Jan., 1902.] Variations in the Water-Snake. 183. 



VARIATIONS IN THE WATER-SNAKE. 



Max Morse. 



The following is the result of a study of the variations in the 

 members of a litter of 2)3 young of the water-snake — jYafr/x 

 fasciata fasciata (X.) The mother was captured at the Lake 

 Shore Railroad bridge No. 13, Sandusky, Ohio, in August, 1901. 

 The head of the mother was so mutilated on being killed that it 

 was thrown away, and hence no comparative study of parent and 

 offspring could be made. The young were very nearly ready to 

 hatch and probabh' w-ould have been born in a few days. With 

 the exception of some of the gastrosteges, all external characters 

 were as in the adult condition. 



Owing to the want of time, only the external characters are 

 considered in this paper. All measurements were made on the 

 fresh specimens, i. c, before the snakes were placed in formalin. 

 Care was taken not to stretch the specimens more than was 

 necessary to straighten them. The counting of scales, etc., was 

 done under a lens and each count was verified twice. 



The results of the study are given in the table on page 186. 

 The following have been considered : 



1. The variation in length from snout to tip of tail. 



2. The variation in length from snout to the tip of the anal 

 plate. 



3. The variation in length from anal plate to the tip of the tail . 



4. The variation in the number of gastrosteges, /. e., the 

 ventral plates. 



5. The variation in the number of gastrosteges from their 

 beginning in the gular region to the umbilicus. 



6. The variation in the number of postoculars. 



The maximum and minimum number, the difference between 

 them, the mode and the mean are given in each case. Obviou.'-h^, 

 as the variates in columns 4, 5 and 6 are integral, the mean in 

 these cases would sometimes be only approximate, i. e., fractional. 

 Under these conditions the probable error of the mean was not 

 calculated, but only approximated, b}' adding an integer to the 

 mean in case the fraction was greater than one-half.* 



Comparing columns 2 and 3, it will be Sccn that there is greater 

 variation in the body region than in the tail. This is contrary to 

 what maj' have been expected, as in certain organs, such as the 

 vertebrae, greater variation occurs in the tail than in the body. 



In many cases the gastrosteges were found to be bifid, as in 

 the normal anal-plate. The number of these that were found, 

 together with their distribution, are given in cokunn 7, where 

 the numbers represent the number of the gastrostege, counting 



♦Davenport, C. B., " Statistical Methods." John Wiley & Sons, N. Y., 1899. 



