1909] The 1907 Alexander Alaska Expedition. 263 
abundantly everywhere except on Admiralty where only a 
single specimen was collected. It was the only shrew taken on 
Baranof and Chichagof islands. 
This series is fairly uniform in coloration. The size varia- 
tion is considerable, some of the larger specimens approaching 
glacialis in external size. The skull variation due to age is 
enormous. Young adults with unworn teeth have large skulls 
with high, inflated brain cases, which surpass greatly im all 
dimensions those of old adults with worn teeth. Unfortunately 
the topotype series of streatori consists chiefly of old adults 
with their dental characters worn away largely. The old skulls 
in the present series agree well with these in size and dental 
characters. There is, however, a color difference between these 
two series, one which practically parallels that found between 
glacialis and longicauda. Compared with typical streatort in 
the same pelage our series is more grayish brown above with 
very little yellowish or brownish wash to the grayish under- 
parts. Wrangell and Revillagigedo islands are represented in 
the Biological Survey collection by series taken in September 
which are scarcely distinguishable in coloration from Yakutat 
material. 
Sorex obscurus glacialis Merriam. Glacier Shrew. 
A good series is in the collection from Admiralty Island. 
These were secured at Mole and Windfall harbors and at Hawk 
Inlet, and Hasselborg Lake. They are also known to occur on 
Chichagof Island. 
A series of twenty-three topotypes was secured at Glacier 
Bay. Glacialis was based on two specimens, a type and topo- 
type from Point Gustavus, Glacier Bay. The large series now 
available seems to confirm the color differences. They differ 
from both alascensis and longicauda in their more grayish 
brown upperparts and the less extensive brownish wash on the 
underparts. There are, however, no very evident skull charae- 
ters or differences in body measurements between these series. 
In coloration they approach very closely specimens of obscurus 
from Bennett, Alaska, to which form they are apparently most 
closely allied. 
