If the correctness of the conclusions so far stated may be assumed, the necessity is at once evident for a whole series of anatomical, physiological, histological and histo-chemical investigations into those differences between male and female types, in the structure and function of the individual organs by which the dowers of arrhenoplasm and thelyplasm express themselves in the tissues. The knowledge we possess at the present time on these matters comes from the study of averages, but averages fail to satisfy the modern statistician, and their scientific value is very small. Investigations into the sex-differences in the weight of the brain, for instance, have so far proved very little, probably because no care was taken to choose typical conditions, the assignment of sex being dependent on baptismal certificates or on superficial glances at the outward appearance. As if every “John” or “Mary” were representative of their sexes because they had been dubbed “male” and “female!” It would have been well, even if exact physiological data were thought unnecessary, at least to make certain as to a few facts as to the general condition of the body, which might serve as guides to the male or female condition, such as, for instance, the distance between the great trochanters, the iliac spines, and so forth, for a sexual harmony in the different parts of the body is certainly more common than great sexual divergence. This source of error, the careless acceptance of sexually intermediate forms as representative subjects for measurement, has maimed other investigations and seriously retarded the attainment of genuine and useful results. Those, for instance, who wish to speculate about the cause of the superfluity of male births have to reckon with this source of error. In a special way this carelessness will revenge itself on those who are investigating the ultimate causes that determine sex. Until the exact degree of maleness or femaleness of all the living individuals of the group on which he is working can be determined, the investigator will have reason to distrust both his methods and his hypotheses. If he classify sexually intermediate forms, for instance, according to their external appearance, as has been done hitherto, he will come across cases which fuller investigation would show to be on the wrong side of his results, whilst other instances, apparently on the wrong side, would right themselves. Without the conception of an ideal male and an ideal female, he lacks a standard according to which to estimate his real cases, and he gropes forward to a superficial and doubtful conclusion. Maupas, for instance, who made experiments on the determination of sex in _Hydatina senta_, a Rotifer, found that there was always an experimental error of from three to five per cent. At low temperatures the production of females was expected, but always about the above proportion of males appeared; so also at the higher