438 U. S. BUREAU OP FISHERIES 



straight diet of beef liver. The mortality, however, was low and 

 practically the same as in the lot fed beef liver only. The results 

 with yearling rainbow tront were practically the same as with 

 fingerlings. 



In order to compare the value of seal meal and salmon-egg meal in 

 the diet of trout, two lots of rainbow fingerlings, at the Leetown, 

 W. Va., station were kept on experimental diets from September 23 

 to December 18. One lot was fed a ration composed of 60 percent 

 beef liver and 40 percent seal meal while for the other lot salmon- 

 egg meal was substituted for seal meal. At the end of the experi- 

 ment the lot fed the seal-meal combination had gained 380.78 percent 

 in weight while the lot fed the salmon egg meal combination had 

 gained only 274.09 percent. The mortality was 10.2 percent and 

 16 percent respectively. 



Selective hreeding. — Breeding experiments designed to develop 

 superior strains of trout were continued along the same lines as in 

 previous years. Eggs from 48 selected pairs of brook trout were 

 incubated at the Pittsford, Vt., station and as usual the best lots 

 of fingerlings will be segregated until they mature. An exception- 

 ally high percentage of these eggs hatched and, since they were taken, 

 from a stock of fish producing highly fertile eggs, it is a reasonable 

 tentative assumption that this high fertility is a result of selection. 



In the breeding experiments at the Leetown, W. Va., station eggs 

 were taken from 38 pairs of selected brook trout, a part of which 

 were select stock of the second generation. In an effort to obtain a 

 strain of brook trout which will produce eggs better adapted to 

 conditions at Leetow^n, stock is being selected at present on the basis 

 of high fertility rather than of rapid growth, body symmetry or 

 other characters. The present standard requirement for a select 

 stock is a minimum of 70 percent reaching the eyed stage, an excep- 

 tionally high figure for brook-trout eggs at Leetown, the average 

 being approximately 43 percent. 



Breeding experiments have also been carried on with rainbow 

 trout and with brown trout. 



Nutrition studies. — Nutrition studies have been continued at the 

 Cortland, N. Y., hatchery by Dr. C. M. I^IcCay and A. V. Tunison. 

 Three lines of research have been followed during the year: (1) 

 Continuation of the study on growth of four species of trout fed 

 similar diets, and maintained under similar conditions; (2) the de- 

 termination of the amount of feed that should be fed per unit of 

 weight to attain the most economical production; and (3) biochem- 

 ical studies to determine the optimum amount of phosphorus and 

 calcium needed by a growing trout. The four species used were: 

 Lake trout {Cristivomer namaycush) ., browm trout {Sahno fario), 

 rainbow trout {Salnio irideus), and brook trout (Salvelinus 

 fontinalis) . 



The experiment to determine the relative ability of the species to 

 convert feedstuffs into body tissues was continued through the 128th 

 week. All four species were maintained under similar conditions 

 and fed the same diet, consisting of two parts fresh beef liver and 

 one part dry skim milk. The numbers were reduced from time to 

 time to eliminate the factor of overcrowding. It was found that 

 lake trout was the most efficient of the four species in the conversion 



