284 BULLETIN OF THE BUREAU OF FISHERIES. 
close to shore in shallow water. As suggested before, in the case of menhaden, 
they may enter these places for food or for protection, but as they subsist to a 
great extent upon minute crustaceans which abound outside they could hardly 
be there solely for the sake of food. 
During the period when they are beyond observation it is, of course, not 
known to what extent menhaden and mackerel feed. But the inference is that 
mackerel, at least, and probably menhaden, do not feed during this time, as the 
fish appearing in the spring, which apparently could be only those of the pre- 
vious season’s production, are about the same size as when they left, and early 
adult mackerel are always lean. A similar condition seems to obtain to some 
extent among the menhaden, for often the early menhaden are very poor, 
according to the reports of the fishermen. It can hardly be imagined that the 
fish would refrain from eating for a long period if food were plentiful unless they 
to some extent hibernate. If they do not hibernate—become more or less 
dormant and inactive—it is hard to explain why they should resort to a foodless 
region when food abounds in other places. 
In respect to the menhaden, not so much is known regarding its food as is 
known of the mackerel, but it has been shown to feed to some extent upon 
minute crustaceans, as do mackerel, and such food abounds in the winter months 
in the localities that both menhaden and mackerel have apparently left. The 
natural inference is then that temperature is the controlling factor. Both 
menhaden and mackerel endure a rather wide range of temperature and both 
linger late in the fall in northern littoral waters. Rather curiously, young 
menhaden seem to remain longer than adult fish and will stand reduction of 
temperature to a low degree. An experiment made by Mr. Vinal Edwards at 
Woods Hole, which will be described, supports the statement. The ultimate 
death of the fish may have been due to the sudden changes of conditions to 
which they were subjected rather than to the low degree of temperature alone. 
On October 16 Mr. Edwards placed in the inner poolof the United States 
fisheries station about 1 barrel of young menhaden from 2 to 6 inches long. 
In three or four days they formed together in one school and continued so until 
December 21, when the temperature fell to 35° F. The school then broke up 
and scattered about in the pool, some swimming near the surface, others near 
the bottom. When the sun was shining all would seek the sunny portions of 
the pool, avoiding the shade. On December 24 the temperature in the pool 
was 31.5° F. and most of the menhaden swam near the surface but seemed 
otherwisé undisturbed. On December 26 the temperature dropped to 30° and 
the fish were less active, some swimming on their sides. On the night of Decem- 
ber 26-27 it became much colder and snowed, and in the morning all of the fish 
were dead. 
Thus it would seem that the disappearance of young and old in the late 
fall is to avoid extreme and sudden changes of temperature. They disappear, 
