SCIENCE. 



[Vol. XII. No. 306 



trade in musical boxes, but a visit to the establishment of M. J. 

 Paillard will convince the most sceptical that automatic musical in- 



struments play an important part in satisfying the musical demands 

 of the public. 



MENTAL SCIENCE. 

 The Illusions of Drawing and Painting.' 



The arts of drawing and painting depend upon the possibility of 

 presenting to the eye a result in two dimensions of space which it 

 will readily transform into one of three dimensions. In this pro- 

 cess there is a large element of illusion — of conscious, designed 

 illusion. The chief factor in this process is perspective. If the 

 spectator take, whether in imagination or reality, the position of the 

 artist when making the design, the image on his retina will be the 

 same as that on the artist's retina, and the design will be recog- 

 nized as the counterpart of the reality, provided the spectator knows 

 in general the nature of the object represented. If the facsimile is 

 to be more exact, color must be imitated, light and shade intro- 

 duced, and the retinal effect copied with all the skill of eye and 

 hand. In order to have an Esthetic effect, the picture must repre- 

 sent known objects : the interpretation of two-dimensional objects 

 into three-dimensional must be rendered easy by the knowledge of 

 the three-dimensional. The artist must not create entirely new 

 forms : exceptions are apparent only, and prove the rule. The poetic 

 monsters are either conventionalized, or unite incongruous but ex- 

 isting forms, — half man and half beast. This is especially neces- 

 sary when the object of the drawing is purely intellectual, — to 

 make clear something not easily expressed in words, such as de- 

 signs for houses, mechanical constructions, and the like. Here a 

 more or less exact knowledge of the type of object represented is 

 needed. To the layman such designs have little meaning. 



In artistic painting, however, it is not the most detailed and ex- 

 act drawing that produces the best result. Photography excels all 

 manual art in this, but its effect is of a lower order. The same 

 can be said of those clever productions by which a bas-relief ap- 

 pears drawn in two dimensions, or the objects of a panorama to 

 stand out in three. One admires the skill, but it is a curiosity 

 rather than a piece of art. But the object of art is not servile imi- 

 tation, not to give the spectator an absolute illusion, but to arouse 

 certain feelings, certain thoughts ; and those details must be chosen 

 that bring to mind the appropriate sentiments. 



The spectator of a painting never loses entirely the sense of view- 

 ing a painted surface : for (i) the drawing is strictly accurate only 

 for one point of view ; every change of position vitiates the per- 

 spective ; (2) the phenomena of binocular vision prevent the illusion 

 (the points of the canvas are seen at the real distance of the eye 

 from the canvas, and not at the various distances required by the 

 perspective ; while, furthermore, the real object would form differ- 

 ent images on the two retins, and the painting gives two nearly 

 alike) ; (3) even in viewing objects monocularly, we get impressions 

 of distance, for the eye constantly moves, while these changes are 

 quite different in viewing a painting with one eye (the illusion of a 

 painting is no doubt increased by regarding it monocularly through 

 a hollow tube); (4) color and light can be imitated, but their men- 

 tal effect is recognizably different from that of the real objects. 



A picture placed in a horizontal position produces the illusion 

 nearly as well as in a vertical position. If it be a marine view, the 

 water does not seem vertical in the former case, though in the lat- 

 ter it seems horizontal. If it be an architectural design, it is not 

 displaced, any more than we confuse directions when we gaze at 



' By M. J. L. Soret, in Revue Scientifique, Nov. 3, 1888. 



an object in a reclining position. This is the result of much prac- 

 tice in seeing the form of representations irrespective of their 

 position, and in transforming the actual retinal image into the 

 one that the artist intends. If you dispense with all light and 

 shade, with all color, with all perspective, and leave simply a bare 

 outline, then we can see in such an outline all the various designs 

 which it can physically represent. If you draw one square within 

 another and join the corners, you can see such a figure either as the 

 description just given, or as the picture of a shallow trough looking 

 into the bottom, or as a view of the same object from the bottom ; 

 and so on. Light and shade, familiarity with the design, decide 

 what we shall see. This does not mean that the artist may neglect 

 perspective, but only that the object of the perspective is to make 

 easy the mental apperception of the spectator. Cases occur in 

 which a painter violates the rules of perspective, if by following 

 them he would produce a scientifically accurate but apparently un- 

 natural result. 



In the perception of distance the objects touching the lower edge 

 of the canvas are, as a rule, meant to be seen as in the plane of the 

 canvas. This gives the spectator his point of view, while the 

 framing of the picture by supplying a vertical and a horizontal, aids 

 very materially his conception of position. If in a landscape we 

 have the ground touching the lower end of the canvas, and the sky the 

 upper, we can judge distances best. If a prominent object is cut 

 at the edge of the canvas, it increases the difficulty of distance per- 

 ception. Of course, the size of the painted objects need bear no- 

 approximation to the actual size. Our eye is trained to perceive 

 form relations independently of size ; and, if the real size of the ob- 

 ject is familiar, we involuntarily suppose a more distant point of 

 view. So, again, we generally underestimate the size of colossal 

 figures, because we allow too much for our distance from them. 



A more complete proof that imitation is not the artist's chief aim- 

 is that he attempts to represent motion in a single view, which 

 physically is impossible. When a tree is represented in a wind, its 

 branches are shown bent and strained in the direction of the wind ;. 

 and this gives us at once the picture of a wind, of motion. So in a 

 figure the attitude characteristic of a series of motions stands for 

 the motion itself. It is not so much the fidelity as the suggestiveness 

 of the attitude that is important. So, again, when objects move 

 very rapidly, they become indistinct to our vision, and by pa:inting 

 them as indistinct the illusion of rapid motion is aided. If the mo- 

 tion is too rapid for the eye to follow, as in the rotation of the spokes- 

 of a carriage-wheel, the peculiar appearance can be imitated on 

 canvas, and suggests extreme speed. 



In the walk or run of an animal, although one position follows 

 another with great rapidity, the eye selects certain positions as typ- 

 ical, and these the artist uses as the presentation of movements 

 Generally the position at the beginning or the end of a step is 

 chosen. Instantaneous photography shows the great variety of posi- 

 tions in passing from one step to another ; but many of these have 

 an unnatural appearance to the eye, and the artist cannot utilize 

 them. 



A very distinctive illusion is shown in many portraits in which 

 the eyes seem to follow the eyes of the spectator. This occurs 

 when the model's eyes are facing the artist's. We assume the po- 

 sition of the artist, and so have the eyes in the picture looking at 

 ours. If we move to one side, we get the illusion of the portrait's 

 turning about, because the eyes still suggest direct vision, and 

 the rest of the pose does not strongly contradict it. This lateral 

 displacement, brought about by a change of position, is very- 

 slight in a painting, while very marked in a three-dimensional ob- 

 ject. Paintings of animals frequently show similar effects. The 

 true artist must understand and utilize such illusions, for the)^ 

 make the difference between what is lifelike and what is artificial. 



The Homing Instinct. — Dr. George M. Gould {Progress^ 

 October, 1888) has collected authentic cases of animals finding 

 their way homeward over long distances. Dogs, even when car- 

 ried away m a blindfolded or drugged condition, find their way 

 home over distances from five to five hundred miles ; and in one- 

 case, when the dog was taken off along the two sides of a triangle, he- 

 came home by the third side. The exquisitely trained instinct of 

 the flying pigeon, and similar capabilities of most animals, show the 



