Before we take up the main questions before us this evening, I must speak of one other form of heredity. In many instances we have evidence that a character is the product of more than a single _mutant_ gene. I say “mutant gene” because in fact every character is no doubt the product of the combined action of many genes, but in addition to this general relation there are many cases now known where there are several specific genes whose _chief_ effect is on one character. Size differences furnish abundant data of this sort. One of the clearest cases is that of the size of the ear of corn. Some races of corn have short ears (and cobs), some long. If two such races are crossed, the hybrid is intermediate with a considerable range of variation. If the hybrid is self-fertilized, the progeny in the next generation shows a still wider range of variation, extending from that of the shorter grandparent to that of the longer. Both grandparental cobs have reappeared, but also many intermediate grades, Fig. 5. [Illustration: FIG. 5. Cross between long- and short-eared corn. Samples of two original types shown in upper part of figure, hybrid offspring in the middle of figure, and samples of 2d generation in the lower part. (After East and Hays.)] Such cases were formerly spoken of as blended inheritance. It was supposed that the materials of the two parents have, as it were, fused in the offspring and have remained fused. Today we have a better explanation. It is this. Besides two major factors that here determine cob length, there are other _minor_ factors, some of which make the short cob longer, others that make the long cob shorter. These go over into the first generation hybrids, and are sorted out in the germ cells of the hybrid. Consequently, when the F₁’s are inbred, there are all sorts of recombinations of the minor factors. This explains the greater variability of the second generation. It is probable that in most of our domesticated animals, including man, much of the variability is due to multiple factors, which makes a study of inheritance in these groups extremely difficult, especially when, as in the case of man, the number of offspring from a pair is small, and critical combinations for study can not be made. If then it is highly improbable that any particular defective trait could ever become widely spread in the human germ-plasm, how does it come about that such defects as feeblemindedness and insanity are so widespread in the racial inheritance? There are several possibilities here to keep in mind, but I think we ought not to pretend that we can give a completely satisfactory account of the situation.