But if the question be asked, how _the first majority_ of determinants varying in the same direction is brought about, there are two possibilities: first, by chance, and secondly, by influences which cause particular determinants of all the ids to vary in almost exactly the same manner. We shall find illustrations of the latter among climatic varieties; but the cases of the first kind are the more important, for they form the foundation and the starting-point for processes of selection of a higher order, for personal selection. It might seem perplexing that processes of such importance should depend ultimately upon chance; but when we remember that there are only two directions of variation, namely a plus direction or a minus direction, we recognize that the chance of a majority in one direction or another is much greater than that of absolute equilibrium between the two, and there is therefore a very strong probability that in many individuals of the species either the upward or the downward movement of a determinant _A_ will preponderate. Now as such variation movements, when they are of a certain strength, increase automatically, we can easily see that they must gradually attain to a level at which they acquire selection value, and how then, by personal selection, the ids with favourably varying determinants may be collected together. Of course it is not possible to state positively the time at which in individual cases a variation acquires a biological significance, that is, selection value. We can only say in a general way that, as soon as it attains this, personal selection either in a positive or a negative sense _must_ intervene; an injurious variation tends to the elimination of its possessor, a useful one increases the probability of its survival. There must, however, be for every variation a stage of development in which it has as yet no decisive biological importance, and this stage need not by any means be so insignificant that we cannot see it, or can hardly do so: in other words, there are characters which have arisen through germinal selection, which are of purely 'morphological importance.'