La lecture en ligne est gratuite

N. Kleitmann
J. E. Schreider
X. - Diurnal variation in oculomotor performance
In: L'année psychologique. 1949 vol. 50. pp. 201-215.
1. In 18 subjects kept awake for more than 24 hours, on one or more occasions, there was a diurnal variation in oculomotor
performance, as judged by the blinking rate, binocular fixation, and lateral oscillations of the eyes.
2. Poorest performance occurred in the early hours of the morning, coinciding with the development of drowsiness.
3. Without intervening sleep, there was a spontaneous recovery of oculomotor performance, along with an increase in alertness,
later in the morning or afternoon.
4. An intermediate degree of oculomotor impairment manifested itself in poorer oscillations in post-fixation, as compared to pre-
fixation, trials, and in the appearance of diplopia toward the end of the five-minute fixation period.
5. In the execution of lateral sweeps, the templeward moving eye completed its excursion faster than did the nasalward moving
eye. This disparity also held in drowsiness, when the sweep was slowed in both eyes.
6. Amphetamine had no effect on oculomotor performance of subjects when the subjects were wide-awake,but ledto
improvement of impaired performance observed during drowsiness.
7. Small amounts of alcohol often imitated the impairment of oculomotor performance that occurred in drowsiness.
Citer ce document / Cite this document :
Kleitmann N., Schreider J. E. X. - Diurnal variation in oculomotor performance. In: L'année psychologique. 1949 vol. 50. pp.
doi : 10.3406/psy.1949.8443
by Nathaniel Kleitman and Jonas E. Schreider
Department of Physiology, University of Chicago.
Changes in activity and position of the eyes and eyelids have
always been associated with falling asleep and awakening.
Gradual drooping of the upper eyelids and the appearance of
a " " stare are sure signs of drowsiness, and a firm glassy
closure of the eyes usually precedes the onset of sleep. The
overt contraction of the orbicularis palpebrarum muscle occurs
at the time when most of the other striated muscles are relaxing.
During sleep changes in the tonus of the extrinsic eye muscles
lead to an upward and outward deviation of the eyes, indicating
that even the parallelism of the visual axes maintained for far
vision is perhaps an active convergence of the eyes from their
resting positions. This, and the additional effort required to
converge the eyes for near vision and for scanning the visual
field, involves an incessant and varying activity of the oculo
motor apparatus during the entire period of wakefulness, irre
spective of whether the other muscles of the body are overtly
contracting or not. The diplopia observed in very drowsy indi
viduals could thus be ascribed to oculomotor fatigue. The
popular association of double vision with alcoholic intoxication,
rather than with drowsiness, may be due to the fact that at the
usual going-to-bed hour oculomotor coordination is not impaired.
Miles (1) made a quantitative study of oculomotor performance
in drowsiness by photographically recording the movements
of the eyes when their gaze was shifted from one luminous dot
to another, through an arc of 40 degrees, in the horizontal plane. 202 PSYCHO-PHYSIOLOGIE
In the alert state such an oscillation was carried out in about
100 ms (milliseconds), but in extreme drowsiness the movement
required from 160 to 250 ms, and the sweep was often broken
up into several saccadic jerks. Drowsiness also led to poorer
fixation of the dots at the termination of the sweep, and the
usual corrective adjustments were either omitted, or grossly
In numerous studies on the effects of prolonged sleep depri
vation we noted a diurnal variation in the degree of drowsiness
overtaking our subjects (2). Extreme, almost overpowering
attacks of sleepiness usually occurred between 3 and 6 a. m.,
but later in the forenoon, and especially in the afternoon, the
subjects were much more alert. In the course of the succeeding
night, however, drowsiness appeared again and became progres
sively worse toward the critical early morning hours. While the
behavioral deterioration from loss of sleep was proportional to
the total duration of experimental insomnia, the waxing and
waning of drowsiness made it easier to keep the subjects awake
in the afternoon of third day than in the middle of the second
night. The purpose of this study was to find out if the diurnal
variation in alertness was associated with a parallel variation
in oculomotor performance.
Subjects, procedures, and methods. — Our observations were
made on 13 men and five women, most of them students in
their early twenties, all in good health at the time they were
kept awake for about 30 hours. Ten of these, six men and four
women, were studied only once, but the remaining subjects
went through two or more periods of sleep deprivation. Usually
four or five subjects were tested at one time, and our analysis
is based on 13 experiments and a total of 50 subject-nights.
The observations were made during the winter and spring, and
the subjects were indoors most of the time. They were not
permitted to lie down, nor to sit with their eyes closed, but
were not regimented otherwise with respect to activity.
The experimental procedure consisted of preliminary testing,
at irregular intervals, during the first day and evening, and
continuous surveillance and regularly spaced (every 2-3 hours)
testing during the night and the following forenoon, sometimes
also afternoon. When several individuals were studied simul
taneously, they were encouraged to engage in group activities,
such as conversation and card-playing. They were allowed to
eat, if they became hungry, but not to consume tea, coffee, or KLEITMAN, J. E. SCHREIDER. ■ OCULOMOTOR PERFORMANCE 203 N.
alcoholic beverages. In some cases body temperatures were taken
at two-hourly intervals throughout the period of observation.
Prior to each test, the subject was asked routinely to give
an estimate of his degree of alertness, and this was supplemented
by an objective description of his condition, furnished by the
test administrator. The quantitative evaluation of oculomotor
performance was based mainly on photographic records obtained
by means of a modified model of an Ophthalmograph, a portable
binocular eye-movement camera, manufactured by the American
Optical Company and designed to furnish information on the
characteristics and maturity of reading in school children (3).
The subjects is comfortably seated, with chin resting on a
adjustable support and forehead leaning lightly against a fixed
barrier, so that the position of the head with respect to the
camera is kept constant. Beams of light from two laterally
located lamps are reflected by the corresponding corneas into
adjustable telescope tubes that focus them as two tiny dots on
moving 35-mm photographic paper or film. As we used the
instrument, the subject was asked to look at a target card,
placed in a holder 33 cm away from the eyes, with a centrally cross and a 2-mm dot at each side edge. The distance
between the two dots was 13 cm, and, in shifting his gaze from
one dot to the other, the subject had to rotate his eyes through
an arc of 22.5 degrees. Records could be obtained of the position
of the eyes when the subject looked fixedly at the central cross
and of eye-motility, as the subject executed alternate sweeps
in looking at the dots.
In standard operation of the Ophthalmograph, only film is
used, and it is run by a synchronous motor at the rate of 13 mm
per second. We were able to reduce this speed ten times by
manually operating a crank attached to the feed-spindle. Fair
constancy of speed was obtained by turning the crank-handle
in synchrony with the second hand of an electric clock watched
by the operator. Another modification of the apparatus involved
the introduction of frosted glass screens in front of the lamps,
to be used when the speed of the film was reduced. In addition,
in some tests an electric timer was placed near the left-side
lamp. The revolving sectors of the timer interrupted the beam
of light reflected from the left cornea ten times per second, and
it was used as a check on the rate of movement of the film.
The oculomotor performance test involved continuous gazing
at the central cross for five minutes, preceded and followed by PSYCHO-PHYSIOLOGIE 204
a variable number of sweeps from one dot to the other, at a
rate of 70 per minute, timed by metronome clicks. During the
period of fixation, the operator counted and recorded the numb
er of blinking movements, as well as slower eyelid droopings,.
that occurred in each of the five minutes of observation. In the
early experiments we recorded either fixations, using Insurance
Bromide paper, or sweeps, on Panatomic-X film, but later we
were able to get both in the same test, by dimming the lamps
and greatly slowing the speed during the five minutes of fixation, using film throughout. The dimmed lights were also more
easily tolerated by the subjects. About ten minutes were requi
red to test one subject, and several subjects participating in
one experiment were usually tested consecutively in the same
order during the night.
The photographic record of binocular fixation consisted of
two lines which were continuous as long as the subject's eyes
remained open, and parallel, if balance was preserved.
A slight head movement resulted in a simultaneous and equal
displacement of both lines and could thus be distinguished from
a record of divergence of the eyes, which, because of the camera
effect, caused the lines to get closer to each other, with conver
gence leading to a greater separation of the lines. Successive one
minute portions of the fixation record were numerically graded
on the basis of the presence and degree of binocular imbalance
and the occurrence of eye closures. The oscillation records were
composed of the lateral sweeps of both eyes, short periods of
binocular adjustement upon reaching the target dots, and two
parallel fixation lines denoting the duration of the intersweep
pauses. The durations of the sweeps were determined for each
eye and type of movement (nasalward or templeward) by pro
jecting the film records, with enlargement, on a grid screen, so
that each division represented 10 ms. In addition, the records
were evaluated as to sharpness of the terminal adjustment and
the parallelism of the fixation lines.
Four subjects who, in previous sleep-deprivation experiments,
exhibited uncontrollable drowsiness, were retested while under
the influence of repetitive 10 mg doses of amphetamine (ben
zedrine). Other subjects were given single doses of
only after they became very drowsy. The effect of similar amounts
of amphetamine on oculomotor performance was also studied
in non-sleepy subjects under routine daytime conditions.
The influence of small amounts of alcohol (6 to 30 cc of 95 % KLEITMAN, J. E. SCHREIDER. OCULOMOTOR PERFORMANCE 205 N.
alcohol, diluted with four volumes of water or fruit juice) on
oculomotor performance was also investigated. Four subjects
were tested a total of 12 times. Three of these subjects were
observed while under the influence of both alcohol and amphet
amine, administered consecutively.
Results. — The subjective experiences and overt behavior of
persons experimentally deprived of sleep has been repeatedly
described and need not be gone into in detail. In the course of
the single night of wakefulness some of our subjects became
very sleepy, while others in the same group experiment remai
ned wide-awake. Also, a particular individual who had no dif
ficulty in keeping alert in one experiment, broke down complet
ely in another. Group activities and the mere presence of other
subjects seemed to delay the development of drowsiness by
three to five hours, or to prevent it altogether. No impairment
of oculomotor performance could be detected in those who did
not become drowsy throughout the period of sleep deprivation.
In the others, oculomotor deficiency was evaluated on the basis
of three criteria : rate of blinking, binocular balance in fixation
of five minutes' duration, and the execution of repetitive hori
zontal oscillations.
The number of blinks per minute varied greatly in different
subjects, and also in the same individual in successive experi
ments. Data gathered on nine men and two women in 24 expe
riments showed that in five subjects (including the two women)
the normal, or control, blinking rate was low, usually under 10,
in five it was higher, ranging up to 40 or 50, and in one it was low
in two experiments and higher in two others. Oculomotor dete
rioration manifested itself mainly in a decrease in the blinking
rate and the appearance of half-closures of the eyes, once to three
times per minute, and an increased flow of tears. The drop in
the frequency of blinking was naturally more pronounced when
the control rate was higher. Thus, in one subject the rate of
blinking fell from 42 at 4:30 a. m., when he was quite alert,
to 4 at 6:30 a. m., when he was very drowsy. However, some
subjects blinked at the same rate, whether alert or sleepy.
Recovery of alertness later in the morning was almost always
accompanied by an increase in the blinking rate (provided there
had been a decrease during drowsiness) to the control values.
In the course of the five minutes of observation the blinking
rate usually rose successively, except for a downward turn that
sometimes occurred in the fifth minute. Drowsiness did not change 206 PSYCHO-PHYSIOLOGIE
this development, but, in general, when the blinking rates were
low, the differences in successive minutes were smaller.
if XI
if t ••• I t
mmmmm ^ <m+m , ( mmmm n
i •
* :
t t
: i — .- |
t \, *
■1 i

A ' «
Fig- 1. — Successive five-minute fixation records obtained during sleep
deprivation in subject AR at 9 and 11 p. m., and 1, 3, 5, 7, 9, and 11 a. m_
Minutes delineated by broken horizontal lines.
Binocular fixation was evaluated in terms of parallelism and
continuity of the two lines that were made by the coalescence
of the dots reflected from the corneas on the photographic paper
or film. As shown in fig. 1, representing the fixations of one sub- KLEITMAN, J. E. SCHREIDER. - — OCULOMOTOR PERFORMANCE 207 N.
ject during a single experiment, a slight impairment appeared
in record 3, taken at 1 a. m., and became progressively greater.
Binocular fixation was at its worst in record 6, at 7 a. m., at which
time the subject was extremely drowsy. The complete disappea
rance of the lines in this record was due partly to divergence
of the eyes, but mainly to their semiclosures and the consequent
accumulation of tears in the narrowed palpebral fissures. For
evaluation purposes, record 6 would be termed oculomotor
breakdown. In the last two strips one can see a partial recovery
at 9 a. m. and further improvement at 11 a.m., both accompa
nying a gradual increase in alertness. It can also be seen that
when performance was deteriorating, impairment was greatest
in the terminal portions of the five-minute fixation record. As
a rule, the rate of recovery was greater than that of the dete
rioration that preceded it.
Of the several components of the records of lateral oscillations
of the eyes, the most rigorous evaluation was made of the dura
tion of the sweeps. A treatment of the records of ten subjects,
taken during 19 experiments, showed that normal, or control,
values were about the same after the five minutes of fixation
as they were before. Thus, the mean pre-fixation sweep times
varied from 85 to 95 ms in different tests, while the post-
fixation means were 83 to 98 ms. Drowsiness always led to a
slowing of the sweep movements and often to their decomposition
into several step-like saccadic jerks. In extreme drowsiness the
eyes executed wave-like sways, and that made it impossible to
determine the sweep-time with any accuracy. The five minutes
of fixation that separated the two oscillation trials usually led
to an accentuation of the oculomotor deficit. Whereas the mean
pre- fixation sweep times in drowsiness were 96 to 117 ms
" denotes the exclusion of sweeps that and up (the term " up
wrere too slow to be evaluated from the calculation of the mean),
after fixation they varied from 118 to 161 ms and up. There
was a definite association between the degree of drowsiness and
the slowing of the sweeps. With the return of alertness later in
the morning there was a concomitant decrease in the duration
of the sweeps toward the control values. Under these conditions,
the mean pre-fixation sweep times were found to vary from 97
to 99 ms, while the post-fixation figures were 107 to 112 ms —
a slight slowing of the sweep movements following fixation.
A sample of the effects of drowsiness on eye oscillations is
shown in fig. 2, all six records of which are of pre-fixation sweeps. 208 PSYCHO -PHYSIOLOGIE
The subject was not very drowsy till about 6:30 a. m., and
record 4 reveals a gradual onset of decomposition of the sweeps,
after the first two or three oscillations. The post-fixation record,
obtained at that hour, resembled record 5, which represents
Fig. 2. — Records of eye-oscillations, at the rate of 70 per minute, of sub
ject ES, during sleep deprivation, obtained at 11:30 p. m. and 1:30,
4:30, 6:30, 8:30, and 11:30 a. m. Although lateral sweep lines are too faint
to be seen in some of the records, they were clearly visible in the enlar
ged projections of the original films.
pre- fixation oscillations at 8:30 a. m., when the subject was
very drowsy. But even record 5 shows fair oscillatory move
ments at the beginning and end of the trial. The post-fixation
record at 8:30 a. m., however, consisted entirely of wavy move
ments. Record 6, at 11:30 a. m., wrhen the subject was no
longer sleepy, is close to normal again, but even at this hour KLEITMAN, J. E. SCHREIDER. OCULOMOTOR PERFORMANCE 209 N.
post-fixation performance was not as good as the pre-fixation one.
It is interesting to note that the two eyes did not move with
the same speed in carrying out the lateral sweeps. In one sub
ject, the right eye was always slower than the
left, but in all the others tested the duration
of the movement depended upon whether the
eye was nearer to or farther from the target
spot toward which the particular oscillation
was directed. The templeward sweep of the
nearer eye was made faster than the nasalward
sweep of the farther eye. The variations of
the mean values of the sweep-times of eight
subjects, tested 278 times in 31 experiments,
are given in Table I. The mean differences
between the durations of the nasalward and
templeward sweeps of either the right or the
left eye ranged, for different subjects, from a
few to over 50 ms. The last-mentioned dis
parity between the two sweep-times were so
marked that they could easily be seen without
magnification and are therefore reproduced in
fig. 3. It should be mentioned that the diff
erences between the durations of the nasalward
and templeward sweeps were also present in
drowsiness, and the mean values given in
Table I were based on all oscillation records
I'iir. .'{. - Record in which sweep durations could be determined I' iioniiiil (»rillii- i.)
with accuracy. For the group as a whole, the linii«. of sul)ji"-l
Al' sh <>\v i ii l1 a mean duration of the nasalward sweeps was
marked disparity 106 ms fort he left eye, and 102 ms for the between the rates
right one, while the templeward sweeps of the of movement of
the two eyes. respective eyes required only 84 and 86 ms.
Templeward ft) An overall numerical evaluation of a sub sweeps of either
ject's oculomotor performance in each test was eye required
much less time made by combining the values for the three than did the na-
components : blinking, fixation, and oscillation. salward sweeps
(n). Direction of Oculomotor deficit was graded up to a maxi
movements remum, arbitrarily designated as — ■ 4, and a comp versed by camera
osite curve, constructed from data obtained effect.
in 40 experiments on 18 subjects, is shown in
well throughout the period of observation, as well as those whose
performance deteriorated early and did not recover at all, the
A. P. VOL. JUB. 14