Table II gives in detail the data yielded by four of the most instructive films. C 3 is the longest record that was obtained; Tu 4 is among the shortest, though it is not the very shortest. H 2 and H 3 show how nearly alike are the simultaneous movements of the two eyes: .07 sec. is the greatest difference recorded on any film between simultaneous movements. All four records show how much less the duration of the slow movements is at the beginning of the record than at the end, and how little the fast movements vary in this respect.

H 2 is given because it is not typical; and about one half of the film itself is reproduced in Fig. 5 (Plate II). It will be seen that at four points there intervened between slow movements (toward the right) a rapid one that was also toward the right. This is the only record in which such a thing happened: and its explanation is problematical. With the subjects C and H, and only very rarely with these, a rapid movement sometimes took the place of a slow one, that is, occurred in the same direction as the slow movements (e. g., Table II, C 3). And a trifle more often, yet very seldom, a rapid movement was relatively slow (e. g., ibid.). With every subject there are a few cases in which the eyes stood still for a small part of a second (e. g., ibid.), and these moments of rest seem to come after a rapid or a slow movement indifferently.

McAllister1 and others have shown that the eyes are seldom at rest even when voluntary fixation is attempted, and these anomalies in the nystagmiform series may well be the result of such random factors, which instead of being always inhibited by the afferent impulses from the semicircular canals, which govern the nystagmus, operate along with these latter, and sometimes even inhibit them. With the exception of these anomalies, the movements recorded in the photographs confirm the observations of Purkinje, Mach, Breuer, Delage, and other investigators.

In conclusion, the sensations of vertigo and of nausea seem not to be essentially connected with the nystagmus. Several subjects were so disagreeably affected by a preliminary rotation that it seemed best not to continue the experiment with them. With those, however, whose eyes were photographed, while they experienced a mild degree of vertigo and nausea during and after the first few rotations, these sensations soon wore off with further practice, while so far as could be observed their eye-movements were as ample and 1 C. N. McAllister: Yale Psych. Studies, Psych. Rev. Mon. Supplements, vol. 7, no. 1, p. 17, 1905.

1

rapid as at first. The introspection of these subjects was that after the rotation the body seemed at rest and the stomach quite settled, while the visual field alone whirled rapidly in the direction opposite to that of the previous rotation.

BY E. B. HOLT

DURING and after a prolonged rotation of the head, the visual field seems to spin around before one's eyes, - a phenomenon that is ordinarily called the "dizziness of Purkinje." Delage describes it as follows: "In the experiment of Purkinje, while we are rotating in a positive sense, space seems possessed of a motion in the opposite direction. . . . This phenomenon is explained by the direction of the nystagmus."

"In the nystagmus," he continues, "the eyeballs execute two well-differentiated motions: one, a compensatory, relatively slow motion, during which images pass across the retina so as to give the appearance of a movement of space in the opposite direction; two, a swift motion opposite to the slow one, and so rapid that the images passing across the retina leave no sensation of their movement."

Now, in a previous paper 2 I have shown that there is a central anæsthesia, or central inhibition of visual sensations, during about the latter two thirds of the time occupied by every voluntary eyejump; and in view of this I was led to enquire whether in fact, as Delage so confidently asserts, it is the speed of these more rapid movements, or some other factor, that causes them to leave no visual sensations. There can be no doubt that they do leave none, since, aside from the statement of Delage, in dizziness the visual field whirls always in only one direction; whereas it should otherwise. appear to swing now to one side, now to the other, as the eyes move back and forth across the objects. I have found but one other mention of this point in the literature. In his Analyse, Mach

3

1 Yves Delage: Physiol. Studien über d. Orientirung (Aubert's transl.), p. 100, Tübingen, 1888.

2 E. B. Holt: Harvard Psych. Studies, Psych. Rev. Mon. Supplements, vol. 4,

p. 1, 1903.

E. Mach: Analyse der Empfindungen, 2d ed., p. 98, Jena, 1900.

says, parenthetically, "(the jerky eye-movement leaves no optical impression)"; but he does not suggest that this is because of its greater speed.

In order to test this point, a 2 c. p. incandescent lamp was so arranged that it could be moved vertically in front of, and about four metres distant from, a rotating chair. Since after a rotation the eyes are oscillating from side to side, if the lamp is moved up and down an obliquely inclined after-image streak must be generated on the retina; and clearly there are four possible positions in which this may lie, as shown in Fig. 1.

The results were absolutely uniform (the author alone as subject); the after-image streak always lay on that side of the moving light toward which the slow eye-movements were directed, that is, the lamp appeared to drift obliquely up or down and in a lateral direction opposite to that of the slow eye-movements. Apart from its vertical displacement, then, the lamp behaved like the less intensely illumined parts of the visual field, seeming to be totally invisible during the swifter eye-movements. Now since the experiment was done in a partially darkened room and the eyes were partly adapted to darkness, the lamp should have been intense. enough adequately to stimulate the retina even during the more rapid movements, and might be expected to leave an after-image streak on that side toward which these rapid movements were directed, and differing only from the streaks seen during the slow eye-movements in being inclined at a less angle from the horizontal. Yet no such streaks were visible.

These observations were made at about the same number of seconds after the rotation stopped, as the photographs were taken that are recorded in the preceding paper of this volume. The rapid movements were therefore about one sixth as long in duration as the slower ones. Since the respective amplitudes of rapid and slow must average very nearly the same, the rapid movements must have been about six times as swift as the slow movements. It needs therefore to be shown beyond a doubt that the 2 c. p. lamp was bright enough, in view of the briefness of stimulation of any one retinal element during the rapid eye-movement, to be above the threshold of perception. For this reason the experiment was not continued with other subjects.

The certainly adequate degree of illumination was realized during the photography of the eyes described in the preceding paper. Here during the post-rotary dizziness an arc lamp (of 6 amp.) was in