10 



A most interesting uniformity is revealed. It is only in a few indi- 

 viduals of a single form {calVtgaster) that 27 rows are ever found, and 

 only in a few individuals of two degenerate forms {elapsoides and 

 virginiana) tliat 15 rows occur. Just as before the presence of 15 

 and of 27 rows was discovered, it was expected, from the table as 

 then worked out, that the change from 17 to 15 rows would involve 

 the loss of the fourth, and the change from 27 to 25 would involve 

 the loss of the seventh, so now we may expect if 29 rows should ever 

 be found, it would be by addition of a seventh row, and that a drop 

 to 13 would involve the fourth. The latter might not hold, however, 

 for we kno^v that some snakes have the same number of rows through- 

 out the length of their body, and that such generally possess a small 

 number of rows, as 19, 17, or 15. It is not impossible that these 

 forms are the end results of degenerative evolution, involving diminu- 

 tion in number of scale rows. This is suggested by the case of 

 elapsoides, the most diminutive and degenerate form in the genus. 

 This form has clearly been evolved from the typical forms by degen- 

 erative evolution. In southern Florida it reaches its greatest reduc- 

 tion and it is here only, out of the whole genus, that individuals have 

 been found which possess the same number of scale rows throughout 

 the length of the body. Several individuals from here have been 

 noted with 17 rows of scales throughout the body. It is as if extreme 

 reduction leads to a minimum number of rows, uniform throughout 

 the length of the body. That an increase from 27 to 29 rows would 

 involve the addition of a seventh row is indicated by the genus 

 Elaphe. Examination of a few individuals of E. laeta, E. guttata, and 

 E. obsoleta gives the following formula for change in number of scale 

 rows in this genus: 



29-27-25-23-21-19 

 7 7 6 6 5 



Thjs, it wdll be noticed, is the same rule as derived from Lampi^opeltis. 

 The rule as derived by Ruthven for the genus ThamnopMs is as fol- 

 lows : 



23-21-19-17-15 

 5 5 4 4 



This formula is the same in kind as for Lampropeltis and ElapTie, but 

 the lower line of figures has been moved one place to the left. The 

 rule for TliamnopJiis appears, quite according to expectation, to hold 

 likewise for Natrix, but as to whether the similarity in formulae 

 between ElapJie and Lampropeltis indicates communit}^ of descent, 

 we can not say. Further investigation may shov/ what value to 

 place upon similarity in method of change in number of scale rows, 

 in establishing relationships between genera. 



The value of a knowledge of variations in scale rows in a study of 

 relationships within the genus lies not in which row^ is dropped, since 



