140 



C. B. Davenport. 



purposes of this paper these two form units may be considered 

 distinct species although some persons, considering facts like 

 those here presented, would regard the two as varieties, but this 

 difference in view does not affect our argument. When shells 

 from Long Island and the Gulf Coast at Tampa, Fla., are com- 

 pared they are found to differ in color, the lower valve of the 

 Gulf shells lacking the blue of the more northern shells and being 

 white or white and red. They differ quantitatively as follows: 



Taken together these three pairs of characters seem satis- 

 factorily to differentiate the two species. But when we study 

 shells from Cape Hatteras (Morehead, N. C.) we find here a 

 form unit in many respects constituting a link connecting the two 

 species. The color is quite intermediate. The number of rays 

 is 17.3, which is intermediate between some Long Island localities 

 and Tampa, but more like Long Island. The globosity of the 

 shell is much like that of Tampa, being 0.319 for shells of a length 

 of 59 mm. The range of globosity is such as largely to bridge the 

 gap between the means of the Tampa and Long Island lots. The 

 average excess of antero-posterior diameter is 2.5 mm. for 59 mm. 

 shells; thus intermediate between the 1.5 mm. of Tampa and the 

 6 mm. of Long Island. 



Finally, important evidence is afforded by a series of fossils 

 from the Pliocene or late Miocene of the Nansemond River 

 (James River system) at Jack's Bank near Suffolk, Va. These 

 fossils are Pectens^ closely related to P. irradians and known 



'Now deposited at the University of Chicago. 



