146 U, S, BUREAU OF FISHERIES 



meat in the mixture is essential. Although Thompson (1929) has 

 reported great success in feeding dehydrated salmon eggs and also 

 a meal composed of a mixture of dried horse liver and blood for 

 several months without the addition of fresh meat to the diet we 

 are still not convinced that trout can be kept on dry foods 

 indefinitely. 



In our experimental work it was found that lake-trout fingerlings 

 started on dried shrimp meats appeared to be perfectly normal 

 for over three Aveeks. At the middle of the fourth week, however, 

 a sudden high mortality set in which was checked by the addition 

 of 50 per cent fresh liver to the ration. The fingerlings were kept 

 on this mixture throughout the summer with normal growth and 

 insignificant losses. 



In common with McKay and his coworkers (McKay and Dilley, 

 1927) we have found that dry foods alone result in stunting and ex- 

 cessive losses with brook-trout fingerlings. (Fig. 10.) Each lot 

 developed characteristic symptoms, the growth falling off and the 

 mortality rising until at the close of the experiments the only sur- 

 viving fish were a few stunted, emaciated creatures, which doubtless 

 would have died too had the ration been continued. 



The methods emjiloyed at various hatcheries in feeding dry pro- 

 ducts differ quite widely. In a few instances they are fed straight 

 the fish being given an occasional meal of fresh meat. More often 

 the dry meals are mixed directly with meat ; and we believe that, in 

 general, this method is greatly to be preferred. In many of our 

 experimental rations the dry meal was moistened with hot water 

 imtil it formed a thick mush before being mixed with the ground 

 meat. HoAvever, we are noAv convinced tluit in most instances better 

 results can be obtained by mixing the dry meal directly with the 

 meat Avithout preliminary moistening and Ave haA'^e adopted this as 

 the standard method foi- use in future experiments. This method 

 results in a somcAvhat more concentrated food, and the meat juices 

 are absorbed more completely by the dry constituent. If the mixture 

 is alloAved to stand for a short time before being fed the dry particles 

 will have become sufficiently soft to permit of their being taken 

 readily by the fish. 



As high as 25 per cent of clam meal or dry milks may be fed 

 successfully to small fingerlings, but it is probable that a larger per- 

 centage aa-ouIcI result in considerable amounts being wasted. With 

 larger fingerlings and yearliiigs, hoAvever, a much higher percentage 

 of meals can be used to advantage. We have not yet determined 

 definitely the maximum amount of the dry constituent which can be 

 safely included in trout rations. Doubtless this Avill vary consid- 

 erably in different combinations. In some of our experiments with 

 yearling trout Ave have obtained good results with as high as 60 per 

 cent of the dry product and it is very probable that the meat con- 

 .stituent can be still further reduced \vithout seriously impairing the 

 efficiency of the ration. 



BIBLIOGRAPHY 

 BiNG, Franklin C ' 



1927. A progress report upon feeding exi)eriment.s with brook-trout finger- 

 lings at the (Connecticut Htate Fish Hatchery, Burlington, Conn. 

 Transactions, American Fisheries Society, vol. 57, 1927, pp. 266- 

 273, 1 fig. Hartford, Conn. 



