1406 



ginous outside. Orbigny states that the species is attached by its left valve, 

 which is, of course, an error. I find no persistent differences between the 

 recent and Pliocene shells. Those mentioned by Conrad, when he described 

 C. cornuta, are not constant in a large series. Fifty-five specimens of the fossil 

 forms show ribs ranging in number from seven to fourteen, averaging 9.6 ribs 

 each. Forty specimens of the recent shell have from five to sixteen ribs each, 

 and average 8.8. There is obviously nothing distinctive in the number of ribs, 

 and I have not found anything among the other characters. The specimens 

 of E. calif ornica in the collection show twenty to twenty-one ribs, and are pure 

 white with a yellowish tinge. 



The shell figured by Holmes under the name of C. arcinella is the young of 

 another species. 



Superfamily CARDITACEA. 

 FAMILY CARDITID^E. 



This family is of ancient origin and is represented in the Mesozoic by sev- 

 eral groups, but, as in other cases, only the Tertiary and recent forms will be 

 considered here. The family is related to the Crassatellitidce, Astartidce, and 

 Chamida, as a scrutiny of its paleontologic history, the anatomy, and the de- 

 velopment of recent forms conclusively show. 



The hinge has been compared to that of the Venerida, but in my opinion 

 the resemblance is slight and superficial. The Venerid hinge has never less 

 than three left cardinals, which show.no traces of torsion, while the Carditida 

 have never more than two, and the posterior one invariably long drawn out, 

 a feature characteristic of the family. The most fully developed hinge is met 

 with in such forms as Carditamera, which exhibits a hinge with the formula 

 L. oi.oioio.oi though the anterior and posterior right cardinals are in the adult shell 



R. io.ioioi.io 



almost obscured by the stem of the anterior right lateral and the nymph re- 

 spectively, though quite recognizable in the young shells. The cardinals are 

 almost invariably finely, transversely striated. In such forms as V enericardia 

 antiquata Linne the formula is reduced to L - - OIor - by degeneration. It is evident 



R. o.ioio.o 



when such changes take place in the growth of the individual, and when certain 

 of the teeth are or may be obsolete in the developed shell, that too much stress 



in classification should not be placed on these mutable features. In fact, the 



i 

 subdivisions of the principal genera must be based chiefly on form, the types 



of which, it must be acknowledged, are rather unexpectedly constant in the 

 faunas following that of the early Eocene. 



Throughout the family the tendency of the lunule is to be very small or 



