370 U-S. BUREAU OF FISHERIES 



skim milk, 2 : 1, was used to determine if two daily feedings of such a 

 diet were satisfactory. Although the growth curve showed good 

 results, it was below the maximum. A diet of fresh beef liver and 

 dry skim milk, 1 : 2, was fed 6 times daily as a basis of comparison 

 for the growth rate obtained from feeding fresh beef liver and dry 

 whole milk, 1 : 2. This experiment with whole milk was included to 

 see if butter fat, which is present in the whole milk but very low in 

 the skim milk, might not be advantageous in feeding young trout. 

 This butter fat provides a diet richer in calories as well as the fat 

 soluble vitamin A. In this combination, however, liver seems to 

 provide adequate supplies of this vitamin, since the growth curves in 

 the two experiments were identical. 



At the same time another lot of trout was fed dry skim milk and 

 raw egg in the ratio of 1 : 1 to test eggs as a source of the growth 

 factor H, which is destro3'ed in the usual drying of fresh meats. No 

 appreciable difference in growth rates was observed. Another diet 

 consisted of dry skim milk, cottonseed meal, and white-fish meal in 

 the ratios 2:1:1, fed regularly with a change to the diet of dry 

 skim milk and fresh beef liver once a week. A slow growth rate 

 from this diet was observed for the first 3 months, but aft^^r that 

 the curves ran parallel with those obtained from better diets. The 

 spray process of dry skim milk 92 percent, and vacuum dried beef 

 liver 8 percent, was combined for the diet of another experimental 

 lot. The liver was dried under 80° C. in an atmosphere of nitrogen. 

 It was then stored under carbon dioxide until ready for use, in order 

 to determine whether factor H found in fresh liver could be pre- 

 served in this way. For the first 20 weeks these fry grew very rap- 

 idly, but at the time thej^ were approaching a mean weight of 2 grams 

 their growth became very slow and they started dying rapidly. 

 These experiments show that fry can pass through the early stages 

 with excellent growth upon a dry diet, but even with the careful 

 treatment it received the liver seemed to have lost much of its 

 factor H. 



In experiments started during the previous year growth curves 

 were constructed for trout fed upon mixtures of dry buttermilk, 

 skim milk, cottonseed meal, and peanut meal. Data were then avail- 

 able onh' for 16 weeks, but these groups of trout were continued for 

 8 additional weeks or a total of 24 weeks. Five hundred brook trout 

 were used in each group at the beginning. At the end of 16 weeks 

 this number was reduced to 300 in order to prevent crowding in 

 the troughs. As far as these growth curves indicate these practical 

 diets are equal in value. Either peanut meal or cottonseed meal are 

 suitable trout feeds if combined with a binder such as dry skim 

 milk. The percent utilization of such feeds, however, cannot be 

 determined until balance experiments can be run. 



During the past year standard fish hatchery troughs were used 

 for running another series of studies upon commercial feedstuff s 

 that had already been tested in previous years. Experimental diets 

 were composed of varying proportions of dried skim milk, cotton- 

 seed meal, white-fish meal, salmon-egg meal, and raw beef liver. 

 Each experimental group was started with 1,250 fingerling brook 

 trout of the Pittsford (Vt.) strain and continued for 28 weeks. 



