24 BULLETIN 01, UNITED STATES NATIONAL MUSEUM. 



for example, that the forms known as elegans and ordinoides intergrad^ 

 in eastern Washington and Oregon, elegans extending to the east of 

 this region and ordinoides westward to the coast. In elegans the labi- 

 als are normally 8/10, but in ordinoides they are generally less, so that 

 there is evidently a decrease in the number of these plates in the region 

 where these forms intergrade. This is shown when specimens from 

 the eastern part of the range of ordinoides are examined. Thus, on 

 the eastern shore of Puget Sound the upper labials in this form 

 are 7 or 8 (rarely 6); the infralabials 8, 9, or 10, the average for the 

 upper labials in a large series being 7.2, for the infralabials 9. 

 Toward the west coast, however, the number decreases very notice- 

 ably, the occurrence of 8 upper and 10 lower labials becomes rarer, 

 the formula tending to become 7/8. This not only indicates a differ- 

 ejice between two forms in the number of plates, but also a geographic 

 variation in this trait within the form ordinoides. As w^ill be shown 

 later, this occurs in several other forms, and there seems to be good 

 reason, therefore, for concluding that whatever the factors may he tJmt 

 injluence the number of labial 'plates, the variations are geographic and 

 have been the basis for the racial differences that now exist. This will 

 come out much more clearly in the consideration of the various 

 forms. 



In attempting to discover the factors involved in the variations 

 in the number of labials it is necessary to inquire more closely into 

 the nature of the latter. In the first place it must be noted that the 

 number of upper and lower labials are correlated. For example, in 

 a series of specimens of ordinoides from Puget Sound, Washington, 

 all but two out of fifteen individuals with 7 supralabials have 9 

 inferior labials or less, while out of the eight with 8 upper labials 

 six have 10 inferior on both sides and one 10 on one side and 9 on 

 the other. In most cases, however, the range of individual variation 

 is too small to show this correlation, and it can only be noted in a 

 o-eneral way. Thus, a review of the different forms shows that in 

 those with 8 superior labials the inferior labials are normally 9, 10, 

 or 11; in those forms, in which the number of supralabials is 7, the 

 infralabials are 8, 9, or 10, while in those with 6 in the upper series 

 there are usually 8 or 9 in the lower. Thus, although not exact, a 

 correlation between the number of superior and inferior plates very 

 evidently exists, and it can be said that generally there are normally two 

 or three more scutes in the lower series of labials than in the upper, but 

 rarely less or more than this number. 



This- correlation might well be expected from the fact that these 

 two series of scutes border the mouth exactly above and below on 

 either side, and ks they thus have the same relative length in each 

 snake, any difference in the length of one series must evidently be 

 correlated with an increase or decrease in the length of the other, and 



