The general truth that extra function is followed by extra growth, must be supplemented by the equally general truth, that beyond a limit, usually soon reached, very little, if any, further modification can be produced. The experiences which we colligate into the one induction thrust the other upon us. After a time no training makes the pugilist or the athlete any stronger. The adult gymnast at last acquires the power to perform certain difficult feats; but certain more difficult feats no additional practice enables him to perform. Years of discipline give the singer a particular loudness and range of voice, beyond which further discipline does not give greater loudness or wider range: on the contrary, increased vocal exercise, causing a waste in excess of repair, is often followed by decrease of power. In the exaltation of the perceptions we see similar limits. The culture which raises the susceptibility of the ear to the intervals and harmonies of notes, will not turn a bad ear into a good one. Lifelong effort fails to make this artist a correct draftsman or that a fine colourist: each does better than he did at first, but each falls short of the power attained by some other artists. Nor is this truth less clearly illustrated among the more complex mental powers. A man may have a mathematical faculty, a poetical faculty, or an oratorical faculty, which special education improves to a certain extent. But unless he is unusually endowed in one of those directions, no amount of education will make him a first-rate mathematician, a first-rate poet, or a first-rate orator. Thus the general fact appears to be that while in each individual certain changes in the proportions of parts may be caused by variations of functions, the congenital structure of each individual puts a limit to the modifiability of every part. Nor is this true of individuals only: it holds, in a sense, of species. Leaving open the question whether, in indefinite times, indefinite modifications may not be produced by inheritance of functionally wrought adaptations; experience proves that within assigned times, the changes wrought in races of organisms by changes of conditions fall within narrow limits. Though by discipline, aided by selective breeding, one variety of horse has had its locomotive power increased considerably beyond the locomotive powers of other varieties; yet further increase takes place, if at all, at an inappreciable rate. The different kinds of dogs, too, in which different forms and capacities have been established, do not now show aptitudes for diverging in the same directions at considerable rates. In domestic animals generally, certain accessions of intelligence have been produced by culture; but accessions beyond these are inconspicuous. It seems that in each species of organism there is a margin for functional oscillations on all sides of a mean state, and a consequent margin for structural variations; that it is possible rapidly to push functional and structural changes towards the extreme of this margin in any direction, both in an individual and in a race; but that to push these changes further in any direction, and so to alter the organism as to bring its mean state up to the extreme of the margin in that direction, is a comparatively slow process.[24]