178 MICHIGAN ACADEMY OF SCIENCE. 



tive number of scale-rows. In the water-snake the normal number of 

 scale-rows is twenty-three and only eighth-two per cent, of the individ- 

 uals examined were normal in this respect, man}' of them having twenty- 

 four and twenty-five rows. In the garter-snake the normal number is 

 nineteen, and of the 148 examined not a single individual showed any 

 variation. 



NO CORRELATION BETWEEN COLOR AND OTHER VARIATIONS. 



When the work upon the age and sex variations had come to an end 

 an attempt was made to find some differences in relative measurements 

 or number of ventral plates, which should correspond to the color differ- 

 ences but it was unsuccessful. The majority of the individuals taken 

 were either Thamnopliis sirtalis sirtoUs or Tliamnopliis sirtalis parietalis, 

 distinguished by the presence of red blotches beneath the lateral stripes 

 and the absence of gastrostegal spots in the latter sub-species. The 

 outline which was followed in the color classification was one which the 

 writer arranged to meet the needs of the case and was as follows : 



Red in lateral stripe 



i!so constant gastr. spots, dorsal stripe not prominent. 

 Constant gastrostegal spots. 

 Dorsal spots evident 

 Back hlacl-isJh 

 Back greenish. 

 Back hrownish. 

 ^0 red in lateral stripes 



Dorsal and lateral stripes yellow 

 Dorsal and lateral stripes green , 

 Dorsal and lateral stripes very dull 



flaking the distinguishing point the presence or absence of red, the 

 results of the tabulation were, briefly, as follows : 



Individuals with red, average relative length of head .0367, without red 

 .0360. 



Individuals with red, average relative length of tail ,222, without red 

 .218, 



Individuals with red, average number of urosteges 68, without red 69. 



Individuals with red, average number of gastrosteges 1.52, without red 

 153. 



Individuals with red, eighty-three per cent, as to labials, without red 

 seventy-nine per cent, regular. 



From these facts it will be seen that there is no difference which might 

 not vary one way or the other with further examination and that the 

 color variation is the only one which warrants any division into sub- 

 species. There are undoubtedly^ marked differences between many of 

 the garter-snakes in this respect, but it would simplify the classifica- 

 tion if this were regarded as a matter of individual variation rather 

 than a sign of differentiation into sub-species. In the first place the 

 age largely affects the coloration. Those who are acquainted with the 

 life history of llie blue-racer will know Ihat the young are very different 

 from the adults, in having a checkered marking on the back, much like 

 that of the common water-snake, and do not acquire the peculiar uniform 



