Dosinia discus (Reeve). 581 

 ^ — = K, a constant, 



Li 



that is, the shell of Dosinia grows in such a manner that the sum 

 of its width and depth divided by its length (as these dimensions 

 have been defined above) is, approximately, a constant. Of course, 

 this expression is only valid after the asymmetry of the valves (i. e., 

 the forward pointing of the umbo) has been started, and would 

 probably not be true for very young shells. The value of K 

 determined from the curves varies from 1,24 to 1,28 in passing 

 from L = 2 to L ^ 7, because the curves do not pass exactly 

 through the origin (Table I). 



Table I. 



Values of K for increasing values of L. 



L K 



2 1,28 



3 1,26 



4 1,25 



5 1,24 



6 1,24 



7 1,27 



The equations which Pearson, Peael, and Hatai have found 

 most useful in the graduation of growth data (Peael, 1909; Hatai, 

 1911, 1913) are logarithmic, and involve several constants. It is 

 clear that they do not apply to the present case; the relative growth 

 of the shell parts in Dosinia parallels rather that of the body parts 

 in the teleost, Cynoscion regalis, where the depth, width, head length, 

 etc., are directly proportional to the total length of the animal.^) 



Growth with respect to time has not been considered in the 

 preceding discussion. According to a quite generally accepted idea 

 — that the dark bands on molluscan shells represent periods of 

 arrested growth, and that the more prominent ones are year bands 

 (Belding, 1911; Rossbach, 1912) — it should be possible to obtain 

 some idea of the age of the shells. On attempting to do this, how- 

 ever, it was found that in Dosinia the dark rings are of very variable 

 number, width and opacit}^, and that no correlation could be made 

 out between shell size and band number. It is possible either that 



1) Ckoziee, W. J., and S. Hecht (1912), in: Bull. U. S. Bureau 

 of Fisheries. (Not yet published.) 



