124 U. S. BUREAU OF FISHERIES 



rainbow trout {/Salmo shasta)^ the steelhead trout {Salmo gaird- 

 iierl), and the hike trout {Christ ovom^&r namoycush)^ have been 

 utilized to some extent. It is planned to conduct more extensive 

 feeding experiments with rainbow and brown trout as soon as facil- 

 ities are available. 



In any consideration of the relative value of different trout foods 

 it is important to bear in mind that rapidity of growth is not the 

 onl}^ factor to be considered. Too often in the past this has been 

 virtually the only criterion used in evaluating a trout food. Of 

 course, the reason for this is perfectly obvious. We have no other 

 means of comparing definitely the results obtained with different 

 rati(ms. We have no '"' yardstick " for making precise measurements 

 of vigor and vitality in fish, yet anyone who has had experience in 

 handling trout knows that the}'^ often show marked differences in 

 this respect. Mortality records are very valuable as an aid in de- 

 termining the relative efficiency of different rations and have been 

 carefully kept in all of our experiments. The interpretation of these 

 records has frequently been complicated, however, by the spread of 

 specific infections among the fish in the experimental lots. The 

 form, color, and behavior of the individual fish are among the most 

 important criteria to be observed in forming an opinion as to the 

 relative hardiness and vigor of different lots. There can be little 

 doubt that trout which closely approximate the wild fish in these 

 respects are more desirable for stocking purposes than the soft, fat, 

 i:)ot-bellied fish not infrequently seen at our hatcheries. 



Available trout foods can be divided into three groups : The first 

 group includes fresh fish and fresh meats, such as horse meat and 

 the liver, lungs, and spleen of cattle, sheep, and hogs; the second 

 embraces the various dried products of animal origin, such as meat 

 meals, fish meals, and dried milks; while in the third group we have 

 the vegetable products including wheat middlings, low grades of 

 flour, shorts, soybean meal, and Mexican pinto beans. Although 

 these products have been used for years by fish culturists, in many 

 cases there is, as yet, no general agreement regarding their relative 

 value as trout foods. 



The experiments which form the basis of this paper were begun at 

 the Manchester (Iowa) station in the spring of 1923. They were 

 continued during the summer of 1924 at White Sulphur Springs 

 (W. Va.) station. In the spring of 1925 it was decided to convert a 

 small liatchery at Pittsford, Vt., into an experimental hatchery, 

 and the feeding experiments were accordingly transferred to this 

 station where they have been carried on each summer since that date. 

 Unfortunately, for various reasons it has been found impracticable 

 to continue the experiments during the winter. 



METHODS 



In all cases the feeding experiments were conducted under as 

 nearly average hatcliery conditions as ])()ssible. Fingerling trout 

 were kej>t in standard hatcliery troughs siijji)lied witli s[)ring water 

 of a uniforiM temperature. In ordei' to reduce the j:)<)ssibility of 

 oxygen deficiency to a minimum, the water was fully aerated before 

 entering the troughs, and the flow was considerably in excess of 

 that required by such small numbers of fish. 



