Let T stand for the tall form and D for the dwarf form. Since the offspring are composed of both the paternal and maternal gamete, we may represent them as TD. But dwarfness is, as we have seen, recessive, so that the offspring all look as though they were pure T’s. When, however, we come to breed these TD’s _inter se_, the gamete or sex-cell of each individual crossed breaks up into its component parts T and D, which unite with other free T or D units to form TD’s or TT’s or DD’s. What are the possible combinations? A D of one parent may meet and unite with a D of the other parent, so that the resulting cells will be pure D, _i.e._ DD, and will give rise to pure dwarf offspring. Or the D gamete from one parent may unite with a T gamete from the other parent, and the result will be a TD cross, but this, as we have seen, will grow up to look like a pure T, _i.e._ will become a tall organism. Similarly, a T gamete from one parent may unite with a T gamete of the other, and produce a pure tall form, or it may unite with a D and produce a hybrid TD, which gives rise to a tall form. Thus the possible combinations of offspring are DD, DT, TD, TT, but all these three last contain the dominant T gamete, and so develop into tall offspring; therefore, _ex hypothesi_, we shall have three tall forms produced to one dwarf form, but of these three tall forms two are not pure, and do not breed true. Mendel’s experimental results accorded with what we should expect to obtain if the above explanation were correct. Hence the inference that there is such a splitting of the gametes in the sexual act seems a legitimate one. Mendel’s experiments are of great importance, for they give us some insight into the nature of the sexual act. But, as is usual in such cases, Mendel’s disciples have greatly exaggerated the value and importance of his work. It is necessary to bear in mind that Mendel’s results apply only to a limited number of cases—to what we may call balanced characters. In the case of characters which do not balance one another, which are, so to speak, not diametrically opposed to one another, Mendel’s law does not hold. A second important point is, that the dominance is in many cases not nearly so complete as it should be if the Mendelian formula correctly represented what actually occurs in nature. Further, the segregation of the gametes does not appear to be so complete as the above hypothesis requires it to be. The phenomena of inheritance seem to be far more complex than the thorough-going Mendelian would have us believe. Let it be noted that it is not to the facts of Mendelism, but to some portions of what we may call the Mendelian theory, that we take exception. Maturation of the Germ-cells