DAVENPORT. — EVOLUTION OP PECTEN. 135 



Firth of Forth. Mean, 17.673 ± .027 rays ; stand, dev., 1.117 i .019 rays ; 

 coef. of var., 6.32-^.11%. 



M2 = 1.247553 /«3 = 0.021655 M4 = 4.422348 



/3i = 0.000241 /3.2 = 2.841431 Crit. funct. (k^) = - 0.000060 



Normal type. (Skewness, Type I., 0.0069.) 



Comparison of Empirical and Normal Polygons. 



In all cases the relation between the observed and the calculated 

 frequencies is fairly close. 



4. Individual and Specijic Variability. 



In TTie Origin of Species Darwin refers to the " notorious " fact that 

 specific characters are more variable than generic ones. The evidence 

 he presents is, naturally enough, wholly qualitative. Brewster (1897) 

 first tested this law quantitatively, making use of various data derived 

 from mammals, and found it to be confirmed. Heincke (1898) finds 

 the law to hold good for races of herring. He says : " Die Rassen des 

 Heriugs unterscheiden sich in sehr vielen und im allgemeinen in denselben 

 Eigenschaften von einander, in denen die Species der Gattung Clupea 

 von einander verschieden sind. Nur sind die Unterschiede der Rassen 

 meistens, aber nicht immer, kleiner als die der Species " (quotation from 

 Duncker, 1899, p. 366). The genus Pecten affords a good opportunity 

 to test this, especially in the case of the relative variability of the ante- 

 rior and the posterior " ear," or half, of the hinge line. For in some 

 species of Pecten the posterior ear is nearly equal to the anterior ear. In 

 other species it is very much less. The anterior, or notched, ear is the 

 more constant ; it may be regarded as a generic character. The posterior 

 ear may be regarded as a specific character ; is it more variable than the 

 anterior ear ? 



a. Geographical Variation of the Ears. — First, a brief consideration 

 of the differences in hinge proportions in the different localities. The 

 different species of Pecten differ greatly in the length of the hinge in 

 relation to that of the main part of the body. It has seemed worth 

 while to calculate the average ratio of length of hinge joint to length 

 of shell for each locality. 



