Professor Mach proceeds to explain the phenomenon as follows:
Fig. 86.
"Moving objects exert, as is well known, a peculiar motor stimulation upon the eye, they draw our attention and our look after them. If the look really follows them ... we assume that they move. But if the eye, instead of following the moving objects, remains steadfastly at rest, it must be that the constant stimulus to motion which it receives is neutralized by an equally constant current of innervation flowing into its motor apparatus. But this is just what would happen if the steadfastly fixated point were itself moving uniformly in the other direction, and we were following it with our eyes. When this comes about, whatever motionless things are looked at must appear in motion."[455]
The knot x, the string, we ourselves, and all our stationary surroundings thus appear in movement, according to Mach, because we are constantly innervating our eyeballs to resist the drag exerted upon them by the pattern or the flowing waves. I have myself repeated the observation many times above flowing streams, but have never succeeded in getting the full illusion as described by Mach. I gain a sense of the movement of the bridge and of my own body, but the river never seems absolutely to stop: it still moves in one direction, whilst I float away in the other. But, be the illusion partial or complete, a different explanation of it from Professor Mach's seems to me the more natural one to adopt. The illusion is said to cease when, our attention being fully fixed on the moving oil-cloth, we perceive the latter for what it is; and to recommence, on the contrary, when we perceive the oil-cloth as a vaguely moving background behind an object which we directly fixate and whose position with regard to our own body is unchanged. This, however, is the sort of consciousness which we have whenever we are ourselves borne in a vehicle, on horseback, or in a boat. As we and our belongings go one way, the whole background goes the other. I should rather, therefore, explain Professor Mach's illusion as similar to the illusion at railroad-stations described above on [page 90]. The other train moves, but it makes ours seem to move, because, filling the window as it does, it stands for the time being as the total background. So here, the water or oil-cloth stands for us as background überhaupt whenever we seem to ourselves to be moving over it. The relative motion felt by the retina is assigned to that one of its components which we look at more in itself and less as a mere repoussoir. This may be the knot above the oil-cloth or the bridge beneath our feet, or it may be, on the other hand, the oil-cloth's pattern or the surface of the swirling stream. Similar changes may be produced in the apparent motion of the moon and the clouds through which it shines, by similarly altering the attention. Such alterations, however, in our conception of which part of the visual field is substantive object and which part background, seem to have no connection with feelings of innervation. I cannot, therefore, regard the observation of Prof. Mach as any proof that the latter feelings exist.[456]
The circumstantial evidence for the feeling of innervation thus seems to break down like the introspective evidence. But not only can we rebut experiments intended to prove it, we can also adduce experiments which disprove it. A person who moves a limb voluntarily must innervate it in any case, and if he feels the innervation he ought to be able to use the feeling to define what his limb is about, even though the limb itself were anæsthetic. If, however, the limb be totally anæsthetic, it turns out that he does not know at all how much work it performs in its contraction—in other words, he has no perception of the amount of innervation which he exerts. A patient examined by Messrs. Gley and Marillier beautifully showed this. His entire arms, and his trunk down to the navel, were insensible both superficially and deeply, but his arms were not paralyzed:
"We take three stone bottles—two of them are empty and weigh each 350 grams; the third is full of mercury and weighs 1850 grams. We ask L... to estimate their weight and tell us which is heaviest. He declares that he finds them all three alike. With many days of interval we made two series of six experiments each. The result was always the same. The experiment, it need hardly be said, was arranged in such wise that he could be informed neither by sight nor by hearing. He even declared, holding in his hand the bottleful of mercury, that he found it to have no weight.... We place successively in his hand (his eyes being still bandaged) a piece of modelling wax, a stick of hard wood, a thick India-rubber tube, a newspaper folded up lengthwise and rumpled, and we make him squeeze these several objects. He feels no difference of resistance and does not even perceive that anything is in his hand."[457]
M. Gley in another place[458] quotes experiments by Dr. Bloch which prove that the sense which we have of our limbs' position owes absolutely nothing to the feeling of innervation put forth. Dr. Bloch stood opposite the angle of a screen whose sides made an angle of about 90º, and tried to place his hands symmetrically, or so that both should fall on corresponding spots of the two screen-sides, which were marked with squares for the purpose. The average error being noted, one hand was then passively carried by an assistant to a spot on its screen-side, and the other actively sought the corresponding spot on the opposite side. The accuracy of the correspondence proved to be as great as when both arms were innervated voluntarily, showing that the consciousness of innervation in the first of the two experiments added nothing to the sense of the limbs' position. Dr. Bloch then tried, pressing a certain number of pages of a book between the thumb and forefinger of one hand, to press an equal number between the same fingers of the other hand. He did this just as well when the fingers in question were drawn apart by India-rubber bands as when they were uninterfered with, showing that the physiologically much greater innervation-current required in the former case had no effect upon the consciousness of the movement made, so far as its spatial character at any rate was concerned.[459]