(_a_) If however knowledge of this kind is really to claim the rank of _connoisseurship_ of the first class it must be both varied in its character and of the widest range. It is an indispensable condition to such that it should possess an accurate knowledge of the well-nigh limitless field of particular works of art both of ancient and modern times, some of which have already disappeared, while others are only to be found in distant countries or portions of the globe, and which it is the misfortune of our situation to be unable to inspect. Add to this that every work of art belongs to _one_ age, _one_ nationality, and depends upon particular historical or other ends and ideas. On account of this it is indispensable that the finest type of art-scholarship should have at its command not merely historical knowledge of a wide range, but knowledge that is highly specialized. In other words, a work of art is associated with particular[22] detail in a peculiar sense, and a specific treatment, is imperative to the comprehension and interpretation of it. And in conclusion this connoisseurship of the finest class does not merely imply like every other a retentive memory, but also a keen imaginative sense, in order to hold clearly before the mind the images of such artistic representations in all their characteristic lines, and above all, to have them ready for comparison with other works of art. (_b_) Within the limits of such a method of study which is primarily historical[23], distinct points of view will soon assert themselves which in the contemplation of such works we are not suffered to lose sight of, inasmuch as they are indispensable to a critical verdict. Such points of view, as is the case with other sciences the commencement of which is empirical, are summarized, after their due collection as separate units and comparison, in general criteria and propositions, emerging in a yet further stage of formal generalization in "_Theories_ of the arts." This is not the place to dwell at length upon literature of this kind; we will merely recall a few specimens of such work in the most general way. There is, for instance, the "Poetics" of Aristotle, which contains a theory of tragedy still of real interest. With still more pertinency among the ancients the "Ars Poetica" of Horace and the Essay on the "Sublime" by Longinus will exemplify generally the manner in which this type of theorizing is carried out. The general theses which are therein formulated are intended to stand as premises and rules, in accordance with which works of art ought to be produced, their necessity being above all insisted on in times of the decadence of poetry and art. They are, in short, prescriptions to the practitioner. The prescriptions, however, of these physicians of art were even less successful in their curative effect on art than are the ordinary ones in the restoration of bodily health.