating, Shansi, 1. 



Two of the specimens from Sianfu are darker and the third much 

 more ruddy than topotypes, and the two former at least probably 

 A'ary toward the race canorios. Those from east of Taiyuanfu might 

 almost as w^ell be placed with chihliensis. Sowerby found these rats 

 most commonly among the rocks along the sides of the valleys. 



RATTUS CONFUCIANUS SACER (Thomas) 



Mus confucianus saccr Thomas, Proc. Zool. Soc. London, 1908, p. 6 (Cliefoo, 

 Shantung, China). 



JSpecimens. — Twenty-two from Hunan; Changsha, 2; Yochow, 20. 



I have had no opportunity to examine topotypes of this race, but 

 as I understand sacer it is an illy defined subspecies representing 

 various degrees of intergradation between confuciarms and chUA- 

 ensis, and farther west, luticolor also. These Yochow skins are con- 

 siderably brighter than true confuciawiis but darker than luticolor., 

 and are variousjly intermediate in other ways. It may be noted that 

 Yochow is not in Honan, as mentioned by Allen (1926) but in Hunan, 

 which puts a somewhat different interpretation upon specimens from 

 that locality. 



RATTUS HUANG HUANG (Bonhote) 



Mris huang Bonhote, Abst, Proc. Zool. Soc. London, 1905, p. 387 (Knatnn, 

 Fukien, China). 



SpecimeTis. — Thirty-nine from Fukien : 70 to 75 miles southwest of 

 Yenpingf u, 31 ; Yenpingf u, 2 ; Peiliang, 1 ; Foochow, 1 : and Kuatun, 4. 



This fine series seems to be perfectly typical, although only the 

 very largest specimens equal the measurements usually given for this 

 species. On the other hand, the smallest adult is considerably larger 

 than the dimensions given for ling. In the series at hand there is 

 comjDaratively little color variation save what is ascribable purely to 

 wear, but there is much as regards spininess of pelage. The fresh, 

 unworn coat has very few and slender spines but the latter increase 

 in number and size as wear of the hairs progresses, until in very rag- 

 ged individuals the middle back and rump are covered thickly with 

 them. 



