1890.] 



occurring in certain Decapod Crustacea. 



451 



Table III.— Ordinates of the Curves of Deviation of Carapace 

 Lengths, each in Terms of its own Q. 



, . , observed ordinate — M 

 Each entry = — 



Grade. 



Jr lymotrtxi. 

 400 specimens. 



CU U.l'11 JJUI v. 



300 specimens. 



fin PPTTI pCQ 



300 specimens. 



^Normal curve. 



5° 



+ 2-42 



+ 2-86 



+ 3-34 



+ 2*44 



10 



+ 1-86 



+ 2-11 



+ 2-29 



^-1-90 



20 



+ 1-22 



+ 1-29 



+ 1-35 



+ 1-25 



25 



+ 1-00 



+ 0-97 



+ 1-02 



+ 1-00 



30 



+ 0-79 



+ 0-70 



+ 0-76 



+ 0-78 



40 



+ 0-39 



+ 0-34 



+ 35 



+ 0-38 



50 



o-oo 



o-oo 



0-00 



o-oo 



60 



-0 32 



-0-37 



-0-35 



-0-38 



70 



-0-71 



-0-80 



-0-74 



-0-78 



75 



-1-00 



-1-03 



-0-99 



-1-00 



80 



-1-28 



-1-27 



-1-28 



-1-25 



90 



-2-10 



-2-05 



-1-97 



-1-90 



95 



-2-63 



-2-68 



-2-41 



-2 44 



The table shows that in all the races the coefficients of Q agree 

 fairly well with those indicated by the normal curve. When these 

 coefficients and the values of M and Q in each case are known, it is 

 evident therefore that the whole curve is known. 



The values of M and Q are as follows : — 



Plymouth M = 250 05 ; Q = 4-53 



Southport M = 248-50 ; Q = 3-17 



Sheerness M = 247'51 ; Q = 3'05 



It thus appears that not only does the average size of the carapace 

 differ in different local varieties, but the range of deviation from that 

 average differs also. Nevertheless, the frequency with which the 

 observed deviations from the average occur is in all the three 

 observed cases expressed by a curve of error. 



Since the variations observed in adult individuals depend not only 

 on the variability of the individuals themselves (which is possibly 

 nearly alike in all races), but also upon the selective action of the 

 surrounding conditions — an action which must vary in intensity in 

 different places — the result here obtained is precisely that which 

 might be anticipated, and it is precisely that predicted by 

 Mr. Galton. 



The same features are presented by the curves derived from the 

 remaining sets of measurements. The following tables give the data 

 for constructing curves of deviation of each organ. 



