Leaving the question of use-inheritance open, I pass on to say a word about variation as considered in itself and apart from this doubtful influence.  Weismann holds, that organisms resulting from the union of two cells are more variable than those produced out of a single one.  On this view, variation depends largely on the laws of the interaction of the dissimilar characters brought together in cell-union.  But what are these laws?  The best that can be said is that we are getting to know a little more about them every day.  Amongst other lines of inquiry, the so-called Mendelian experiments promise to clear up much that is at present dark.

The development of the individual that results from such cell-union is no mere mixture or addition, but a process of selective organization.  To put it very absurdly, one does not find a pair of two-legged parents having a child with legs as big as the two sets of legs together, or with four legs, two of them of one shape and two of another.  In other words, of the possibilities contributed by the father and mother, some are taken and some are left in the case of any one child.  Further, different children will represent different selections from amongst the germinal elements.  Mendelism, by the way, is especially concerned to find out the law according to which the different types of organization are distributed between the offspring.  Each child, meanwhile, is a unique individual, a living whole with an organization of its very own.  This means that its constituent elements form a system.  They stand to each other in relations of mutual support.  In short, life is possible because there is balance.

This general state of balance, however, is able to go along with a lot of special balancings that seem largely independent of each other.  It is important to remember this when we come a little later on to consider the instincts.  All sorts of lesser systems prevail within the larger system represented by the individual organism.  It is just as if within the state with its central government there were a number of county councils, municipal corporations, and so on, each of them enjoying a certain measure of self-government on its own account.  Thus we can see in a very general way how it is that so much variation is possible.  The selective organization, which from amongst the germinal elements precipitates ever so many and different forms of fresh life, is so loose and elastic that a working arrangement between the parts can be reached in all sorts of directions.  The lesser systems are so far self-governing that they can be trusted to get along in almost any combination; though of course some combinations are naturally stronger and more stable than the rest, and hence tend to outlast them, or, as the phrase goes, to be preserved by natural selection.