NATUEAL HISTORY OF IMPORTANT FOOD-FISHES. 233 



tliis way a larg-e number would be able to discliarge their spawn without 

 any interruption, the result of wliieli should have been seen in an in- 

 creased number of young' fish. This reasonino-, however, would hardly 

 explain the presence of so many young fish in 1871, since the traps were 

 in position i,n 1870, before the spawning fisli arrived. 



According to Mr. Edwards, scup were still more scarce than usual at 

 Wood's Hole iu 1872 up to the 1st of July, a few small ones only being 

 seen, and none of any size taken with the hook. After that date, how- 

 ever, the two-year fish made their api)earance in considerable numbers. 



In reference to the movement of the scup in Narragansett Bay, the 

 testimony taken both by the Rhode Island commission, and by Mr. Theo- 

 dore Lyman and myself, was quite coutradictory, some maintaining that 

 they enter by the west passage, and, passing round the north end, fall 

 into the traps set for them in the eastern passage ; others insisting that 

 the fish enter both passages at the same time. The general im])ression, 

 however, seemed to be that the army of old fish did uot})ass u[) into the 

 bay, but that probably while the main body kept along the shore, from 

 headland to headland, only those that were originally spawned up the 

 bay turned off and proceeded up toward its head. The success of the 

 traps at Seaconnet is jirobably due in large part to the peculiar funnel- 

 shape of the river, by which the tish at flood-tide would be carried out 

 of their course. The traps there being always set toward the north, it 

 is likely that during- the flood the fish pass up along the channel, and as 

 the tide turns, losing theiv direction, they are scattered toward the 

 shores, and in following down the ebb are taken in the traps. 



According to Captain Thomas Hinckley, after passing Seaconnet Point 

 and entering Buzzard's Bay, the scup keep along the northern shore and 

 malce ahuost the entire circuit of the bay before appearing at Quisset 

 Harbor and Wood's Hole, their appearance being always later there 

 than at the head of the bay or about New Bedford. Whether it is the 

 fish alone that belong to Buzzard's Bay that enter it, or whether others 

 pass directly between the Elizabeth Islands and Martha's Vineyard, is 

 not yet satisfactorily ascertained. We know, however, that they reached 

 Waquoit, the first pound on the north side of Vineyard Sound, in 1871 

 as early as April 25, but that the largest numbers were taken from the 

 10th to the loth of IMay. This gives about a week's difference between 

 this point and Xewport. 



On the south side of Vineyard Sound the fish are netted at Menem- 

 sha Bight, where there are several large and effective pounds, three 

 days or a week earlier than at Lombard'sGove, and nearly two weeks 

 earlier than at the Wood's Hole pound. 



According to Mr. Luce, breeding-fish enter the tidal ponds on the 

 north side of Martha's Vineyard (formerly in large numbers,) where they 

 s[)awn, accomplishing this operation by the end of June, tiie ponds 

 being filled with young in August. As soon as frost comes these fish 

 leave for their winter abode. 



As a general rule, in their movement along the coast the scup are not 

 found in water shallower than a few fathoms ; and it sometimes happens, 

 in the course of heavy storms, that in consequence of the discoloration 

 of the water near the shore, tlie fish move farther out to sea, and on such 

 occasions measurably escape falling into the traps. 



The scup is very largely a bottom-feeder, and depends very much upon 

 moUusks or shell-iish for subsistence. I have been informed by the fish- 

 ermen that they may frequently be seen feeding upon small bivalves 

 of different species, rooting them out of the sand or mud. The stom- 

 achs of about tw^o hundred 1^-pouud scup were examined at one time iu 



