In a graphic representation of objects one of two points of view may be taken: it may be considered as essential that all the characteristic features be shown, or the object may be drawn as it appears at any given moment. In the former case our attention is directed primarily towards those permanent traits that are most striking and by which we recognize the object, while others that are not characteristic, or at least less characteristic, are considered as irrelevant. In the latter case we are interested solely in the visual picture that we receive at any given moment, and the salient features of which attract our attention. This method is more realistic than the other only if we claim that the essence of realism is the reproduction of a single momentary visual image and if the selection of what appears a salient feature to us is given a paramount value. In sculpture or modelling in the round these problems do not appear in the same form. Here also attention may be directed primarily towards the representation of the essential, and the same principles of selection may appear that are found in graphic art, but the arrangement of the parts does not offer the same difficulties that are always present in graphic representation. As soon as man is confronted with the problem of representing a three-dimensional object on a two-dimensional surface and showing in a single, permanent position an object that changes its visual appearance from time to time, he must make a choice between these two methods. It is easily intelligible that a profile view of an animal in which only one eye is seen and in which one whole side disappears may not satisfy as a realistic representation. The animal _has two_ eyes and _two_ sides. When it turns I see the other side; it exists and should be part of a satisfying picture. In a front view the animal appears foreshortened. The tail is invisible and so are the flanks; but the animal has tail and flanks and they ought to be there. We are confronted with the same problem in our representations of maps of the whole world. In a map on Mercator projection, or in our planiglobes, we distort the surface of the globe in such a way that all parts are visible. We are interested only in showing, in a manner as satisfactory as possible, the interrelations between the parts of the globe. We combine in one picture aspects that could never be seen at one glance. The same is true in orthogonal architectural drawings, particularly when two adjoining views taken at right angles to each other, are brought into contact, or in copies of designs in which the scenes or designs depicted on a cylinder, a vase, or a spherical pot are developed on a flat surface in order to show at a single glance the interrelations of the decorative forms. In drawings of objects for scientific study we may also sometime adopt a similar view-point, and in order to elucidate important relations, draw as though we were able to look around the