NOTES ON MICHIGAN SNAKES. 173 



■wii'hes were distributed at a small reception and some thirty students 

 of both sexes ate them, not knowing what they were. All agreed there 

 was nothing disagreeable about them, while many of the young ladies 

 thought them very good, and some supposed they were chicken. The 

 meat is white, somewhat like that of frog's legs, but rather more tough 

 and stringy. 



^'(iriatioii. — The question of variation in snakes demands careful study 

 from living specimens, no preserving fluids serving to keep the colors 

 or pro})ortions with accuracy. Among our Michigan snakes there is no 

 one which will better repay investigation than the common milk snake 

 {Osceola doliata triangula). I have seen few specimens from this State, 

 but they are so entirely different from the Massachusetts specimens 

 of what is supposedly the same subspecies that there is little doubt in 

 my mind that thej- are distinct. — In my paper on water-snakes, 

 already referred to (American Naturalist, January, 1903), several points 

 in regard to the variability of Natriw are brought out, which are of some 

 interest, the most important being the apparently greater variability of 

 females. Investigations on these points are now being carried on as 

 extensively as living material permits. The recent warm weather brought 

 out the snakes in large numbers and 109 specimens have been examined. A 

 hasty survey of these statistics confirms my observations of last year 

 showing that females are externally distinguishable from males, that they 

 are much more variable, and that the lower jaw is much more variable 

 than the upper. — In the ''Ohio Naturalist" for January, 1902, is an inter- 

 esting account by Max Morse of variations observed by him in a brood 

 of thirty-three young water-snakes {Natrix fasciata sipedon), which were 

 taken from the body of the mother. I have recently examined a similar 

 brood of thirty specimens of the same species, and a comparison of our 

 results is interesting. Unfortunately my specimens are alcoholic, and I 

 have therefore made no measurements. On the other hand, Morse makes 

 no reference to sex (which can be easily determined even in these unborn 

 young) nor to the labial plates, nor to the urosteges, while his remarks 

 in regard to the number of scale rows indicate that he does not under- 

 stand that the number of longitudinal rows where most numerous 

 is what is meant by "number of scale rows." His statement in regard 

 to rows converging is true of practically all snakes. As a result of my 

 inability to make measurements, and his failure to record at least four 

 important points, our observations are comparable only in the number 

 of postoculars and gastrosteges and in the position of the umbilicus. Of 

 his thirtj'-three specimens, only eighteen had three postocular plates on 

 each side of the head, three others had three on the right side, one had 

 three on the left side, while eleven had only two on each side. Of my 

 thirty specimens, only one, a female, fails of the normal number, she 

 has two on the right side, three on the left. Here is certainly a "notable 

 difference between the Ohio and Michigan broods. In number of gastro- 

 steges, Morse gives the maximum as 150, the minimum as 142, the mean 

 as 144.8. My specimens show a maximum of 146, minimum of 139, mean 

 of 142.3. At first sight this seems like a notable difference, but I think 

 it probable that it is not real. In my figures, the anal plate and the 

 more or less triangular plate on the chin which heads the series of gas- 

 trosteges arc never included. There is reason to believe that Morse has 

 included them, and if so. then the Ohio and Michigan snakes are strik- 

 ingly alike in the number of g-Qstrosteges. Morse does not list any half 



