.DAVENPORT. — EVOLUTION OF PECTEN. 127 



IV. Results. 



1. Change of Proportions with Age. 



In order to make use of the dimensions of shells of different sizes, it 

 is important that we should know how and to what extent the shells 

 change in proportions as they grow larger. The change in the propor- 

 tions of the two chief axes perpendicular to the transverse interests us 

 most at this time, and are given in Table I for all localities. 



TABLE I. 



For each Class of Dorso-ventral Diameters are given the Frequency and 

 the Ratio of Anteroposterior Diameter to Dorso-ventral Diameter 

 for the Right Valves of all Sizes from the Three Localities. 



Firth of Forth. 

 /. Ratio. 



613 



This table shows, first, that the Eddystone shells are the smallest, and 

 the Firth of Forth shells the largest, the shells from the Irish Sea being 

 intermediate in size ; secondly, that there is an increase in the ratio from 

 near or even below unity iu the smaller shells to a point much above 

 unity, 1.06-1.10, in the largest shells; and thirdly, that the shells of a 

 given dorso-ventral diameter are longest at Eddystone and roundest at 

 the Firth of Forth. Thus for shells having a length of 67 mm. the 

 ratios are: 1.067, 1.061, and 1.039 in the three localities. 



A similar change of proportion with size I have found in Pecten irra- 

 diuns from the east coast of the United States and in Pecten rentricosus 

 from the coast of southern California. Assuming that the small shells 

 retain embryonic characters more than the large ones do, then in accord- 

 ance with the biogenetic law we would be led to conclude that the ances- 

 tral pectens had a relatively greater dorso-ventral diameter, and that 



