THE DOUBLE-CRESTED CORMORANT. 21 



by enemies before reaching breeding maturity. The birth rate 

 we can assume as being sufficient under normal conditions. 

 But the introduction of man with his efficient and deadly methods 

 of taking fish after they have passed the gauntlet of natural 

 enemies, has introduced a decidedly adverse and abnormal 

 influence against the species. The mature fertile fish is nature's 

 finished product, the outcome of that fine balance it has taken 

 ages of adaptation to evolve and secure. Loss at this stage will 

 be more keenly felt by the species than at any other time in its 

 life-history and, unless compensated for, is bound to have a 

 marked effect upon the total numbers. It is not only the 

 individuals themselves that are lost in this case, but the gener- 

 ation they are about to give rise to, and for which the whole 

 economy of the species has been formed to produce. The 

 proper correction for this is the hatchery which, in so far as it 

 compensates for the eggs of those fertile individuals taken by 

 man, should be sufficient to keep the salmon to the standard 

 allowed by the resultant of other forces. However, no increase 

 in final population beyond the normal capacity of the stocked 

 waters can be expected from these means unless there is food 

 supply for an increased number of the species at all ages. This 

 food supply has to be considered throughout the species' range. 

 The salmon in the sea seem to feed largely upon such fish as 

 herring, and upon Crustacea, both of which are plentiful, and there 

 is little probability of their insufficiency. In the streams it is 

 another question and a determination is less easily arrived at. 

 The fry undoubtedly feed upon microscopic animalcules, the suffi- 

 ciency of which may be tentatively assumed with moderate 

 safety, but we have yet to learn the food supply of the fingerlings 

 or parr. The water of the streams visited is absolutely clear, 

 the bottoms are clean gravel and rock, with no visible abun- 

 dance of algal, plant, or insect life. There is undoubtedly a 

 certain amount of some such food, but it is too scattered to 

 attract the attention and is obviously insufficient for the support 

 of an unlimited number of fish. 



We found that the parr do eat fry of their own species most 

 greedily. As all the fry in the salmon streams are those of salmon 

 and of a few trout the inference is plain, that a large part of the 



