VARIATION AND MENDELIAN INHERITANCE 387 



For relative tail length, relative width of the tail stripe, and 

 number of vertebrae no corrections have been introduced, since 

 the first of these characters is correlated feebly and the last two 

 probably not all with body length. 



Another way of comparing the mean values of these various 

 characters in series of animals which differ in size has been 

 employed by me in earlier papers. This is to divide each series 

 into groups containing individuals of approximately the same 

 size, and to compare the means of the corresponding size groups 

 of different series. The mean difference between the two series 

 under comparison may readily be computed according to a 

 simple formula.^^ Such mean differences I have not yet calcu- 

 lated, however, for the present material. 



Figure 2 is based upon the corrected averages referred to 

 above, the mean of the figures for the two sexes (not weighted) 

 being employed for each race. The differences between these 

 various racial means are plotted to scale along the vertical lines. 

 The bone measurements have thus far been taken for only four 

 of the eight races, though the various bones have already been 

 prepared for measurement. 



It will be noted that for no two of the characters considered 

 is the arrangement exactly the same. The nearest approach to 

 agreement is found between the scales for tail and toot length, 

 the chief difference being that for the latter character there are 

 fewer distinguishable grades. The order for tail stripe is the 

 same as that for tail length with this important exception, that 

 Berkeley has been transferred from near the bottom of the scale 

 to a point above the middle. 



Skull length in the four races for which figures are available 

 follows nearly the same order as tail and foot length. 



In respect to the number of caudal vertebrae. Eureka stands 

 at the top of the scale, as was true for all of the four characters 

 just considered. The differences among the other four sta- 

 tions are, however, of doubtful significance in the case of the 

 vertebrae. 



1' Journal of Experimental Zoology, April, 1915, p. 346. 



