We would ask our readers to distinguish carefully between the facts we set forth, and the conclusions we draw therefrom. The former, being facts, must be accepted. The interpretations we suggest should be rigidly examined, we would say regarded with suspicion, and all possible objections raised. It is only by so doing that any advance in knowledge can be made. By inheritance we mean that which an organism receives from its parents and other ancestors—all the characteristics, whether apparent or dormant, it inherits or receives from its parents. Professor Thomson’s definition—“all the qualities or characters which have their initial seat, their physical basis, in the fertilised egg cell”—seems to cover all cases except those where eggs are parthenogenetically developed. The first fact of heredity which we must notice is that inheritance may take several forms. This is apparent from what was set forth in the chapter dealing with hybrids. Types of Crosses In considering the phenomena of inheritance it is convenient to deal with crosses in which the parents do not closely resemble one another, because by so doing we are able readily to follow the various characters displayed by each parent. It may, perhaps, be urged that such crosses occur but rarely in nature. This is true. But we should bear in mind that any theory of inheritance must explain the various facts of cross-breeding, so that, from the point of view of a theory of inheritance, crosses are as important as what we may term normal offspring. As inheritance is so much easier to observe in the former, it is but natural that we should begin with them. Our deductions must, if they be valid ones, fit all cases of ordinary inheritance, _i.e._ all cases where the offspring results from the union of parents which closely resemble one another. Now, when two unlike forms inter-breed, their offspring will fall into one of six classes. I. They may exactly resemble one parent, or rather the type of one parent, for, of course, they will never be exactly like either parent; they must of necessity display fluctuating variations. The cases in which the offspring exactly resemble one parent type in all respects are comparatively few. They occur only when the parents differ from one another in one, two, or at the most three characters. Thus when an ordinary grey mouse is crossed with a white mouse the offspring are all grey, that is to say, they resemble the grey parent type. Although they are mongrels or hybrids, they have all the appearance of pure grey mice. This is what is known as unilateral inheritance.