22 



THE AMERICAN NATURALIST. [Vol. XXXVII. 



In the Olivet College museum there is a specimen of erythro- 

 gaster, a female, taken at Olivet in the spring of 1889, which is 

 the largest water snake I have yet seen. The tail is broken, 

 but careful calculation for the lost portion shows that the speci- 

 men was certainly over 1300 mm. in length. The eye is 6.5 mm. 

 in diameter, 22 per cent of the head length; there are 154 

 gastrosteges and 2 5 rows of scales. There are no dorsal mark- 

 ings of any kind, but the ventral surface is mottled with a 

 great deal of slate color along the sides, especially near the 

 vent, on the posterior gastrosteges and the urosteges. The 

 coloration is therefore No. 8 of Table VI. 



We have here, then, three (or possibly four) snakes which 

 seem to be what might be considered connecting links between 

 erythrogaster and sipedon, or some other form of fasciata. Are 

 they really such? Two facts must be noted : first, all are old 

 specimens, the most recent having been taken thirteen years 

 ago, and that one is most nearly a typical erythrogaster; second, 

 all are females, the variable sex, and are aberrant in number of 

 scale rows, urosteges, or gastrosteges. They are not, therefore, 

 actually intermediate forms, but individuals which have varied 

 from the normal in color as well as in some other particular. 

 The smallest of my specimens of ejythrogaster, a male 760 mm. 

 long, was kept in captivity for six weeks, at the end of which 

 time he shed his skin. Although when captured his coloration 

 was perfectly normal, without a trace of markings, his new 

 suit showed along the sides faint indications of lighter, vertical 

 bands, visible only in just the right light. Might this not indi- 

 cate the ancestry, as the spots on the breast of a young robin 

 indicate its ancestry, without making the individual in any sense 

 a connecting link ? 



All of the evidence so far collected seems to me to show 

 that we have in Natrix erythrogaster a well-defined species of 

 water snake, probably derived from some form of fasciata, 

 though probably not sipedon. Possibly the separation has been 

 completed during the nineteenth century and the specimens in 

 the National Museum, referred to above, are some of the last 

 connecting links, though I am inclined to regard them merely 

 as unusually aberrant females. At any rate, what we need 



