206 PROCEEDINGS : BOSTON SOCIETY NATURAL HISTORY. 
and I have a skin from New Berlin, Florida, having an irregular 
black streak down the middle of the back. 
On Cumberland Island Mr. Brown took eighteen gray squirrels. 
These specimens are peculiar, being different from those taken any 
where else along the Georgia coast. In many respects they appear 
more nearly like /S. extimus. The tail is shorter than that of S. 
carolinensis typicus , the foot is smaller, and the hand is considerably 
smaller, the markings are very fine, and there is a pale, yellowish 
mealy cast to the whole upper parts. In all these characters they 
approach very nearly to S. carolinensis extimus, but differ from 
that form in being slightly larger, not so gray, and in having much 
less white in the tail. On the other hand, specimens from south of 
Cumberland Island on the Florida coast are true 8. carolinensis , 
and the Cumberland Island form, though it appears to tend towards 
S. extimus , is completely cut off from contact with that form by true 
S. carolinensis. Perhaps there is something in the conditions of 
life on Cumberland Island, differing from that of the main land of 
Georgia and northern Florida, and approaching that of southern, 
tropical Florida. More probably, however, the Cumberland Island 
form is an incipient insular species, and its likeness to S. extimus 
purely fortuitous. 
I have an enormous series, including specimens from Hursman’s 
Lake, St. Catherine’s Island, Harris Neck, Montgomery, Barrington,, 
St. Mary’s, and Cumberland Island, Georgia, and Rose Bluff (St. 
Mary’s River), New Berlin, Gainesville, and Citrus County, Florida. 
Sciurus carolinensis extimus Bangs. 
Sciurus carolinensis extimus Bangs, Proc. Biol. soc. Wash., 1896> 
vol. 10, p. 158. 
Type locality. Miami, Dade County, Florida. 
The everglade gray squirrel occupies southern, tropical Florida 
and passes into true S. carolinensis about half way up the penin¬ 
sula. It is a strongly marked form and very much smaller 1 and 
grayer, with much more white in the tail, than S. carolinensis 
typicus. 
I have specimens from Miami, Oak Lodge, and Eau Gallie, Florida. 
U typographical error crept into my average measurements for this form in “A 
review of the squirrels of eastern North America,’’ p. 158, where the measurement for 
hind foot should read 57 m. instead of “ 47.” 
