And yet mankind does possess knowledge of conduct, which does not differ in kind from scientific knowledge; there is, in fact, but one kind of knowledge, which passes into scientific knowledge as it becomes more definite and articulate. The knowledge that mankind possesses consists in what we have thus far taken for granted, that under the same circumstances of outward environment and inward character, human conduct does not change. Of society, as of an organism, we cannot say _a priori_ that it is so and could not have been otherwise; we can only show, _a posteriori_, how different parts mutually imply each other, so that, given the whole, we can see that any particular part could not have been otherwise. Our gain from such knowledge is the recognition that there may be discoverable laws of growth essentially relevant to our investigation of conduct. So long as reasoning was conducted upon the tacit assumption that social phenomena can be satisfactorily explained by studying their constituent elements separately, attention was diverted from the important principles of the interrelation of parts to the whole. The theory of evolution brings out the fact that every organism, whether social or individual, represents the product of an indefinite series of adjustments between it and its environment. Every race or society is part of a larger system, product of the continuous play of a number of forces constantly shifting with an effort towards general equilibrium, so that every permanent property represents, not an accidental similarity, but a correspondence between the organism and some permanent conditions of life. To solve the problem of existence by calculation is an impossibility; but our own lives are working it out; the evolution of history is the solution of our problem. And when we fully recognize that a problem is being solved, we have only to gain some appreciation of its general nature and conditions, in order to reach some important, though limited, conclusions, which may fairly be called scientific, as to the meaning of the answer. These conclusions are not scientific in the sense of giving us quantitative and precise formulae, but they may be so far scientific as to be certain and reliable. Thus we may be able to show how a given set of instincts corresponds to certain permanent conditions under which they were developed, and (returning to the problem of differing theories of morals with which we started) to show what is the cause of differing opinions. Our investigations of the problem of morality have nothing to do, in the first instance, with moral principles which are, or profess to be, deduced from pure logic, independent of any particular fact; they deal with actual moral sentiments as historical facts. The word moral, as used in our considerations, does not, therefore, refer to an ideal moral code, but to the one actually existing in the case considered.