The student may feel that these are unusual conditions; that few houses would be of just the proportion of two cubes, - and this is of course true. It is not a difficult matter, however, when a cube has been drawn as a unit, to add one or several more in any direction, or portions of one. If the house just considered was to be thirty feet long, for instance, instead of forty, the second cube could be easily cut in half, the correct perspective distance being judged by the eye, or the diagonals of its nearest face could be crossed which would give the correct point of intersection for the cut.

Once the main proportions have been established the doors and windows, roof overhangs, etc., can be added and the whole completed. Experience will show many uses of diagonal lines in locating centers and measuring distances, and other short cuts which will prove a saving of time and an aid to accuracy.

Sometimes it is desired to show buildings entirely above the eye, as on a high hill or mountain, and again it is a part of the problem to represent them below the eye. Sketch "2" illustrates these conditions in a simple way.

Now whether buildings are above or below the eye or at its level and whether simple or complex, the same general principles hold. But when a building is complicated in its masses, or irregular in plan, it is usually best to think of it as inclosed within a more simple mass, drawing this mass first, and then subdividing it into the smaller parts. Sketch "3" was designed to illustrate this thought, the dash lines showing the simple mass which was drawn first.

When the larger proportions of a building are established there are many details to be added and Sketch "4" pictures a few typical ones in a very meager way. Many towers are based on pyramids and cones such as those shown at "A" and "B." One should practice these, then, and should try his hand at steps, chimneys, arches, dormers, etc., until he feels able to sketch any of the more commonly seen details easily and well, either from the objects themselves (which is excellent practice) or from memory.

Sketch "5" is to show that when furniture is to be represented it is often well to first block it in very simply so far as mass is concerned, just as we did the building in Sketch 3. For the chair at "A" two cubes were drawn as shown by the dotted lines, and the seat below was sketched within a square prism. When objects are thus inclosed within simple forms or "frozen into a block of ice" one is less likely to get them incorrect in perspective. It is suggested that as a means of adding to one's ability to do this well lie cut out prints of buildings and pieces of furniture and sketch simple shapes around them with a few lines, preferably straight, for this will help one to realize that all objects are comparatively simple in basic form.

Figure 10. Illustrating the Application of Perspective Principles.

Photographs or prints can help us in another way in the study of perspective, for we can lay a ruler on them or a T-square or triangle and produce with a pencil the various series of parallel lines to or towards their vanishing points, locating and draw-ins; the eye level or horizon line first of all. This will help one to understand the perspective phenomena more quickly, perhaps, than any other one thing.

We should not close without some reference to the perspective of interiors, though a brief word will suffice, for it is hardly necessary to say more than that interiors are done in just the same way as are exteriors, only we are looking at the inside of the cubes and prisms instead of at the outside, which means that we simply remove those faces which are the nearest to us. Rooms themselves are usually very simple as to form; it is in the furniture, turned at various angles and of irregular shape, that one encounters the greatest difficulty. A little practice, however, will give one considerable proficiency in all of this work.

One should be cautioned, nevertheless, that the difficulties are not few, for whether one is drawing interiors or exteriors or small objects it must be borne in mind that in theory he is supposed to look in the same fixed direction constantly until the draw-ing is finished, and although in practice this is not especially hard to do when an object is so small as to come entirely within the range of vision without the gaze being shifted, when it comes to large objects or entire rooms or buildings we are so accustomed to glancing about from place to place that it is not easy to keep from making a sort of composite sketch in which the various small parts may be correct in themselves, but wrong when considered in relation to one another and to the whole. When drawing a room, for instance, it is easy to go astray by looking first at a window and drawing that, and next doing a door, and so on, one thing at a time.

When this method is followed the whole is quite sure to look distorted. For this reason one should locate a horizon line on the drawing whenever possible and if vanishing points would naturally come within the paper area find them also, and in sketching the main lines try to give them the right proportion and perspective convergence, for if a sort of framework can be correctly built up for the whole it will not be hard to add the detail; therefore spend plenty of time on this first work. If too much trouble is encountered when drawing from actual buildings sketch from photographs for a while as this will be much easier to do. Then go on to portions of interiors and exteriors before attempting them in an entirety.

The excellent drawing by Mr. Watson on page 33 shows a type of subject which would prove extremely difficult to block out because of the great number of converging lines, unless one was familiar with the perspective principles involved; and if a subject of this nature is not correctly constructed the errors will usually be glaringly apparent, regardless of the quality of the technique.

Pencil Sketch By Ernest W. Watson Pennsylvania Station, Pittsburgh, Pa.