Besides cases such as these where a simple form of Mendelian inheritance is obviously indicated, there are others which are more difficult to read. Of some it may be said that on the whole the peculiarity behaves as though it were an ordinary dominant; but that exceptions occur in which affected children are born to unaffected parents. It is not impossible that the condition may, like colour in the sweet pea, depend upon the presence or absence of more than one factor. In none of these cases, however, are the data sufficient for determining with certainty whether this is so or not. A group of cases of exceptional interest is that in which the incidence of disease is largely, if not absolutely, restricted to one sex, and so far as is hitherto known the burden is invariably borne by the male. In the inheritance of colour-blindness (p. 117) we have already discussed an instance in which the defect is rare, though not {176} unknown, in the female. Sex-limited inheritance of a similar nature is known for one or two ocular defects, and for several diseases of the nervous system. In the peculiarly male disease known as haemophilia the blood refuses to clot when shed, and there is nothing to prevent great loss from even a superficial scratch. In its general trend the inheritance of haemophilia is not unlike that of horns among sheep, and it is possible that we are here again dealing with a character which is dominant in one sex and recessive in the other. But the evidence so far collected points to a difference somewhere, for in haemophilic families the affected males, instead of being equal in number to the unaffected, show a considerable preponderance. The unfortunate nature of the defect, however, forces us to rely for our interpretation almost entirely upon the families produced by the unaffected females who can transmit it. Our knowledge of the offspring of "bleeding" males is as yet far too scanty, and until it is improved, or until we can find some parallel case in animals or plants, the precise scheme of inheritance for haemophilia must remain undecided.