But although the aspects of autotelic artistic activity give us at once a datum and a problem on which we may confidently base our research, we must not overlook the peculiar difficulties that will necessarily arise from the exclusively psychological, non-historical character of this basis. A historic study of art shows us that the artistic activity proper can never be explained by examining concrete works as we meet them in reality. Whenever we have to deal with art as autotelic, the need of theoretical abstraction forces itself upon us with irresistible cogency. It is of no avail to argue from the data of art-history, because we can never fully know the mental origin of the works. The problem presented to us by the tendency to engage in artistic production and artistic enjoyment for their own sake can only be solved by studying the psychology both of artists and of their public. The “art-impulse” and the “art-sense,” as referring to subjective tendencies in creators and spectators, are the chief notions with which we have to operate in such an investigation. And when we are obliged to introduce the notion of the “work of art” we have to remember that this term, strictly speaking, refers to an abstract and ideal datum. Only by thus restricting our attention to the psychical facts can we attain any clear conception of that autotelic aspect of art on which so much stress has been laid in all æsthetic philosophy. It is needless to say, however, that even a purely philosophic interpretation of art would be impossible without a knowledge of its works and manifestations as they appear in real life, with all their extraneous, non-æsthetic elements. The psychological examination must therefore necessarily be supplemented by an historical one. The methods of the latter research cannot be the same as those used in a strictly æsthetic inquiry. And the words will naturally be employed in a different sense. We do not at that stage demand of a poem, a painting, or a drama, that it should fulfil more than the technical requirements of the several arts. The ornamentation of a vase, _e.g._ is in this sense a work of art even if it serves a magical, _i.e._ a supposed practical purpose. Indeed it is most advantageous, if we wish to bring out the influence of sociological factors with the greatest possible clearness, to concentrate our attention upon the very qualities which we have to disregard in the treatment of purely artistic activity. The productions of primitive tribes, in which art is so closely connected with life, supply the most profitable material for such a study. After having examined, in these simple forms, all the sociological aspects of art, it will be possible to place the two art-factors in the most illustrative antithesis and to study their mutual influence. From this it should be possible to suggest—although in this work no detailed attempt will be made to follow out the reasoning—why it is that the