2 G. C. Embody. 



or the impracticability of propagating amphipods as food for fishes and 

 in case of the former to determine which species would lend themselves 

 most readily to such a procedure. During the progress of this study a 

 few supposedly new facts came to Ught which had no apparent rela- 

 tion with the original problem. It is believed however that these are 

 of sufficient importance to other lines of investigation to justify their 

 inclusion here. 



I desire to express my gratitude to Dr. J. G. Need ham of Cornell 

 University, under whose helpful guidance this investigation was carried 

 on. His constant sympathy with the work and his generous and kindly 

 criticism contributed largely to its successful completion. 



U. Relative sizes of the forms studied. 



The maximum sizes attained by adults of the four species in the 

 water examined are contrasted on plates I and II, and likewise in 

 table P). It will be noticed that G. limnaeus is larger than the other 

 three; G. fasciatus ranking second; E. gracilis, third and Hyalella 

 being the smallest. 



Table I. Lengths of largest adults collected. 



Miss Weckel states (1907 p. 33) that adults of E. gracilis vary in 

 length from 7 to 18 mm. At Ashland, Va., the writer found them as 

 long as 13 mm, and at Irvington, Ind., in pools not inhabited by fish 

 individuals as long as 15 mm were common. The writer is unable to 

 account for the small size attained by those in waters at Ithaca, un- 

 less it be from the fact (Table XI, p. 23) that these amphipods grow 

 much less rapidly than the other three forms and also that fishes and 

 other enemies are abundant enough to devour them before the maximum 

 size is reached. 



III. Distribution about Ithaca, New York. 



G. fasciatus so far as known is confined to Cayuga Lake and 

 its immediate tributaries. Its distribution in these waters varies with 

 the season. In March just after the disappearance of ice, it may be 



^) The lengths of the smallest sexually mature females are given in table XII, 

 pag. 25. 



