MEATS AS TROUT FOOD 131 



from fresh meat, although with these older fish the difference was 

 less marked. 



From these results we can not recommend fresh fish as an impor- 

 tant item in the menu of trout. In the case of hatcheries located 

 near an abundant supply cheap fish mip:ht prove economical when 

 fed in conjunction with meat, but its sole use emphatically is not 

 recommended. 



COOKED MEATS 



Most fish culturists agree that cooked meats do not prove as satis- 

 factory for trout as raw meats, and our own experiments have 

 further verified this belief. In all cases poorer growth and higher 

 mortality followed the use of cooked meat which was not eaten by 

 the fish as readily as the raw product. An experimental lot of 

 brook-trout fingerlings kept for 103 days on cooked liver suffered a 

 loss four times greater than a lot on raw liver. There thus appears 

 to be no advantage in cooking meats before feeding. Although fish 

 on cooked meats might appear healthy for some time following its 

 use it is safe to say that continued feeding of nothing but cooked 

 meat will bring very unsatisfactory results. 



DRIED ANIMAL PRODUCTS 



Prior to 1929 practically all the dry products used in our experi- 

 ments were mixed with warm water so as to form a thick mush 

 before being incorporated with the ground meat; and, unless spe- 

 cifically stated otherwise, it is to be understood that this was the 

 method employed. In moistening the meals the amount of water 

 required to make a mixture of the proper consistency was carefully 

 determined in the case of each meal, and this amount was rigidly 

 adhered to in all our Avork. The actual percentage of each dry 

 product used in our experimental rations can be determined by 

 making allowances for a water content as follows : Clam meal and 

 buttermilk, 50 per cent ; shrimp meal, 60 per cent ; haddock and 

 menhaden meals, 65 per cent ; and cod-liver meal, 67 per cent. 



This method was adopted since it was thought that by moistening 

 the dry products before they were mixed with the meats a mixture 

 was obtained which was eaten more readily than when the dry 

 meals were added directly to the ration. Carefully controlled 

 experiments during 1929 have convinced us, however, that with 

 most dry products this is not the case, although there are some, such 

 as coarse shrimp meal, which can not be fed without preliminary 

 moistening. 



" CLAM heads" 



" Clam heads " is the trade name given to a by-product of the 

 clam canneries. It consists almost entirely of the siphons and small 

 fragments of the gills and mantle, which are discarded by canners. 

 This product is dried and either placed on market as it comes from 

 the driers or ground into a coarse or fine meal. " Clam heads " has 

 given us by far the best results of any dry product with which we 

 have experimented ; but, unfortunately, there is only a limited quan- 



