92 BULLETIN Gl, UNITED STATES NATIONAL MUSEUM. 



witliin this area of oak openings and open o;roves on tne hio;her p;roiind, 

 which Transeau has shown to be associated with a rainfall-evapora- 

 tion ratio of 80-100 per cent, that Butler's garter-snake is found (see 

 Transeau's map, 1905, 885). 



Specimens of hutlcri have been examined from the following locali- 

 ties: Ann Arbor, Chelsea, Ypsilanti, Washtenaw County, Brighton, 

 Livingston County, Olivet, Eaton Coinity, Pontiac, Oakland County, 

 Micliigan; Turkey Lake, Kosciusko County, Lake Maxinkuckee, 

 Marshall County, Waterloo, De Kalb County, Indiana; Sandusky, 

 Erie Comity, Columbus, Franklin County, "Miami River," Ohio; 

 Franklin, Venango County, Port Allegheny, AIcKean County, Penn- 

 sylvania. 



From these records it will be seen that Butler's garter-snake ranges 

 over most of Indiana, Oliio, southern Michigan, and western Penn- 

 sylvania, but its exact limits can nowhere be more than approxi- 

 mately fixed, for the locality records are extremely few. In Michigan 

 it has not been taken north of Pontiac, Oakland County, nor west of 

 Olivet, Eaton County, but it is a very common form in both of these 

 localities and doubtless extends farther north. South of tliis latitude 

 it is a common snake in eastern Michigan, while the various localities 

 in Indiana represent nearly the entire length of the State (Waterloo, 

 De Kalb County; Turkey liake, Kosciusko County; Lake Maxin- 

 chukee, Marshall County; and Richmond, Wayne County). Although 

 the specimens in the Academy of Natural Sciences of Philadelphia 

 labeled "Miami River" are also marked "Indiana," wliich would 

 indicate that they are from the extreme southeastern corner of the 

 State, the accuracy of the record can hardly be relied upon, and, as 

 Mr. Brown has remarked (Ruthven, 1904, 289), they are probably 

 from Ohio. 



At tlie present time we know of but two authentic locality records 

 for Oliio. We have already noted that one of the specimens which 

 Morse (1904, 134) records as sirtalis ohscura is a hutleri. An exami- 

 nation of his material shows that his specimens ot ohscura comprise 

 four garter-snakes from Columbus, Ohio, collected by KeUicot (not 

 as Morse gives it, Sandusky and Columbus, collected ])y liimself). 

 Two speciments are sirtalis, but one an undoubted hutleri. Although 

 Morse apparently did not find this snake at Sandusky, it is a very 

 common form there, and the Museum of the L^niversity of Michigan 

 has received a number ot specimens from Mr. E. L. Mosely, which 

 were collected in that locality. Columbus is the southernmost defi- 

 nite locality from which hutleri is known in Ohio, which is in nearly 

 the same latitude as the most southern Indiana record, that of the 

 type (Richmontl, Wayne County). How much farther southward it 

 occurs can only be conjectured, but as it is apparently less common 



