VARIATIONS OF GARTKR-SN AKES. 75 



ing sometimes as many as 30 or 40." I have examined pregnant 

 females in July, August, and September in western Iowa that con- 

 tained from 17 to 25 young, and have kept females in captivity that 

 gave birth to young on August 31 and September 7, 8, 29, and 30 

 (1907). By the size of some of the embryos examined I believe 

 that broods may appear as early as the latter part of July. 



The period of gestation is uncertain. Coues states that "individ- 

 uals were taken in coitu in September and part of October. So as it is 

 unlikely tliat young are born after this date, tliis observation might be 

 taken to indicate a period of gestation protracted for tlie greater part 

 of the year." Tliis observation is difficult to explain. It is hardly 

 possible that more than one brood is raised each year, and those that 

 appear in October are probably belated first broods. Likewise it is 

 highly improbable that the period of gestation is protracted over the 

 winter, since such is not the case in the other species of the genus in 

 which coition has been observed. The probabilities are that coition 

 takes place in the spring, and that Coues was mistaken in liis obser- 

 vation or was viewing abnormal cases. 



Coues (1878, 278-279) writes that i^adix is abundant along the 

 northern boundary of the United States, and Branson (1904, 364) 

 adds that "It has a wider distribution and occurs in greater num- 

 bers than any other Kansas garter-snake." I have already ob- 

 served that it is very common in northwestern Iowa. At times it 

 occurs here in such numbers as to become a veritable nuisance. 

 Particularly was this true in the summer of 1892, when it became 

 extraordinarily abundant. But, while it occurred in numbers about 

 the barns and outbuildings on higher ground, it was most noticeably 

 abundant in the sloughs and about the lake shores, wliicli fairly 

 teemed with them. Hundreds of individuals could be observed lying 

 on the rocks along the shores of these lakes or swimming freely in 

 the water. Tliis remarkable development was doubtless due to the 

 advent of unusually favorable conditions, and it may be significant 

 that this was an exceptionally wet year. After 1892 the number 

 apparently fell off rather suddenly, for they were not observed to be 

 noticeably abundant until 1896, when, according to a number of 

 residents, they again became very numerous, although not as much 

 so as in 1892. 



The abundance of radix over most of the prairie region and the 

 fact that it is here of general distribution and not confined to a par- 

 ticular habitat, indicates, it seems to me, that, unless it changes its 

 habits, we may expect that toward the limits of the prairie it will 

 become more closely confined to the conditions to wliich it has 

 become adjusted in the heart of its range, and will finally be limited 

 in a general way by tlie boundary of the prairie-plains conditions. 



