The Migration of Fishes. 31 
the period when they should be gathered together in the 
southern waters, prove that a portion at least remain. It is 
difficult to imagine how a shad or a river-herring, spawned in 
the St. Lawrence river or any northern stream, could avoid 
entcring a more southern river, if in its vicinity; but if any 
fact has been well established of late years in the history of 
fishes, it is that the anedromous fish, or such as run up the 
rivers from the sea to spawn, will return, if possible, to the 
river in which they first saw the light. So true is this, that 
where there may be two or three rivers entering the sea in 
close proximity, which have become destitute of shad or her- 
ring, in consequence of long-continued obstructions, and the 
central one only has been restocked by artificial means, the 
fish, year by year, will enter that stream, while those adjacent 
on either side will continue as barren of fish as before.” 
ft. EDR INFLUENCE. OF TEMPERATURE .ON TEE 
MIGRATION OF FISHES, ILLUSTRATED BY A 
STUDY OF THE PERIODICAL MOVEMENTS OF 
THE SGHOOLS OF MEMEADEN. 
I. COMMENTS ON THE TEMPERATURE TABLES USED. 
The influence of ocean temperature on the menhaden is 
not at all well understood, and I can here record only crude 
generalizations founded upon very unsatisfactory data.ai gl 
have before me three tables, showing the variation of temper- 
ature by monthly means for Key West, Fla.; Jacksonville, 
Fla.; Savannah, Ga.; Charleston, S. C.; Wilmington, N. C.; 
Norfolk, Va.; Baltimore, Md.; New York City ; New London, 
Conn.; Woods Holl, Mass.; Portland, Me.; and Eastport, Me, 
Table I. shows the monthly means of surface temperature ; 
table II. the temperature at the bottom near the shore, and 
