Though it is possible to purchase a certain number of salmon ova, this is not at all a satisfactory way of obtaining them. To begin with, it is impossible to get them in sufficient numbers to carry out operations on a large enough scale. Salmon ova are also expensive; and it is no use working with less than half a million in several stations if the river is of any size. It is advisable that the ova should be obtained from the fish. This may be done either by collecting the ova deposited by the fish in the spawning beds or from the gravid females. The latter course necessitates the ripe female and male fish being caught and artificially spawned. As in nature, at best but a comparatively small percentage of the ova are impregnated, and by artificial spawning over ninety per cent. of them may be successfully hatched out, there can be but little doubt as to which is the better way. It is difficult to make sure of catching the fish just at the time they are ripe, so it is advisable to impound them in a fenced-off portion of the river, where they may be got at easily. In the ripe female the ova flow out very readily, and but little pressure is necessary. Hard pressure on the abdomen should never be applied, as it is sure to injure the fish. A ripe female having been obtained, from which the ova flow readily, the female is held over a perfectly clean tin or earthenware dish--wet, but containing no water--and the ova are caused to flow into it by gently but firmly pressing the hand on the abdomen, and stroking it down towards the vent. Milt from a ripe male fish is then allowed to run over the ova in the dish, and is made to run well between them by tilting the dish about from side to side. The ova will now adhere together, and some water should be added. This water should be poured off and fresh added till the superfluous milt is washed away, when the ova should be left in the water till they separate, which will be in about twenty minutes or half an hour. The fertilized ova thus obtained may either be laid down in artificially protected hatching beds, or may be transferred to a hatchery. The latter proceeding, of course, requires a hatching house specially built and arranged, and as this is outside the scope of the present work, I would refer my readers to larger works upon the subject, such as _An Angler's Paradise_, by J. J. Armistead. Of course, by using a hatchery a large number of the eggs will be saved, ninety per cent. of them should hatch out. This is, therefore, obviously the best way to proceed. A very much larger number of eggs will, however, be hatched out in properly-chosen artificial beds than would be the case if they were left to nature.