No. 437-] VARIATION IN LITHOBIUS FORFICATUS. 



cients of correlation of the pores of the right side were, according 

 to Table 8 : Anal, .227 ; 14th, .308 ; 13th, .298 and 12th, .205. 

 That is, the anal and [2th leg coxal pores are more independent 

 of the length of the animal than are those of the 13th and 14th 

 legs. In other words the correlation is less at the ends of a 

 linear series. I had expected the anal pores, the youngest onto- 

 genetically, to vary quite closely with the length of the animal, 

 the fewer pores on the shorter animal and vice-versa. This is 

 found to be true if a group of shorter animals, 6-15 mm. in 

 length be measured. The coefficient of correlation of the pores 

 of the right anal legs with the lengths in a group of 49 young 

 individuals is .88. 



4- In the first part of Table 8 are correlated the coxal pores 

 of the pairs of legs, right with left. Here again the correlation 

 is smaller at the ends of the series. The correlation is much 

 closer than it was with the length but the pores of the anal pair 

 have a coefficient of .575 and the 12th of .58 against .686 for 

 the 13th and .69 for the 14th. The order is the same as in the 

 length correlations except that now it is the anal pair of legs 

 which shows the least correlation whereas it was the 12th which 

 corresponded least closely to the length. 



I tried also one pair of diagonal cross correlations and some 

 serial correlations with very interesting results. The coefficient 

 of correlation of the coxal pores of the right anal legs and the 

 pores of the 12th legs is .43. That of the anal pores R. with 



