If what Weissman says is true, that no variation or development which comes to an animal during his lifetime can be transferred into his own germ cells and handed on to his children, then it becomes evident that we must find some cause of variation that acts within the germ cells. This is the difficulty which Weissman meets. He says that there are small particles in the nucleus of each cell; that these particles which he calls determinants decide the form and the course of development of that cell; that when that cell divides to produce another cell it gives to this other cell one-half of each determinant. As a result the second cell grows to be like the first. This tells us why offspring are like their parents. There is nothing in the theory thus far to show us why offspring are not exactly like their parents. In other words, there is no accounting, thus far in the theory, for variation. When the biologist studies carefully the history of an egg while it is being formed, he sees that at one stage in its development it throws away not one-half of each determinant, but one-half of the determinants. When an egg does this, it deliberately casts aside one-half of the possibilities of its own development. This throwing away is quite as effective for all its descendants. Any ancestral quality now lost is lost from the line forever. In the formation of the sperm cell set free by the male a similar throwing away of one-half the characters has taken place. The egg cell and the sperm cell fuse together. There are as many possibilities now as there were in either parent, but not all the potentialities of both parents. Half the possibilities of each have been thrown away, and hence cannot appear in the offspring. By this constant process we get, in every generation, new combinations of qualities. This is the main cause, says Weissman, for variations. There is, however, another possible cause. Each cell has enough determinants in it for many individuals, and it seems to be more or less a matter of accident which qualities shall come out. It has been suggested that as an egg lies within the gland, a blood vessel may bring blood to it in such way that a determinant, lying in a certain position in the egg, may get the richest supply of blood, and hence develop at the expense of the less nourished determinant. By these two methods variation comes into an animal's life, if Weissman and his school are to be believed.