We have said that whitefish work up through the trays and pass over the rim into the reservoirs. This is the habit with whitefish and with shad, but the salmon, salmon trout and brook trout, work down into the bottom of the boxes, and if the eggs are left in them until they are completely hatched the fry must be taken out when the. trays are removed. There is some danger connected with this peculiarity, as the little fellows until the umbilical sac is absorbed, crowd together one upon another at the bottom, instinctively hiding from imaginary enemies, and will often get smothered. They can be drawn out by a syphon consisting of simply a piece of ordinary rubber pipe leading into a pail of water and will not be injured by the operation; or, if the fish are to be left in, a tin tube pierced with fine holes is put in the box and the upper end of the syphon led into that. However, the tin pipe where it enters the bottom of the box is usually constructed with a hole and a cork fitted to it, by means of which the fry can be drawn off and the box emptied.

The method of inspecting the eggs in this hatching box is as follows: Two wooden wedges are inserted, one on each side of the uprights attached to the bottom frame to hold them in place. One box is always left empty. The trays in the adjoining box are raised so as to bring the upper one just below the surface of the water and are wedged ; this is examined, and after the dead eggs are removed it is placed in the empty box, lowered under water and wedged. When this is finished another is raised in the box which is being examined, and transferred after inspection to the other box. By wedging them each set of trays is kept in its place, and can be examined systematically. When they are all finished, that box is empty and may be cleaned if necessary. Then the trays in the next adjoining box are examined and transferred to it in the same way, and so on with one after the other till they are all completed and an empty box left at the further end of the row for use on the next occasion. If the water is pure this operation will not have to be performed more than once in two days, as the tendency of these boxes is to clean themselves and to prevent the retention of sediment.

By the use of the Holton boxes an immense condensation of room can be effected ; each tray will hold more than ten thousand brook trout eggs, and the total number in each box will be about two hundred thousand. Our usual practice with trout eggs is to remove them from the boxes when they are about hatching and place them on the trays or gravel in the troughs where they will have more room, and can get out of their shells more readily. The empty shells will otherwise, sometimes clog the wires and stick to the embryos, unless the boxes are examined very frequently during the last stages of development.

Glass Jars

A pretty and ornamental modification of the Holton box, not essentially differing from it on principle, can be made by the use of glass, and will answer on a moderate scale. It consists of round glass jars filled with trays with the water entering at the bottom, which should be cast in the shape of a funnel to which is attached an India rubber pipe. The water passes out above through another pipe to the bottom of the next, and so on. The advantages claimed for this arrangement are the neatness, beauty and cleanliness, and the ease with which the eggs can be watched and studied at all times. It is well to have the supply pipe smaller than the connecting pipes between the jars, for fear of overflow in case of accidental obstruction, and from the first there should be a gradual descent, each jar standing a little lower than the one preceding it. When it is desirable to empty any jar it can be done with a tube of rubber used as a syphon, the supply of water being shut off. By drawing off the water into a basin, any young fish carried with it may be saved.

Self-Picker

Another modification of the Holton box was made at the state establishment. It has been called a "self picker." It consists simply in removing all the trays and placing the piece of tin over the hole through which the water enters, which should be in the center of the bottom in such a way as to force the water to rise gently on all sides close to the edges. The box is shaped so that the sides converge towards this center and the water makes a moderate current along them upwards. The eggs which are placed loosely in the box are kept in a gentle bubbling or boiling motion by this arrangement. They are caught by the current, carried upward, and then dropping into an eddy, descend to be again whirled upward. When the fish hatch the general motion of the water is strong enough to carry the shells over the outlet, and it is manifest that by such an arrangement no sediment can at any time collect on the eggs. It was feared at first that the continual motion might wear through the shells. It has been used with trout, and is the proper arrangement if shad eggs are to be hatched in the Holton box, as for them trays are not necessary. If salmon or trout eggs are to be treated, a much smaller number must be put in each box, than it they were distributed on the trays. The original arrangement of the boxes is the best for general purposes, and these modifications, are only mentioned in case they should fit peculiar circumstances.

Glass Trays

It seems hardly necessary to refer to trays which have small glass tubes instead of wire netting on the bottom. These were among the first plans tried and will answer very well. But they are expensive to make and easy to break, while they are no better than our cheaper and stronger trays.

Our Practice

A number of other arrangements could be detailed, but these are the best, and cover all conceivable circumstances. We have given so many that perhaps we had better state exactly what course we follow as the best, simplest, cheapest, easiest and most certain. We hatch whitefish in the Holton box, completing the operation there and allowing the fish to break their eggs, to work their way up through the wires, and pass over the upper rim into a reservoir, the egg shells going over at the same time. We have given up trying to feed whitefish and distribute them at once to the lakes they are to inhabit by depositing them in the water suitable for them.

Brook-trout, salmon trout and salmon eggs we develop in the Holton boxes till they are about to burst their shells, then we place them on trays in the troughs. As they remain much longer in the boxes than in the troughs less room is needed in the latter than would be required if they were entirely hatched there. We give them a good current, and as soon as they are hatched and the egg shells removed, we spread gravel over the bottom of the troughs, or remove them to troughs with gravel on the bottom, as we believe they are more contented, so long as the umbilical sac remains, if they can hide their heads, or imagine that they are hid between the stones. We observe that if they are kept at this stage in bare troughs they are uneasy and keep struggling about from place to place. Just before the sac is absorbed we watch them carefully, especially the brook-trout, to prevent their crowding on one another, and when they do so they are removed in a broad flat net from the head of the trough where they congregate to the lower end, or are otherwise spread out. This is done at night as well as during the day when there is a large number together. As soon as the sac is absorbed we remove the gravel and keep them in clean troughs with nothing in them to catch or conceal dirt. We again increase the current and feed the young six times a day on liver comminuted to the utmost possible degree, and diluted with water. We distribute the fry as soon as possible thereafter, and put them in the smallest spring runs connected with the streams or ponds they are to stock, and not directly into the ponds or streams themselves.