Lyman.] 126 [February 6, 



army pass from south to north, parallel with the coast, turning In at 

 the different rivers as they go; but that those inhabiting each river 

 lie, during the winter season, in the deep water off its mouth, where 

 they wait the return of the breeding time. The theory corresponds 

 with the pretty well established fact, that this fish, as well as the 

 salmon, returns always to the river in which it was hatched. Their 

 manner of spawning is as follows. They work up a little circular sand- 

 bank, on which the spawns are lodged, and are guarded from that de- 

 struction to which they would be exposed from the small fish, did not 

 the male constantly play round the deposit. This cui'ious sentinel 

 duty may account for the singular fact, that very few male shad are 

 seined in our rivers, except in the night time. It may be that they 

 pass the day in watching the eggs, at the bottom of the stream, after 

 the manner of the stickleback. On the other hand, the females are 

 mostly netted during the day; and they may, perhaps, deposit their 

 spawn only in the night season, when they would lie below the sweep 

 of the seine. The separation of males and females seems to be car- 

 ried out even in the migrations ; for, on the coast of North Carolina, 

 the males arrive several days before their partners. The advance of 

 the mioration up a river is by no means a simultaneous movement, but 

 progresses in a series of " runs." The first run is of fish smaller and 

 not so numerous as those of the grand run which follows. These 

 small fish are probably the progeny of the tributary streams, whose 

 waters are warmer than those of the main river, but do not furnish so 

 nourishing food for the fry hatched in them. As the first shoals enter 

 the river, a few individuals, In whom the ova are probably already 

 mature, make haste to the upper spawning grounds. The main body, 

 however, proceeds leisurely, and while in tidal water, the fish even 

 turn down stream at the flood. In pleasant weather they swim high, 

 and sink to a lower stratum when the surface is ruffled by a cold east 

 wind. After the great run, which may continue from ten days to 

 three weeks, follows another scattered run; and it is a singular fact, 

 that the spawn in the late fish is less developed than it was in the ear- 

 lier ones. Once in their native river, there is little doubt that the 

 companies proceed each to its own spawning ground ; some in the 

 tributaries, some in the main stream, and some, even, in the lakes that 

 communicate with it. Thus it is plain that the operation of spawning 

 cannot be simultaneous for the whole of a river ; on the contrary. It 

 extends over several weeks, some fish being ready to deposit on their 

 first advent, while the eggs of others are comparatively immature. 

 The time of the return of the parent fish to the sea is not so accur- 

 ately known as could be wished ; but, after spawning In May and 

 June, they doubtless pass down gradually from the middle of June to 

 the first of August ; after which there are still to be found stragglers 



