even now certain generalizations summarising some of the ranges of human belief and conduct might already be formulated. Such an inquiry, however, is not the purpose of this essay, and these considerations have been advanced, in the first place, to suggest that theory indicates that the problem of sociology is not so hopelessly difficult as it at first appears, and secondly, as a justification for an examination of certain aspects of human conduct by the deductive method. The writer would contend that while that method is admittedly dangerous when used as a substitute for a kind of investigation in which deductive processes are reduced to a minimum, yet it has its special field of usefulness in cases where the significance of previously accumulated facts has been misinterpreted, or where the exacter methods have proved unavailing through the investigator having been without indications of precisely what facts were likely to be the most fruitful subject for measurement. This essay, then, will be an attempt to obtain by a deductive consideration of conduct some guidance for the application of those methods of {15} measurement and co-ordination of facts upon which all true science is based.
A very little consideration of the problem of conduct makes it plain that it is in the region of feeling, using the term in its broadest sense, that the key is to be sought. Feeling has relations to instinct as obvious and fundamental as are the analogies between intellectual processes and reflex action; it is with the consideration of instinct, therefore, that this paper must now be occupied.
II. P SYCHOLOGICAL A SPECTS OF I NSTINCT.
Many years ago, in a famous chapter of his Text Book of Psychology, William James analysed and established with a quite final delicacy and precision the way in which instinct appears to introspection. He showed that the impulse of an instinct reveals itself as an axiomatically obvious proposition, as something which is so clearly “sense” that any idea of discussing its basis is foolish or wicked.1