PROGRESS IN BIOLOGICAL INQUIRIES, 19 31 501 



Fish meals of various kinds have been used extensively in previous 

 experiments and, on the whole, have given quite satisfactory results. 

 This is especially true of the vacuum and steam-dried meals, which 

 are superior to flame-dried and more uniform in quality. A haddock 

 meal dried by a new process at low temperature gave excellent re- 

 sults in combination with beef liver. In fact, this mixture gave the 

 best growth that has been obtained Avith any fish meal and was only 

 slightly behind the salmon eggs in this respect. The mortality was 

 also very low. 



In previous years several attempts have been made to carry experi- 

 mental lots through the summer on dry foods alone, but all such 

 attempts have encled in failure. During 1931 another attempt was 

 made to keep fingerling trout on a ration containing no fresh meat. 

 These fish were fed a ration composed of 58 parts dextrin, 34 parts 

 haddock meal, and 8 parts dry beef liver. The haddock meal and 

 beef liver were dried by the special low-temperature process referred 

 to above. The fish made a good growth, and for the first time in the 

 history of these experiments it was possible to carry them through the 

 summer without feeding any fresh meat. The mortality, however, 

 was higher than normal, and it is not believed that the use of this 

 ration for any length of time is to be recommended. 



In addition to the experiments with fingerling trout, 20 lots of 

 yearling brook trout were placed on experimental diets May 27 in 

 small outdoor pools. These experiments were to have been con- 

 tinued through the summer, but owing to an unsuitable water supply 

 this was found to be impossible. Unlike the hatchery troughs which 

 are supplied with spring water, the only supply for these pools is 

 from a neighboring brook. On account of exceptionally hot, dry 

 weather early in the summer, the temperature of the brook rose to 

 72° F. This brought on an epidemic of furunculosis, which caused 

 such a heavy mortality — in one lot reaching 86 per cent — that the 

 experiments were discontinued on July 9. 



Although these experiments with yearling brook trout were car- 

 ried on for such a short time that the results are of little value, there 

 is one aspect of the case which is of more than passing interest. 

 During the previous summer there was some evidence that the inclu- 

 sion of salmon eggs in the diet tended to make the fish less susceptible 

 to furunculosis, and experience with the yearling trout strongly 

 supports this conclusion. With one exception, the fish which showed 

 the lowest mortality were on diets containing 50 per cent of salmon 

 eggs. The single exception was a lot fed a mixture of equal parts 

 salmon eggs and pig melts, in which the mortality was double that 

 of any other lot receiving an equal amount of salmon eggs. As 

 previously mentioned, this ration resulted in such heavy mortality 

 with fingerling trout that it was discontinued early in the season. 

 There can be no doubt that the lower mortality in the lots fed salmon 

 eggs was due to the food, since in all other respects the experimental 

 lots were treated alike. 



Breeding experiments. — The work on selective breeding, which 

 was started several years ago, has been continued along the same 

 general lines as in previous years. Although it will be some time 

 before the full possibilities of this work will be evident, each year 

 shows a marked advance in the results attained. Progress would 



