The point which most concerns us is in how far biological data can be applied to the sex problem in society.  Systematic dissections or breeding experiments upon human beings, thought out in advance and under control in a laboratory, are subject to obvious limitations.  Surgical operations, where careful data are kept, often answer the same purpose as concerns some details; but these alone would give us a fragmentary record of how a fertilized egg becomes a conscious human being of one sex or the other.  The practice of medicine often throws light on important points.  Observation of abnormal cases plays its part in adding to our knowledge.  Carefully compiled records of what does occur in inheritance, while lacking many of the checks of planned and controlled experiments, to some extent take the place of the systematic breeding possible with animals.  At best, however, the limitations in experimentation with human subjects would give us a rather disconnected record were it not for the data of experimental biology.

How may such biological material be safely used?  Indiscriminately employed, it is worse than useless—­it can be confusing or actually misleading.  It is probably never safe to say, or even to infer directly, that because of this or that animal structure or behaviour we should do thus and so in human society.  On this point sociology—­especially the sociology of sex—­must frankly admit its mistakes and break with much of its cherished past.

The social problem of sex consists of fitting the best possible institutions on to the biological foundation as we find it in the human species.  Hence all our reasoning about which institution or custom is preferable must refer directly to the human bodies which compose society.  We can use laboratory evidence about the bodies of other animals to help us in understanding the physical structure and functions of the human body; but we must stop trying to apply the sex-ways of birds, spiders or even cows (which are at least mammals) to human society, which is not made up of any of these.

It is possible to be quite sure that some facts carefully observed about mammals in a biological laboratory apply to similar structures in man, also a mammal.  Because of this relationship, the data from medicine and surgery are priceless.  Thus we are enabled to check up our systematic experimental knowledge of animals by an ascertained fact here and there in the human material, and to get a fairly exact idea of how great the correspondence actually is.  Gaps thus filled in are narrow enough, and our certainty of the ground on either side sufficiently great, to give a good deal of justifiable assurance.

If we use our general biological evidence in this way, merely to help in clearing up points about human biology, we need not be entirely limited to mammals.  Some sex phenomena are quite general, and may be drawn from the sexual species most convenient to study and control in experiments.  When we get away from mammalian forms, however, we must be very sure that the cases used for illustrations are of general application, are similar in respect to the points compared, or that any vital differences are understood and conscientiously pointed out.