144 V. S. BUREAU OF FISHERIES 



that at some hatcheries trout appear to do well on a ration which 

 at other hatcheries has made a relatively poor showing. In such 

 cases, however, it would jjrohably be found that a more suitable 

 ration would give correspondingly better results. The average trout 

 culturist is satisfied if his fish appear to be making a good growth 

 without excessive mortality, and he naturally chooses the cheapest 

 food which will produce these results. 



Discrepancies in results with similar rations due to age and 

 species of fish are even more important than those traceable to 

 environmental conditions. Our experiments indicate clearly, for in- 

 stance, that many foods which are suitable for trout 4 to"^ 5 inches 

 long and upward are entirely unsuited to small fingerlings. Fur- 

 thermore, a food which makes a good showing with one species of 

 trout may yield very unsatisfactory results with another species. 

 This means that each species of trout must be treated as a more or 

 less independent problem, although no doubt in most instances quite 

 similar results will be obtained with different species on the same 

 ration. 



In the p)ast there has been a tendency to place too much emphasis 

 on the nutritive elements in the food and too little on its physical 

 consistency and palatability. It is obvious that no ration can give 

 satisfactory results unless it can be fed with little waste and is eaten 

 with avidity by the fish. It may contain the proper proportions of 

 proteins, carbohydrates, fats, and other essential elements; but if it 

 disintegrates quickly in the water or is not relished by the trout it can 

 not fail to give unsatisfactory results. In many instances we have 

 found that the growth and vigor of the fish are directly proportional 

 to the amount of food they consume. Xot infre(iuently when fish are 

 not doing well on Avhat theoretically should be a satisfactory diet we 

 have found that they were eating only a portion of their food. This 

 is especially true when using combinations of fresh meat and dry 

 products; in many instances the fish simply pick out the meat and 

 leave the rest. In this respect one of the most surprising results to us 

 IS that with certain combinations the fish will readily eat the entire 

 mixture, while with others — differing only in the meat constituent — 

 they will cat only the meat, disregarding the rest. Consequently, in 

 making up food combinations, every effort should be made to produce 

 a mixture which will be eaten readily with as little waste as possible. 



In general, straight meat diets when properly ground are of the 

 ])roper consistency to feed to trout with comparatively little waste. 

 Different meat products, however, differ considerably in this respect 

 and some are characterized by much more waste than others. This 

 is especially true of liver which contains a considerable percentage of 

 soluble constituents. Moreover, when finely ground, much of the 

 matei'ial is carried away in suspension before the fish can get it. If 

 it were not for this fact beef liver would probably prove to be the 

 most satisfactory food available for advanced fry and very small 

 fingerlings; but, owing to the lower mortality of these fish on beef 

 heart, we consider the latter somewhat superior to beef liver. Later, 

 as emphasized elsewhere, beef liver gives better results if not ground 

 too fine, and in combination with beef heart appears to make an ideal 

 meat ration. 



