96 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



great, especially taking into account the shallowness of the estuary in question. The 

 same has been noticed elsewhere where conditions are somewhat similar. But in some 

 of our deeper lochs, where the water would not be affected so easily, the herrings still 

 remain. 



The migration of the herring is not to and from the Arctic regions, as was once 

 supposed and as was reported by all the older naturalists, but simply from deep to 

 shallow water, and I am convinced that they really remain comparatively near our 

 islands all the time. Their migration, it has been found, depends largely on temper- 

 ature. The study of the movements of these fishes is a very interesting one, and well 

 worthy of a considerable amount of time, labor, and money being spent on it. No 

 one who has not seen a herring shoal can form a conception of what it is like or of the 

 enormous numbers of fish it contains. If we allow a herring for every cubic foot and 

 assume a shoal to be a square mile in extent and 18 feet deep, it would contain 

 500,000,000 of fish ; and there are a great many shoals of vastly larger dimensions. 

 Such, indeed, is the enormous rate of increase that the whole quantity caught by man 

 does not appreciably affect their numbers, and they would literally choke up the sea 

 if they were not largely destroyed by other fishes as well as by birds. One of the 

 fishes most destructive to the herring is the cod, which follows the shoals as well as 

 the salmon, to which I have already referred. 



The return of this latter fish to the rivers is an exceedingly important economic 

 question. During autumn, when they run up the river to spawn, they take, perhaps, 

 less food than at any other time, but the abundance amongst which they have been 

 living previously has caused such a development of fat that the fish are really pro- 

 vided by nature to a great extent with the food which they require at this particular 

 time. Very much the same has been observed with regard to domesticated fish. On 

 a trout farm, for instance, it is observed that as soon as we reach October the fish 

 which are in the habit of spawning at that time almost cease to feed, whereas the occu- 

 pants of other ponds which do not spawn until January go on feeding until hard frost 

 sets in and the water becomes very low in temperature. 



Fish being cold-blooded animals pass into a more or less torpid condition dur- 

 ing cold weather. The influence of an approaching thunder storm in preventing 

 fish from rising is well known. Fish also often do not rise when the barometer is 

 falling, whereas when we have a steady rise in the barometer from the west a good 

 run of salmon is often noticed in the west coast rivers. But when the barometer 

 reaches 29.50° the run ceases. It has also been observed that they run better in 

 west coast rivers when the barometer is lower on the west coast than it is on the 

 east coast. When it is rising and reaches 29° the best run occurs; and from 29.50° 

 seems to be the most favorable point. 



Temperature is also an important factor in connection with the migration of fishes 

 of probably all kinds. It must generally happen that the temperature of a river dif- 

 fers from that of the sea or the estuary. Many writers have remarked that the low 

 temperature of the sea induces the salmon to leave it and seek the higher temperature 

 of our rivers. This has been particularly remarked about our North Sea and east 

 coast rivers, which are earlier than those on the west coast. Fish-culture has taught 

 us a good deal here. It was found that in cold weather fish were later in spawning 

 than in mild weather: that during a hard frost they spawn very tardily j and when 

 there is a mild rain and rapid thaw they spawn more freely than under any other 



