commencing reproduction.[34] There is, indeed, among fishes, at least one case which appears very anomalous. The male parr, or young of the male salmon, a fish of four or five inches in length, is said to produce milt. Having, at this early stage of its growth, not one-hundredth of the weight of a full-grown salmon, how does its production of milt consist with the alleged general law? The answer must be in great measure hypothetical. If the salmon is (as it appears to be in its young state) a species of fresh-water trout that has contracted the habit of annually migrating to the sea, where it finds a food on which it thrives--if the original size of this species was not much greater than that of the parr (which is nearly as large as some varieties of trout)--and if the limit of growth in the trout tribe is very indefinite, as we know it to be; then we may reasonably infer that the parr has nearly the adult form and size which this species of trout had before it acquired its migratory habit; and that this production of milt is, in such case, a concomitant of the incipient decline of growth naturally arising in the species when living under the conditions of the ancestral species. Should this be so, the immense subsequent growth of the parr into the salmon, consequent on a suddenly-increased facility in obtaining food, removes to a great distance the limit at which assimilation is balanced by expenditure; and has the effect, analogous to that produced in plants, of arresting the incipient reproductive process. A confirmation of this view may be drawn from the fact that when the parr, after its first migration to the sea, returns to fresh water, having increased in a few months from a couple of ounces to five or six pounds, it no longer shows any fitness for propagation: the grilse, or immature salmon, does not produce milt or spawn.