[Editor's note: These excerpts from Elie de Cyon's book represent, to our knowledge, the first emergence of the role of the neck muscles and of the oculomotor muscles in postural control]

(...) When I started my experiments on the motor deficiencies that, according to Flourens' data, generally occur following lesions of the semi-circular canals, in the head muscles as well as in the body ones, I felt it necessary to begin by determining the nature of those motor deficiencies: co-ordination disorders or innervation disorders? Following that line of thought, I first had to answer the question concerning the measure in which an abnormal attitude of the head is likely to disturb the animal's sense of balance [Editor's note: our emphasis] and induce motor anomalies.
     We already had, on that subject, Longet's very interesting experiments dating back to the forties, but which have since almost fallen into oblivion. Longet showed that the occurring motor deficiencies are not linked in any way to the discharge of the cerebro-spinal liquid, but constitute a secondary phenomenon, consecutive to the cutting of the nape muscles during the operation. And indeed, whenever Longet merely severed those nape muscles, without opening the vertebral canal, he immediately observed the same phenomena. Furthermore, those phenomena were completely absent whenever he let the cerebro-spinal liquid discharge through a small opening made at the level of the occipital ligaments without causing more or less significant lesions to the nape muscles. Longet explained his experiments by claiming that the unusual attitude of the head following the cutting of the nape muscles had the immediate consequence of a loss of the sense of balance, but that such loss would be the cause of the ensuing motor deficiencies. To achieve the same result, it is not even necessary to sever all nape muscles. The cutting of the posterior right muscles is amply enough to make the animal's gait insecure and unsteady.
     When I reproduced Longet's experiments, I easily observed the phenomena he had described. Right after the cutting of the right muscles of the head, the big and the small posterior ones, a sway on both sides appeared in most dogs. When forced to walk, they spreaded their legs, most of them walked very slowly, their head slightly bent down. Those animals put their paws on the floor extremely cautiously and always in order to spread the front paws as far as possible from one another. When running, which was very hard for them, the animals often fell, and they needed some efforts to rise to their paws again. Such phenomena generally disappeared after five or six days - the head, until then leaning on the chest by the chin, took back its normal attitude, and at the same time the walking movements became normal as well.
The fact observed by Longet, as well as the conclusions he drew from it, have been fully confirmed and the importance of a normal attitude of the head for keeping one's balance came into light. If the observed balance disorders and the motor deficiencies are not as important as after the cutting of the semi-circular canals, neither should we forget that the changes brought to the attitude of the head are far from reaching the same degree as in that last operation.
     A second series of experiments, initiated by the same line of thought, had consisted in artificially giving to pigeons, without causing any lesion of important parts, an attitude of the head similar to that we observed most of the time following the destruction of the semi-circular canals. That fairly complicated attitude is characterised by the fact that the beak is directed upwards and the occiput, on the contrary, downwards, usually towards the ground. The animals can easily be given that attitude of the head, by fixing the head to the sternal region with a few cutaneous stitches. The animals whose head is fixed that way behave exactly like those whose semi-circular canals, horizontal as well as vertical, have been destroyed - they cannot keep their balance, and as long as they are standing, they keep staggering on both legs and try to find a third point of support by leaning on their tail. But they mainly fail to and fall backwards, often over the head, around the transversal axis of the body. They also perform roundabout movements, mainly in the same only way. In short: we observe in them very acute disorders in the whole locomotion sphere. Once the stitches are removed and the head has a normal attitude again, all disorders immediately disappear and the locomotion goes back to normal. Those experiences therefore show very clearly how much a normal attitude of the head is important for the animal to be able to keep its balance and to perform rational movements.
     But during a shift of the head such as the one occurring in such experiments, our judgement on the origin of sound is not less wrong than that bearing on the position and the distance of objects seen. Now, the following experiment showed with all possible evidence that errors in visual perceptions, at least when they happen suddenly, can have the effects of an insecurity in the walk and disorders of the sense of balance. I fastened in front of the eyes of a pigeon a pair of prismatic glasses - the animal thus affected by an artificial strabismus then showed a series of motor deficiencies that present very clear analogies with the disorders observed following the cutting of the semi-circular canals. In some cases of such strabismus, the sway movements of the head correspond to those occurring after the cutting of the two horizontal semi-circular canals. In one of the pigeons affected by artificial strabismus, I observed, during the first moments, roundabout movements. Errors in visual and auditory perceptions therefore seem to be the most important of all those here considered.

[Editor's note: The following excerpt may well explain why Elie de Cyon's works have not been given the importance they deserved]

     As may be concluded from the preceding lines, my conception of the role of head movements in the physiological functioning of the semi-circular canals completely diverged from that of Goltz. Goltz saw the starting point of that functioning in the alleged shifts of the endolymph of the canals during the various head movements. Head movements would therefore act as real excitation means. In my conception, on the contrary, as I expressed it as early as 1873, those motor deficiencies would be the direct consequence of errors in visual and auditory perceptions, which are supposed, in normal conditions, to inform us on the situation of objects in the exterior space and on the position of our body in that space. Then if the attitudes of the head play any role in the functioning of the semi-circular canals is concerned, it is only as far as they help us avoid errors in those perceptions.
     Carrying on with my experiments, I developed that perspective even more and I supported it with the important observation that we can observe the most acute balance disorders in pigeons, in the absence of any sway movement of the head. I will deal with that question more thoroughly in the following paragraphs - let me just add here that many experiments carried out later by Bornhardt, Spamer, Ewald and I have also shown the inconsistency of Goltz's theory, concerning the currents of the endolymph.

[Editor's note: The following excerpt represents a first sketch of clinical study of the neck reflex, fifty years before the publications of Tadashi FUKUDA who, before 1980, did not know this work by Elie de CYON]

(...)      After a few preliminary researches, I finally gave preference to the following, very simple, method. A sheet of paper is carefully fixed on an exactly vertical board, at level of the head of the person submitted to the experiment. That person is standing, blindfold, and draws horizontal and vertical lines with a pencil and a ruler. Even if all the subjects were blindfold, all drawings have been executed in a completely dark room. The subject begins by setting the ruler in the direction he considers horizontal or vertical. We must take care that the ruler and the hand that holds it are taken off the paper as soon as a line has been drawn. The same goes for the right hand and the pencil. That way, we are sure that any new direction drawn is not under the influence, by mediation of the hands, of the line previously drawn.
For the reproduction of sagittal or transversal directions, we fixed the sheet on a table the surface of which was exactly horizontal - the subject was sat, his head and the superior part of his body upright. We will describe later (§ 6) the experiments performed in order to know if the straight lines thus drawn really correspond to the sagittal direction. A discussion on the mode of interpretation of the drawings obtained will also be given then. Those drawings allow us to measure exactly the changes that the imposed conditions of the experiments bring to the three factors we have just evoked. The mere aspect of the drawing already provides information on the way of the error in each direction.
     In order to measure that error, we just have to draw on the sheet of paper, after each experiment, the normal direction. The angles formed by the two verticals with the two horizontals, etc., generally correspond to the intensity of the error. Most of the time we know that intensity thanks to the mere aspect of the intersection point between the vertical and the horizontal lines, or the sagittal and transversal lines drawn in the dark - we just have to measure the width of the angle those lines respectively form with one another. However, the widths of the angles that are systematically indicated on the drawings must not be considered as the absolute measure of the intensity of the errors. The first experiments have notably shown that the intensity of the errors, in the same experimental conditions, is not necessarily proportional to the widths of the angles. It also happens that those widths do not always precisely indicate the way of the error. But even in the case when the way of the errors is identical and when their intensities are proportional to one another, the widths of the intersection angles cannot always help measure such intensities - the latter can stay equal to 90° or slightly under 90°, whereas the error has been very important. Those angles widths notably inform us on the relations existing between the errors in the various directions, that is to say on the most important factor and , for us, the most instructive. The study of the errors of direction in man is of the utmost interest, most of all because it can give us information on the mechanism of the formation of our representations of time and space. The fact to know in which way we make an error, under given circumstances, on one direction or another, is certainly already interesting in itself, but that does not tell us in which organ the sensations of direction are formed, or how our representation of a tridimensional space is created from the perception of the various directions.

[Editor's note: "Person C", whose test results de Cyon then presents, is no other than de Cyon himself, which accounts for certain incoherence between his results and those now described from statistical studies]

§ 4 - Errors of perceptions of the vertical and horizontal directions, during the rotations of the head around its sagittal axis.

     The errors of perception occurring in the oblique positions of the head, and more particularly when it is bent on the right or left shoulder, have already caught the scientists' attention.
(...) The study of the influences that the head rotations, and consequently that of the two labyrinths, have on our perceptions of direction, is of a particular interest. During the reproductions of Aubert's experiments by Nagel, Sachs and Meller, etc., it already appeared that the degree of inclination of the head on the right or left shoulder is likely to have an undeniable influence on the production of the error, as well as on the intensity of the apparent obliqueness of the vertical line in the dark (fig. 1).

§ 5 - Errors occurring during the rotations of the head around its vertical and horizontal axes.

     We will first reproduce a few figures showing the more frequent errors observed during the rotations of the head around its vertical axis. As we can see, the vertical lines but slightly deviate from the normal direction, at least no more than in the dark and with an upright position of the head - the deviation goes the same way for the two rotations, and that way coincides with that of the error which I generally make when drawing in darkness, even with my head upright. As I have just said, I usually make the same deviation from the vertical line, even if to a lesser degree, when drawing in the light.
     The horizontal lines suffer, mostly during the rotation of the head to the right, a more considerable deviation, but that deviation is still going, during the rotation of the head to the left, the same way as when the head is held upright. Figure 9 shows the deviations I make, when the body turns around the vertical axis at the same time as the head, that is to say in such a position that during the rotation to the left, the right part of the body, and during the rotation to the right, the left part of the body, is in front of the sheet of paper designed to be drawn on. As we can see, the deviations from the directions then go exactly the same way as when only the head turns.

FIG. 9 - Person C. Rotation of the head and of the body around the vertical axis. During the rotation to the right the deviation of the angle is of 10°, the rotation to the left of 2°.