''m'i^] ALLEN — MAMMALS 57 



smaller throughout. The upper unicuspids form a graduated series, the 

 anterior-most the largest, with a longer base relatively than those succeed- 

 ing. The gradation between the second and third unicuspids is rather 

 more abrupt than between first and second or third and fourth, so that a 

 distinct tendency to a paired appearance results. The fifth unicuspid is 

 rather large and fully in the tooth-row. The tips of all are pigmented. 

 The protocones of the upper molariform teeth also are well pigmented, and 

 the hypocones of the first and second molars are barely touched with color 

 in the type and in four of the other specimens in the series of six. The 

 middle lobe of the first lower incisor is the best developed, the first is much 

 lower and broader, the third smallest of all. As in S. buxtoni, there is a 

 tendency to the development of a minute posterior cusp on the long- 

 crowned second lower incisor. In three of the six specimens this cusp is 

 very distinct, but it is unpigmented. 



Measurements. — The type was measured by the collector, as follows: 

 total length, 85.5 mm.; tail, 28.5; hind foot, 12.5. The extremes of the 

 six specimens are: total length, 80-94 (average 84.5); tail, 28-34 (average 

 29); hind foot, 11.5-12.5. 



Skull: condylobasal length, 17.7 mm.; basal length, 15; palatal length, 

 7.3; least interorbital width, 3; greatest width outside last molars, 4.2; 

 greatest width of brain-case, 8.3; maxillary tooth-row, 7.3; mandibular 

 tooth-row, 6.9. 



Remarks. — The apparent relationship of this shrew to macropyg- 

 maeus of the Kamtchatka peninsula I have ventured to express by 

 the trinomial. It is a shorter-tailed, paler animal, with nearly 

 white instead of ' broccoli-brown ' under parts, and with regularly 

 graduated unicuspids. Its low-arched, narrow skull is strikingly 

 smaller than that of huxtoni (of which I have had topotypes for 

 comparison, and which Thomas believes to be synonymous with 

 macropygviaeus). The Japanese S. haivkcri, lately described by 

 Thomas, seems to be a more southern form of the same group, with 

 regularly graduated unicuspids, though its tooth-row is very much 

 shorter. It is possible that S. amiexus of Korea and S. centralis 

 from the Lake Baikal region also are related members, having no 

 very close connection with S. minutus (which in Japan is repre- 

 sented by the small S. m. gracillimus) . 



