One of the most valuable, practical inventions in reference to the hatching of the eggs of the salmonidae was made by Marcellus Holton, while in the employ of the New York fishery commission. It consisted of a device for utilizing the upward flow of water among and through the eggs. It had been often noticed that trout sought as a favorite locality for depositing their eggs the parts of the bottom of the ponds or streams through which a spring made its way. Nature taught them thus, to secure a regular flow of clear, unpolluted water of even temperature, working its way between the gravel and eggs of which their nests were composed, and much larger percentage of the spawn deposited in such places hatched than when it was under any other natural conditions, for it was certain to receive precisely what it needed, a steady current of fresh and well aerated water, not increased or diminished in volume, not fouled by rains, not warmed by the sun or cooled by frost, and not bearing with it the germs of fungus or disease. Changes of temperature or condition are injurious to spawn, and by this plan, changes are avoided.

Holton in the arrangement of the hatching box which he invented, created an artificial spring of great volume, conveniently located and thoroughly under control. It was impossible to deposit the eggs over the natural springs, so the springs were brought into the hatching house. This was done by leading the water properly screened through a pipe into the bottom of a box, and allowing it to pass out over the top. A deflector was placed over the aperture in the bottom so as to break the current and distribute the flow equally in all directions, and trays were laid, one upon another in the box so that the water would have to rise through them all before it could escape, and the eggs being on these, must necessarily be kept directly under the influence of a steady but gentle and natural current.

These boxes were constructed especially for the incubation of white fish and shad eggs, although salmon, salmon trout, etc., can just as well be hatched with them. They are about twenty inches square and two feet deep, and will hold eighteen wire trays placed one upon another. The water enters from the bottom, passes up through the box, falls evenly into little troughs which run around the four sides, and is thence carried into a reservoir. The trays are lifted for cleaning by arms at two sides of the box fastened into a lower frame. The trays are made of wire cloth of so small a mesh that the eggs will not pass through, but large enough to permit the passage of the whitefish fry, which, as soon as they break the shell, are carried up by the current and into the reservoir, from which they may be taken out as required. Whitefish eggs placed in one layer will number sixty four to the square inch. Each tray will hold about 18,000 eggs, and each box about 325,000. They can be made larger or smaller and will work equally well.

These boxes are the only device which will hatch white-fish thoroughly well, for whitefish are the most delicate and difficult to manipulate of all the salmon family. The eggs are much smaller than those of the trout and the fry are more tender than even the tender trout fry, which test the fish cullurists best skill to raise satisfactorily. One especial advantage is, the difficulty that sediment of any kind finds in lodging upon the eggs. As it is carried upward, if at all by the current, it is swept along over the upper rim into the reservoir, and rarely catches upon the eggs. It has little opportunity to settle, and the number of eggs killed by this fruitful source of trouble is reduced to the minimum. Where the water passes from above down, it naturally deposits whatever it holds in suspension on its way, but by this reverse process, the sediment cannot fall, and cannot cling to what it touches, because it comes in contact with the underside. The aeration of the eggs is perfect, none are covered up by others so as to be shut out from the passing current and suffocated. In troughs the water strikes the body of eggs on the edge and then simply passes over and along them, not penetrating, but merely coming in contact with their surface; in the Holton box, the water enters every part, works up between every egg, and even if these are piled two or three deep, will vitalize all of them.

It must be remembered that impregnated eggs require the oxygen in the water as much as the living fry and will smother if shut out from it. Anything that does this, whether it be dirt, sediment, insufficient current or other eggs will destroy them and no plan has yet been invented which so absolutely prevents this danger. The economy of room is another important consideration where fish culture is to be prosecuted on a large scale. As we have said, eight boxes will contain and hatch two millions of white fish eggs, and an equal number of boxes of trout eggs will supply the place of a large establishment laid out in the old fashioned and ordinary method.

The openings through the wires are not round or square but a long narrow parallelogram, differing in width for the different sizes of eggs which are to be hatched. To understand this it is only necessary to suppose the wires stretched on the trays at proper distances from one another, and then bound together by cross wires at three-quarters of an inch from one another, or sufficiently near to hold all securely together. The wires for white fish eggs are nine to the inch ; those for brook trout are eight to the inch; seven for salmon trout, and six and a half for salmon. In like way they can be adapted to any other species of fish that it is proposed to hatch. They could be used for shad where it is not desirable or convenient to use the regular shad box which is described in the chapter on shad hatching. For shad, the mesh of the trays is a little different being of heavier wire, square and twenty-two to the inch. These trays, it is unnecessary to say are painted with gas tar like the wood work of the troughs, and inside of the boxes and all other things with which the eggs come in contact. If the wires should not be, they would rust and the rust would penetrate and destroy every egg that touched it. The current may be caused by the natural flow of the water and by artificial means such as the use of a pump, and for shad and whitefish hatching, no trays are absolutely necessary. The boxes may be round or square and made of tin or wood.