14 BULLETIN" 114, UNITED STATES NATIONAL MUSEUM. 



of tlie eye, and beneath tlie newly added labial in the upper row. 

 That this is normally the case seems evident from a study of pyr- 

 rhomelaena and caUigaster (fig. 6), but that the extra scute is often 

 acquired behind the eye must also be admitted. When the latter is 

 the case the added scute may be due to the division of the seventh 

 or to the addition of one between the seventh and the eighth (fig. 5), 

 or to the division of the ninth, or to a shortening of the last and a 

 migration into the labial series of one of the gular scales. The few 

 specimens that possess eleven lower labials usually show that this 

 is due to the addition of one before the eye and one behind. It is 

 not improbable that when the change from nine to ten is due to the 

 division of the seventh that this is an accident in place of the third, 

 and that the next division would then be due to occur before the eye 

 instead of behind. We may therefore express the normal change in 

 the lower labials as follows : 



8-9-10-11 



7 4 8 



in which the upper line refers to the number of labials and the lower 

 to the number of the scute that is involved in the loss or gain, as the 

 case may be. In case of increase it might be more accurate to replace 

 the second row above with the figures 6, 3, and 7, to indicate which 

 scute is divided to increase the number, since it more often looks like 

 a case of division than like an interpolation. 



If we express the changes in the upper labials, as we have done for 

 the lower, we get the following: 



8-7-8 

 6 3 



This fits the scheme for the lower labials, and the two may be com- 

 bined thus: 



6-7-8-9-10-11 

 6 3 7 4 8 



We may express this in the following general rule, which seems to 

 hold throughout the genus, subject to the exceptions explained 

 above: In reduction — ^when the number of labials is odd, a scute is 

 lost behind the eye, and when the number is even, one is lost in front 

 of the eye; in addition — when the number of labials is odd, a scute 

 is added in front of the eye, and when the number is even, a scute 

 is added behind the eye. The two labials of each series that lie 

 directly beneath the eye are never concerned in any change. 



The actual method of change in the number of labials is of scant 

 importance in determining the relationships of the forms within the 



