The Migration of Fishes. ; 55 
(1) It presupposes less sudden changes of temperature than 
that of hybernation. It has been shown that hybernation of 
fishes is never voluntary, but is a state of torpidity, induced, 
like that of zstivation, by a change of temperatures and sur- 
roundings, which they have no power to avoid. Before enter- 
ing upon hybernation or estivation, fishes retreat to the 
deepest water, and only become torpid when they are followed 
thither by the changed conditions of existence. In the fresh 
water of temperate countries fishes do not become entirely tor- 
pid in cold weater, but are sufficiently active, to be taken with 
hooks from under the ice. This is also the case in sub-polar 
regions. The kalleraglitz, or American turbot (Retnhardtius 
hippoglossoides), is taken with hooks in the dead of winter 
under the floe-ice of North Greenland at a depth of 300 
fathoms, in South Greenland, on the oceanic banks, at 60 and 
and 80 fathoms, and at Fortune Bay, Newfoundland, it is cap- 
tured in the shore herring-seines at the same season. 
So long as the menhaden can avoid the extremes of tem- 
perature, which they so carefully avoid in summer by seek- 
ing congenial warmth in the ocean strata under the Gulf 
Stream, need we suppose that they will plunge into the colder 
strata below? 
(2.) It involves less radical changes than hybernation in 
the habits of the fishes. Some fishes, like the mud-minnow 
(Umbra limi) of the eastern United States, are peculiarly 
adapted for a life in the mud. Others, such as the “com- 
pound breathers” (Ladrynthici) of India, are said to respire 
with ease with their head covered by liquid mud. Such fishes, 
however, are totally different in organization from the free 
swimming species of the open seas. All free swimmers are 
especially heedful to avoid contact with the bottom. This is 
so in the case of the herring family, of which the men- 
