"Quite generally, then, there are two conceivable methods by which the sex might be artificially influenced in any particular case; firstly, if means could be found of ensuring that any particular fertilized ovum received the required chromosomes; and, secondly, by the discovery of methods which always effect the ovum or embryo in such a way as to produce the desired sex. Many suggestions for applying both methods have been made, some of which have attained considerable notoriety, but hitherto none of them has stood the test of practical experience. In the case of the higher animals, especially of the mammals, in which the embryo develops in the maternal uterus until long after the sex is irrevocably decided, it is obviously difficult to apply methods which might influence the sex after fertilization, even if it were certainly known that such methods were ever really effective. "Apart from the few experiments like those of Hertwig on rearing tadpoles at different temperatures, there have been a very few cases in which there is even a suggestion that the sex of the fertilized egg can be modified by environment, and the belief that this is possible has been entirely abandoned by many of the leading investigators of the subject. It is probable, therefore, that if it will ever be possible to predict or determine artificially the sex of a particular child, the means will have to be sought in some method of influencing the output of germ-cells in such a way that one kind is produced rather than the other. It is in this way that Heape and others interpret the results of their investigations; they find that certain conditions affect the sex-ratio of cells, and they explain the result by assuming that UNDER SOME CIRCUMSTANCES MALE-DETERMINING OVA ARE PRODUCED IN EXCESS, AND UNDER OTHER CIRCUMSTANCES, FEMALE-DETERMINING."