These fish which are a variety of the salmon-oids are to be treated very differently from the trout and salmon, for their eggs are exceedingly adhesive. They ascend the rivers in early spring. In the neighborhood of New York they are caught largely in the Raritan and Passaic rivers.

About five miles above the city of New Brunswick there is a dam which blocks the river, and which the smelt cannot surmount. The fishing grounds, extend from the old wooden city bridge down the river for two or more miles; very little, if any fishing being done above the bridge, on account partially of the little depth of the water, partially because the smelt appear to pass down the river again, after being impeded in their onward course by the dam. The smelt are caught entirely with seines, which include in their sweep, nearly the entire breadth of the river, averaging about thirty rods.

The seines vary from thirty to sixty fathoms in length, one hundred and eighty to three hundred and sixty feet, and are about fifteen feet in breadth, with meshes one-half inch square. The time of working the seines depends much upon the state of the weather and the water, but as a rule, the fishermen are engaged early in the morning and again in the afternoon.

The smelt spawn throughout the month of March, the eggs are small and so adhesive that they must be deposited upon the trays where they are to remain. There are about forty thousand eggs to each medium-sized fish, and they will hatch in about a month with a temperature of water of from thirty-five degrees to forty degrees, or in the ordinary water in the river in about eighteen days. The spawning fish, as fast as captured, should be placed in tubs, or, if not ripe, they may be kept in ponds till the eggs mature. When they are to be handled a tray dipped in water should be placed in a tin pan without any water in it. The eggs are stripped directly on the tray, and the milt, as soon thereafter as possible, then a little water should be added, just enough to cover the tray, and the whole shaken about till the eggs are evenly distributed. A few minutes expire before they adhere finally, but when adhesion once takes place they must remain undisturbed till they hatch. The time of development is so short that there is no trouble in their management, and they may be hatched in unlimited numbers. The spawners may be stripped directly into a shad hatching box and that left in the current of the river, and a large number hatched in an ordinary fish car, in which the parents had been confined to mature their eggs and in which they had spawned of themselves. The trays are removed to the hatching boxes after the eggs have adhered by the hardening of the mucous matter that surrounds them and then treated like trout eggs except that the dead fish cannot be removed.