Such a procedure, however, is encumbered with difficulties, of which the writers on speculative æsthetic were scarcely aware. When theories of art and beauty were based on general _a priori_ principles, there could not possibly be any doubt as to the point of departure in the several researches. But when we have no assumptions to start from, the very demarcation of the subject may become a matter of uncertainty. In the philosophy of art, to which department of æsthetic I wish to restrict my researches in the present work, this difficulty of formulating the data and _quæsita_—the facts which we have to go upon, and the facts which we wish to find out—constitutes the first, and by no means the least important, problem. If we are to embark upon a scientific treatment of art without any preconceived definitions, the aim and conditions of such treatment can only be determined by examining the prevailing notions on the subject, as they are expressed in language and in literature. As an interpretation for general use and of general applicability, a theory of art can claim attention only if it conforms to the recognised usage of the principal æsthetic terms. In the various definitions of art which are contained in the different æsthetic systems, we must therefore try to find some point of unity from which to approach our subject. The difficulties of such a task are evident to any one who has gone through the discouraging experience of reading a history of æsthetic. The investigator who seeks an accurate demarcation of the whole area of art, as distinguished from other departments of life, meets with partial definitions which can be applied only to certain fixed forms of art. We need mention but a few of the most typical instances. Even an ardent admirer of Taine is compelled to admit that his generalisations are too exclusively derived from the study of poetry and the formative arts. In the same way it is only by laborious adjustments that the theory of Vischer can be applied to music and lyric poetry; the aphorisms of Ruskin do not even pretend to apply to any but the formative arts; and Mr. Marshall’s _Æsthetic Principles_—to adduce one of the most recent attempts in general art-theory—are too obviously those of an expert in architecture. In none of the modern systems has sufficient room been made for certain forms of art which, from the evolutionist’s standpoint, are of the highest importance: such as acting, dancing, and decoration. All the one-sided definitions are, moreover, so inconsistent with each other that it seems impossible to make up for their individual deficiencies by an eclectic combination. It is not to be wondered at, therefore, if some writers on art, confused by the bewildering contradictions of æsthetic theories, have called in question the very existence of any universal art-criterion.[4]