Sleep Disorders: Disorders of Arousal?

Author(s): Roger J. Broughton

Source: Science, New Series, Vol. 159, No. 3819 (Mar. 8, 1968), pp. 1070-1078
Published by: American Association for the Advancement of Science
Stable URL: https://www.jstor.org/stable/1723604
Accessed: 11-03-2019 19:23 UTC

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1723604?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

American Association for the Advancement of Science is collaborating with JSTOR to

digitize, preserve and extend access to Science

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
 References and Notes
symbolized
symbolizedin
inFig.
Fig.2,2,B Band
and
C,C,
is,is,
how-
how- 8. S. G. Smith, Can. Entomologist 94, 941
(1962); Can. J. Genet. Cytol. 7, 363 (1965).
ever,
ever, aa relatively
relativelyminor
minorone,
one,
and
and
Fig.
Fig.
2C 2C 1. E. Mayr, Systematics and the Origin of 9. J. L. Aston and A. D. Bradshaw, Heredity
Species (Columbia Univ. Press, New York, 21, 649 (1966).
should definitely not be considered as 1942); "Change of genetic environment and 10. T. A. Vaughan, Evolution 21, 148 (1967).
an allopatric model in the classical sense evolution," in Evolution as a Process, J. 11. M. J. D. White, R. E. Blackith, R. M.
Huxley et al., Eds. (Allen and Unwin, Lon- Blackith, J. Cheney, Australian J. Zool. 15,
of the dumbbell diagram. don, 1954); "Species concepts and defini- 263 (1967).
The concepts presented here are put tions," in The Species Problem (AAAS, 12. H. Lewis, Science 152, 167 (1966).
Washington, D.C., 1957), pp. 371-388; Animal 13. G. C. Webb, unpublished work.
forward in the hope that they may Species and Evolution (Harvard Univ. Press, 14. M. J. D. White, Australian J. Zool. 5, 285
stimulate renewed interest in the cyto- Cambridge, Mass., 1963). 1957); and L. J. Chinnick, ibid., p.
2. G. L. Stebbins, Processes of Organic Evolu- 338.
genetic processes involved in animal tion (Prentice-Hall, Englewood Cliffs, N.J., 15. M. J. D. White, H. L. Carson, J. Cheney,
1966). Evolution, 18, 417 (1964).
speciation. It is beginning to appear 3. A. K. Lee, Australian J. Zool. 15, 430 (1967). 16. M. J. D. White, unpublished work.
that there are more different kinds of 4. M. Wasserman, Proc. Nat. Acad. Sci. U.S. 17. L. Sandler and E. Novitski, Amer. Natural-
46, 842 (1960); Univ. Texas Publ. 6205, ist 91, 105 (1967).
mechanisms involved than was suspect- Studies in Genetics 2, 85-117 (1962).
18. S. G. Smith, Chromosoma 18, 380 (1966).
ed a few years ago. Differences in 5. H. L. Carson, F. E. Clayton, H. D. Stalker, 19. A. Meylan, Rev. Suisse Zool. 72, 636 (1965);
Proc. Nat. Acad. Sci. U.S. 57, 1280 (1967). 73, 548 (1966).
modes of speciation are clearly related 6. M. D. McCarthy, Amer. Naturalist 79, 104 20. R. Matthey, ibid. 73, 585 (1966).
to differences in population dynamics (1945); H.-G. Keyl, Chromosoma 13, 464-514 21. Some of the cytogenetic studies discussed in
(1962); R. W. Dunbar, Can. J. Zool. 27, 497- this article have been supported by Public
and population structure; but they may 525 (1959); Nature 209, 597 (1966); P. 0. Health Service grant No. GM-07212 from
Ottonen, Can. J. Zool. 44, 677 (1966); K. the Division of General Medical Sciences,
also depend on differences in the ge- Rothfels and M. Freeman, ibid., p. 937. U.S. National Institutes of Health, and by
netic system. 7. H.-G. Keyl, Chromosoma 17, 139-180 (1965). a grant from the Reserve Bank of Australia.

which time the enuresis occurs. The

subject is difficult to arouse, confused,
disoriented even to the extent of deny-
ing that the bed is wet, and completely
unable to recall any dreams.
When no organic cause (for example,
Sleep Disorders: pathology of the genitourinary system,
epilepsy, and so on) is known, enuresis
Disorders of Arousal? often is described as "idiopathic" or
"essential." It has been interpreted as
an expression of aggression, Oedipus
Enuresis, somnambulism, and nightmares occur in fixation, or pathologically deep sleep
(3). The unmotivated nature of the
confusional states of arousal, not in "dreaming sleep."
episode and the social ostracism in-
curred have been well described by
Roger J. Broughton George Orwell (4), who suffered from
childhood enuresis:

I knew
knewthat
thatbed-wetting
bed-wettingwaswas
(a) wicked
(a) wicked
and (b) outside my control. The second
fact I was personally aware of, and the
Nocturnal enuresis, somnambulism, psychological
psychologicalmechanisms
mechanisms areare
also
also
pre-
pre-
first I did not question. It was possible,
the sleep terror, and the nightmare are sented. Until these mechanisms are therefore, to commit a sin without know-
important social and medical problems. understood, treatment will remain ing you committed it, without wanting to
Their social impact can be appreciated empirical. commit it and without being able to avoid
it.
from the extent to which they appear in I begin with a summary of the essen-
the literature of various cultures (ex- tial features of each of the four attacks. 2) In a typical attack of somnambu-
amples are Lady Macbeth's sleepwalk- 1) Nocturnal enuresis, or bed-wet- lism or sleepwalking the individual sits
ing and the nightmares of Dante's Souls ting, is a common symptom in children up quietly, generally an hour or two
in Purgatory). They pose medical prob- and young adults. Incidences have been after falling asleep, gets out of bed, and
lems because of their frequency, their reported as follows: 10 to 15 percent moves about in a confused and clumsy
unresponsiveness to treatment, and their for "nervous" children and 30 percent manner. Soon his behavior becomes
similarity, in certain respects, to other, for institutionalized children (1); 1 per- more coordinated and complex. He may
more dangerous sleep disorders, espe- cent for U.S. naval recruits who have avoid objects, dust tables, go to the
cially nocturnal epileptic seizures. Re- been previously screened for the dis- bathroom, or utter phrases which are
cent studies have helped clarify their order (2); and 24 percent for naval usually incomprehensible. It is difficult
pathogenesis. Data are presented here recruits discharged on psychiatric to attract his attention. If left alone, he
which support the hypothesis that they grounds (2).
occur independently of typical periods Typically the subject awakens to find The The author
authoris is
assistant
assistant
professor
professor
in the
indepart-
the depart-
of dream activity, a view in direct con- himself in soaked bedclothes. An ob- ment of neurology and neurosurgery, McGill Uni-
versity, Montreal, Canada, and clinical neuro-
tradiction to widespread long-standing server usually notes movement succeed- physiologist, Montreal Neurological Institute. This
assumptions. The results of recent in- ed by several seconds of tranquillity, article is based upon a paper presented before
the Association of Psychophysiological Study of
vestigations of their physiological and with apparent continuation of sleep, at Sleep, Gainesville, Florida, in March 1966.
1070
SCIENCE, VOL. 159

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
goes back to bed. A great deal of stim- the sleeper's chest. This is beautifully bursts of rapid eye movements (9, 10).
ulation is required to awaken him. And illustrated in Fig. 1, and in the etymol- The incidence of dream recall (that is,
when he is awakened he has little if any ogy. The root word in the case of the the probability that a subject will recall
recollection of his sleepwalking activi- English nightmare, the German Nacht- having dreamed) following arousal
ties and no recollection of dreaming. mar, and the French cauchmar is the from this state has been found to vary
Nevertheless, these remarkable attacks ancient Teutonic mar, meaning devil. between 70 and 90 percent (10, 11).
are almost universally interpreted as an Thus nightmare and Nachtmar mean Since publication of these initial reports,
"acting out of dream activity" (5). nocturnal devil, and cauchmar, which other biological changes have been doc-
3) The sleep terror or pavor noc- derives also from caucher, an ancient umented as characteristic of this state.
turnus attack of children is very differ- French verb meaning "to press," means Excellent reviews are available, by De-
ent. The child abruptly sits up in bed "pressing devil." The Italian incubo ment, Jouvet, Snyder, Roffwarg, and
and screams. He appears to be staring comes from the Latin in and cubare, others (12, 13).
wide-eyed at some imaginary object; his signifying "to lie upon." The neurophysiological changes (12,
face is covered with perspiration and Freud (8) believed that both the in- 13) typifying rapid-eye-movement sleep
his breathing is labored. Consoling stim-cubus attack and the night terror of include cerebral activation affecting at
uli have no effect. After the attack, children express "repressed sexual im- least the neocortex, specific sensory re-
dream recall is rare and usually frag- pulses." But the four types of attack lay nuclei, and the so-called limbic
mentary, and the dreams recalled re- have more generally been interpreted as system and require participation of
semble daytime fantasies. The child has being due to dream activity, despite the brainstem structures, including a re-
no recollection of the episode the fol- infrequency of dream recall. And the stricted pontine area (the nucleus retic-
lowing morning. amnesia has often been considered a ularis pontis caudalis and nucleus
These three disorders are often found result of "repression." coeruleus). The state has been called
in combination. All three may be mani- Recently, the "dream hypothesis" of rapid-eye-movement (or REM) sleep,
fested by a single subject at different the pathogenesis of the sleep disorders stage-I REM sleep, desynchronized
times. Or all may appear in the family has become verifiable, as true dreaming sleep, paradoxical sleep, and rhomben-
history; thus, 34 percent of enuretic has been shown to occur in association cephalic sleep.
naval recruits have a personal history, with a recurrent and distinct physiolog- After infancy, the sleep of normal
and 25 percent have a family history, ical state. subjects always begins with a state qual-
of sleepwalking (6). Finally, they may In this article I examine a number itatively different from REM sleep and
be combined in a single attack, a sleep of questions. What are the known divisible
phys- by electroencephalographic cri-
terror or an enuretic episode being fol- iological correlates of dreaming? Do teriathe
into stages I to IV in the Dement
lowed, for example, by prolonged sleep disorders occur when dreaming and isKleitman classification (10). The
somnambulism. taking place? If not, when do they first sign of decreasing vigilance is usu-
The nightmare or incubus of adults occur? To what extent might some or ally a slowing and diffusion of the alpha
has received particular attention over all of the symptoms of these apparently rhythm (stage IA1) (14, 15), which be-
the centuries (7). During an attack the "psychological" disorders be explicable comes fragmented (stage 1A2) before
adult suddenly cries out in his sleep and by physiological factors associated with being replaced by diffuse theta activity
shows all the signs of intense anxiety: changes in the waking-sleeping cycles? (stage IB). Soon sharp waves with
sweating, a fixed facial expression, di- I review investigations that have pro- phase reversal around the vertex ap-
lated pupils, difficulty in breathing. The vided some answers to these questions pear, followed by 12- to 14-cycle-per-
essential symptoms of the nightmare are and report new studies concerning second spindle activity occurring syn-
respiratory oppression, paralysis, and others. What neurophysiological mecha- chronously in both central regions
anxiety, usually occurring in that order. nisms underlie the common physiolog-(stage II), and by K-complexes repre-
The anxiety is believed by many to be ical correlates of the disorders? How senting cerebral potentials evoked by
the most intense ever experienced by relevant are they in explaining the various stimuli.
man (7). Well-structured dream activ- symptoms of the attacks? To what ex- As vigilance continues to diminish,
ity is rare. [This is in contrast to the tent, if any, is the attack pattern in a background slow-wave delta activity (1
terrifying dream (see cover) which con- given individual determined by preced- to 3 cycles per second) increases. When
tains complex imagery, remains a se- ing physiological rather than psycholog- this is present in less than 50 percent
quence in time, and is not accompanied ical events? I conclude with a discussion of the tracing, the electroencephalogram
by respiratory oppression (7).] Reports of the possible relationships of the is classified as stage III; when it is pres-
of apparent dreaming are usually poor attacks to various types of mental ent in more than 50 percent, the classi-
descriptions of oppressive situations, activity. fication is stage IV (9, 14, 15). Stages
such as being locked up in a tomb, or III and IV are often referred to together
having rocks piled on one's chest. The as "slow-wave sleep"; stages I to IV, as
following morning the subject has abso- Physiological Correlates of Dreaming non-rapid-eye-movement or non-REM
lutely no recollection of the attack. Sub- sleep. Such sleep is characterized also
jects typically show intense daytime For over a decade it has been known by a subsequent low incidence of dream
anxiety. that dreaming is statistically related to recall,
a generally given as 5 to 10 percent
Writers of antiquity, the Middle distinct and cyclically recurring type (9-11),
of although higher figures have
Ages, and even the late 18th century sleep characterized briefly by a low- been reported.
believed that nightmares were caused voltage electroencephalogram lacking The physiological correlates of non-
by a nocturnal demon pressing upon "sleep spindles" or K-complexes, and by REM sleep include decreased function
8 MARCH 1968 1071

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
of the brainstem "reticular activating If the sleep disorders are related to, showed that nocturnal enuresis in chil-
system" (resulting in the slow waves of and perhaps even generated by, dream dren and in young naval recruits oc-
cortical origin) and synchronization of activity, they would be expected to oc- curred during slow-wave sleep, was not
unspecific thalamocortical connections cur mainly or exclusively during REM associated with dreaming, and, in fact,
(responsible for the sleep spindles). sleep. Sleep studies of the actual attacks actually retarded the appearance of the
Non-REM sleep therefore has been de- therefore are of particular interest. first rapid-eye-movement period. These
scribed as synchronized or telencephalic observers nevertheless interpreted enu-
sleep. resis as a "dream equivalent"-that is,
In the adult, a night of normal sleepSleep Studies of Sleep Disorders a phenomenon replacing the initial
consists of a series of cycles, each con- "dream period."
taining non-REM, then REM, sleep in- Nocturnal enuresis has been studied Bental then published a study of two
terspersed with occasional awakenings. more carefully than any of the other teen-age brothers having almost nightly
The first cycle usually has much slow- sleep disorders considered here. I there- enuresis, both of whom passed into long
wave activity and may lack a rapid-eye- fore discuss some of the features com- (about 2.5-hour) "dissociated states"
movement stage. As the cycles pass, the mon to all four in the context of this (states in which electroencephalograms
proportion of REM sleep increases and disorder. typical of wakefulness were associated
that of non-REM sleep, especially slow- In 1955 Ditman and Blinn (15) with behavioral sleep) before one wet
wave sleep, decreases. The average dur- reported that micturition in young hischil-
bed and the other awoke to void
ation of the sleep cycles increases with dren coincided with electroencephalo- (17). Saint-Laurent et al. (18), in a
age, from 45 to 55 minutes in the infant graphic patterns of slow-wave sleep, pilot study at Gastaut's laboratories in
to 75 to 90 minutes in the young adult and in young adults, with patterns re- Marseilles, demonstrated that, in epi-
(13). sembling wakefulness. Pierce et al. (16) leptic children, nocturnal enuresis takes

Fig. 1. "The Nightmare," painting by Johann Heinrich Fiissli (Henri Fuseli), 1741-1825. The monster sitting on the patient's chest
and the female horse leering in the background refer to the ancient Teutonic word mar, meaning devil, from which nightmare is
derived, and the English word mare, which it suggests. [Courtesy of the Detroit Institute of Arts]
1072 SCIENCE, VOL. 159

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
place during arousal from slow-wave with electroencephalographic patterns The Confusion of Awakening
sleep and is almost always nonepileptic of slow-wave sleep, light sleep, or even
in nature. wakefulness. To a considerable extent The most striking aspect of our initial
this depends upon the recording meth-
In order to clarify the contradictions studies of these sleep disorders was the
in the literature, a program of enuresisod. That is, if the recorder is far fromfinding that all four took place during
research was initiated in 1962, under the meatus (it is often placed under arousal
the from slow-wave sleep rather than
the supervision of Gastaut, in which bed sheets), micturition which started in, or during arousal from, REM sleep.
more sophisticated recording techniques in slow-wave sleep may not be signaled Furthermore, they were associated
were employed: use of specially deviseduntil considerably later in the arousal. with body movement, autonomic activa-
bikini-type shorts housing electrodesMore recent results (21) again indicate tion, and the following six symptoms:
which were short-circuited by the firstan age factor: in children nocturnal (i) mental confusion and disorientation;
drops of urine at the meatus and, inenuresis occurred during slow-wave (ii) automatic behavior; (iii) relative
nonreactivity to external stimuli; (iv)
later studies, continuous direct-currentsleep, but in adults it occurred 12 to 45
recording of intravesicular pressure byminutes after electroencephalographic poor response to efforts to provoke be-
means of an indwelling catheter. patterns of non-REM sleep had been havioral wakefulness (in the sense of
These studies (14, 19, 20) showed replaced by patterns of wakefulness full and lucid contact with the environ-
that enuresis in childhood and in ado- associated with behavioral sleep (22). ment); (v) retrograde amnesia for
lescence almost always occurs during a Again it did not take place during many intercurrent events (an entire
partial or complete arousal from slow- rapid-eye-movement periods, and there sleepwalking episode, the activity of the
wave sleep as a part of a more or less was no related dream recall. investigators, and so on); (vi) only
constant sequence of events which we In subsequent studies, sleepwalking fragmentary recall of apparent dreams,
called the "enuretic episode." The epi- (limited by the length of the electrode or none at all.
sode usually began with body movement wires to a distance of about a meter) It was hypothesized that these six
associated with increased muscular tone, was recorded in 18 of 30 subjects said common symptoms might be attribut-
tachycardia, tachypnea or apnea, a de- to be suffering almost nightly from these able to the arousal process itself. We
crease in skin resistance, and other signs attacks. Like enuresis, it was found to therefore investigated normal subjects
of arousal. Micturition itself took place be related to arousal from slow-wave and sleepwalkers during abrupt arousal
when electroencephalographic patterns sleep (14, 19) except for one episode from the various stages of sleep (14,
were still those of stage-IV sleep or, that occurred during transition from 19). We paid particular attention to
much more frequently, had become stage-II REM sleep (14, 19, 23). It was comparisons between arousal from
those of stage III, II, or I or even wake- never related to dream recall. The as- slow-wave sleep (stages III and IV)
fulness. The period of latency from the sociation of sleepwalking with arousal and from REM sleep. Telemetric rec-
first electroencephalographic or poly- from slow-wave sleep has been con- ordings of the subject's electroenceph-
graphic indication of arousal to the firmed by Jacobson et al. (24). alogram patterns and eye movements,
beginning of the micturition was found Seven episodes of night terror or filmed recordings of his behavior, and
to increase more or less linearly with pavor nocturnus were recorded in sevenmagnetic-tape recordings of his speech
increase in vigilance, as judged by elec- children. These attacks occurred during were made. Arousal was achieved by
troencephalographic criteria. Thus the sudden and intense arousal, again from suddenly standing the subject up and
average periods of latency from the ini- slow-wave sleep. The electroencephalo- questioning him.
tial sign of arousal during stage-IV sleep gram became that of wakefulness (14, In the case of REM sleep the thresh-
(bursts of higher-voltage delta-wave 19). old for arousal was high. Nevertheless,
activity, tachycardia, or other sign) to Finally, ten typical nightmare attacks the subjects became lucid almost imme-
micturition associated with an electro- were recorded during similar intense diately and generally (70 percent) re-
encephalogram typical of stage III, II, arousal from slow-wave sleep in five called dream activity. The threshold for
or I or wakefulness were 37, 68, 100, adults and one adolescent. A further arousal in slow awakening from
and 140 seconds, respectively. nightmare in an adult also followed slow-wave sleep was also high. But
Few enuretic episodes (three of 22) arousal from stage-IV sleep, but only forced arousal from slow-wave sleep, in
were related in any way to REM sleep. after stage-IA patterns had been re- striking contrast to arousal from REM
One occurred during passage from stage corded for 25 minutes. None of these sleep, frequently produced periods of
II into REM sleep, and two occurred attacks were associated with REM mental confusion with poorly coordi-
after REM sleep had been well estab- sleep or with detailed dreams (14, 19). nated automatic behavior, distant or
lished. The "wet dreams" so often men- Apart from a case report (25) and in- slurred speech, relative nonreactivity,
tioned in the literature of enuresis were cidental mention, in two instances, of retrograde amnesia, and a lack of dream
present only if the bedclothes were left apparent nightmare attacks during in- recall-in short, all the common symp-
wet and the subject awakened during vestigation of other phenomena (26), toms described for the sleep attacks.
subsequent REM sleep-a situation I have found no other reports of studies This was true whether the forced
which allowed the incorporation of ex-of either the night terror of children orarousal occurred during continuous
teroceptive stimuli into dreaming. the nightmare in the literature. slow-wave sleep or following the body
The finding that enuresis occurs at An interesting feature of the investi-movement that frequently "spontane-
some point along a continuum of in-gations of these sleep disorders was the ously" terminates such electroenceph-
creasing vigilance (arousal) from slow- finding that the incidence of the attacks alographic patterns. The intensity and
during recording sessions was lower
wave sleep to a stage of lighter sleep or duration of confusion were greatest in
even wakefulness reconciled the diver- than the incidence when the subject was children and in sleepwalkers. Kales et al.
gent views on whether it is synchronous in his home milieu. (27) recently reported similar findings.
8 MARCH 1968 1073

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
 Confusional episodes associated with were studied. Only findings relevant to ways. [Interestingly, Shakespeare, dur-
awakening from an apparently "deep" the present discussion are given here. ing Lady Macbeth's sleepwalking (32),
sleep have been known for centuries; Stimulation was provided by means has the Doctor of Physic say, "You see,
they are referred to in the German liter-of a sound-attenuated Grass stroboscope her eyes are open," to which the Wait-
ature as Schlaftrunkenheit (28) and in(peak intensity, 106 lumens per square ing-Gentlewoman responds, "Ay, but
the French literature as l'ivresse du meter) placed 25 centimeters from the their sense is shut."] It should be noted
sommeil (29). They were induced ex- bridge of the nose. The electroencepha- that in experimental studies of the cat
perimentally in man as early as 1897, logram was amplified by four Grass by Walsh and Cordeau (33), arousal
by de Manaceine (30). P511 preamplifiers (50-percent attenua- from slow-wave sleep produced long
We went on to induce attacks of som- tion at 0.3 cycle and 500 cycles per sec- periods of inhibition of the primary
nambulism experimentally in sleepwalk-ond), then stored on an Ampex 1200 visual cortex, usually preceded by
ers or, sometimes, normal children bytape recorder, and the responses (N= transient facilitation. Furthermore, such
giving the subjects a basic biological 100, occasionally 50) were summated arousal is associated with reduced rates
drive for micturition (a full bladder by a Mnemotron CAT 400B digital of background firing of unit cells in the
produced by forced intake of water be-computer and written out on an X-Y visual cortex (34, 35) and with a de-
fore retiring) and suddenly arousing plotter. The reference electrode was on crease in evoked unit activity (35).
them in stage-IV sleep (14, 19). The the earlobe. Superficial electromyograms Following arousal from REM sleep,
subjects walked about (usually to a dis- of scalp muscles were recorded. however, cerebral reactivity to visual
tant toilet), voided, and returned to Arousal from slow-wave sleep and stimuli returns quite directly to that of
bed. All had amnesia when forcibly from REM sleep had markedly different the waking state and, not surprisingly,
awakened at the end of the episode. The effects on the visual evoked potential. there is no mental confusion. It is in
electroencephalographic record made Following one-fourth of the arousals fact disputable whether the term arousal
during these confusional periods gen- from slow-wave sleep, the visual evokedshould be applied to awakenings from
erally showed a diffuse and rather high- potential at the standard occipital or REM sleep, where, contrary to the case
amplitude alpha rhythm which reacted vertex electrode initially contained ap- for awakening from non-REM sleep,
poorly or not at all to intense light parent carry-over of components typical the brain shifts from one complex pat-
stimuli, or else showed patterns of of slow-wave sleep despite the return of tern of intense activity to another and
low-voltage delta and beta activity, lack- electroencephalographic patterns char- some cerebral functions might even be
ing spindles. That forced arousal from acteristic of the waking state. These per-viewed as becoming less active (12, 13).
slow-wave sleep can produce sleepwalk-sisting components were either a surface The importance of these findings
ing has been confirmed by others (27, positivity peaking at 75 ? 15 milli- would appear to be as follows. If the
30a). seconds or a large negativity peaking at altered cerebral responsiveness to ex-
The belief that arousal from stage-III 250 ? 50 milliseconds, or both; they teroceptive stimuli in arousal from slow-
or stage-IV sleep is the optimum, if notgave rise to a visual evoked potential wave sleep indeed represents a carry-
the necessary, condition for the confu- intermediate between that of slow-wave over of non-REM sleep reactivity or
sional syndrome common to all the sleep and that of wakefulness (Fig. 2). some degree of inhibition or decrease in
sleep disorders was the basis of recentEven when there was no such carry- integration of sensory systems, confu-
investigations performed at the Mon- over, the visual evoked potential regu- sion would certainly be expected. Con-
treal Neurological Institute. larly showed decreased amplitudes and fusion from any etiology (drug intoxi-
increased latencies of later components cation, epilepsy, head trauma, and so
(Figs. 2 and 3), despite marked mydri- on) is associated with a "liberation" of
Physiology of the asis, before reverting to the amplitudes behavioral automatisms, the organism
Confusion of Awakening and latencies of presleep wakefulness reverting to, or adapting by means of,
(31). Following arousal from REM simpler and less conscious activity. The
If the confusional state following sleep, on the other hand, the visual decreased behavioral reactivity to other
arousal from slow-wave sleep were, as evoked potential was essentially that ofenvironmental stimuli during the confu-
our results implied, one of impaired presleep wakefulness (Fig. 3), and no sional episode can reasonably be com-
cerebral responsiveness or of "function- changes similar to those in subjects pared to the altered cerebral reactivity
al deafferentation," this would at least awakened from non-REM sleep were to at least visual stimuli. And when one
partially explain all the common symp- observed. considers that the movement, increased
toms. The hypothesis is amenable, with- The effects of the two types of arous- tone, and autonomic phenomena are
in certain limitations, to investigation by al on the visual evoked potential paral- part of the arousal itself, it becomes ev-
scalp recording of the brain's responses leled the behavioral changes. After ident that arousal from slow-wave sleep
to controlled stimulation. arousal from slow-wave sleep, mental can produce all the symptoms com-
The visual evoked potential was confusion was associated with altered mon to the four major sleep disorders.
studied during a presleep period of re- cerebral reactivity to visual (and prob- The postarousal state following slow-
laxed wakefulness following 20 to 30 ably also other) stimuli despite the re- wave sleep would therefore appear to be
minutes of dark-adaption with the eyesturn of an alpha rhythm characteristic the necessary, but not sufficient, condi-
closed, during various stages of sleep,of wakefulness. The nature of this alter- tion for the occurrence of the confu-
and during and following arousal from ation indicates either a continuation of
sional sleep disorders, and to explain
slow-wave sleep and REM sleep. No responsiveness typical of non-REM the common symptons, although not the
hypnotics were given, and the subject's sleep or changes in later visual-evoked-initiating or triggering mechanisms. But
pupils were not dilated by means of potential components suggestive of what determines the nature of the at-
drugs. Ten subjects, aged 6 to 30 years,inhibition or occlusion of visual path- tacks?

1074 SCIENCE, VOL. 159

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
Pathophysiology of the Attack Type self consisted of an increase in the The differences in bladder physiol-
frequency and magnitude of these ogy con-between enuretic and nonenuretic
Continuous recordings of the bladder tractions, culminating in micturition, children
at observed throughout sleep sug-
pressure of enuretic and normal chil- which time the peak pressure attained gested that other aspects of sleep, and
dren during sleep had previously shown values up to 10 cm-H20 before drop- especially of automatic function, might
quantitative differences between the two ping to near-zero levels (14, 20). In be affected. A statistical study of the
groups during and, interestingly, inde- REM sleep, however, there were few sleep cycles and of cardiac and respira-
pendently of the attack (20). Specifical- spontaneous bladder contractions, and, tory functions was therefore made from
ly, the enuretic group had an increase in of course, there was no series of con- data accumulated in Marseilles. In this
the number and intensity of spontane- tractions leading to enuresis. study, data for ten enuretic children on
ous and evoked primary detrusor con- A few of the recordings of bladder 13 nights on which enuretic episodes
tractions during non-REM sleep. Blad- pressure of enuretic subjects were, it occurred were compared with data for
der contractions could be evoked by turned out, made on nights when there ten nonenuretic children on 13 nights;
such stimuli as clicks, hand clapping, or were no enuretic episodes. Excessive the experimental and control groups
other noises, even when these caused bladder contractions were observed were matched for age (6 to 16 years)
no observable electroencephalographic throughout these nights. But these con- and sex (six males, four females).
changes. Occasionally a "spontaneous" tractions never reached the magnitude During the sleep of normal subjects
bladder contraction would precede a K- of those observed during nights when the periods of slow-wave sleep (stages
complex, indicating that these potentials such episodes occurred, and they never III and IV) are often terminated by
can be evoked by interoceptive as well culminated in a series of contractions movement, autonomic discharge, and
as exteroceptive stimuli. Any arousal such as generally precedes an enuretic other signs of arousal. (Direct passage
during non-REM sleep provoked pri- episode. from slow-wave REM sleep is also ob-
mary bladder contractions. Normal subjects, by comparison, served, but it is rare.) This recurrent
During stages III and IV, the spon- showed very few primary detrusor con- phenomenon was designated the "slow-
taneous increases in bladder pressure tractions during sleep, and the increases wave arousal episode." In the study,
were of the order of 3 to 15 cm-H20, in pressure were almost never above 25 each enuretic episode was compared to
but they sometimes reached 50 cm-H20 to 30 cm-H20. Baseline pressures, the corresponding slow-wave arousal
or even more. The enuretic episode it- moreover, were generally lower. episode that terminated the identi-

Waking ....
Presleep \ \ I

Stage IV /

~ ~~~/\Arousal . ....- "">.

,/ .. 120iiV .

+ 2 min

Arousal REM
+5 minmV Arousal

0 100 200 300 400 500 5V

msec
Fig. 2 (above). A comparison of the visual evoked potential during
presleep wakefulness, stage-IV sleep, and the confusional period
following arousal from stage-IV sleep for a 16-year-old subject. (The 0 100 200 300 400 500
recording is from the occipital electrode to the earlobe; 100 responses msec
were summated.) During slow-wave sleep the visual evoked potential
shows consistent marked positivity at the occipital electrode (positivity is in the downward direction), with a peak latency of about
85 milliseconds and a later negativity at 250 milliseconds. The visual evoked potential 2 minutes after arousal in this case shows
definite carry-over of components of the wave form of stage-IV sleep (superimposed dotted line), producing a potential intermediate
between sleep and wakefulness despite the presence of an alpha rhythm in the electroencephalogram. The visual evoked potential
5 minutes after arousal has a wave form similar to that of wakefulness, although various components have slightly greater latency
and lower amplitude than are typical of wakefulness for this subject. Fig. 3 (right). Visual evoked potential for a 27-year-old
subject during a confusional period 4 minutes after arousal from slow-wave sleep (stage IV) and during a lucid state 4 minutes after
arousal from REM sleep, relative to presleep wakefulness (dotted line). (The recording is from the occipital electrode to the earlobe;
100 responses were summated.) Note the decrease in amplitude and especially the increase in latencies of the later components of
the potential after arousal from slow-wave sleep. The potential after REM sleep is within the variability of that of wakefulness.
8 MARCH 1968 1075

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
Table 1. Comparison of relative intervals (group means) toa different sleep stages and to the enuresis (usually stage-II sleep) and in
slow-wave arousal episode for enuretic patients (children) and matched control subjects. The
means values for the two groups are essentially identical [P > .05 (Student's t-test)] The onlysubsequent REM sleep. Cardiac varia-
aspect of sleep cycles approaching significant levels is the longer interval to initial REM bility (arrhythmia) was correspondingly
sleep in enuretics (P = 08).
significantly lower in these children.
Interval to stage onset (minutes) Respiration (Table 3) showed a decrease
in variability only in stage IV for the
Subject Recording Last IA- Last IA- Last IA- Last I-slow-wave Last 1A-
ons et- stage II stage III stage IV arousal episode 1st enuretic
last IA
REM group. The data for the control
group of normal children as a function
Enuretics 38.89 2.96 8.82 15.62 72.13 174.96 of stages of sleep are qualitatively sim-
Controls 38.51 3.92 9.65 15.96 65.44 141.01
ilar to data for adults reported by other
workers (36, 37).
This statistical investigation of the
cal cycle for the matched control sub- proached significance was the consider- sleep of enuretic children confirmed the
ject. able increase in mean latency to the hypothesis suggested by the all-night
The intervals to the onset of the vari- first REM-sleep period in enuretic sub- studies of bladder function that auto-
ous stages of sleep and to the slow-wave jects (13 minutes later in these subjects,nomic changes occur throughout sleep,
arousal or the enuretic episode were P .08). Rapid-eye-movement sleep independent of the enuretic episode it-
measured (see Table 1). Also, the aver- always followed enuresis, usually after self, and, moreover, are of the type that
age rates and percentile variability (de- 5 to 15 minutes of further non-REM would predispose the subject to micturi-
gree of arrhythmia) were determined sleep. tion. The enuretic episode in fact ap-
for cardiac and respiratory function The heart rate of enuretic chil- pears to represent simply an increase of
(Tables 2 and 3). Variability was calcu- dren (Table 2) was significantly these higher changes to a clinically overt level.
lated for each minute as the longest than that of normal children before And by interrupting the normal flow of
minus the shortest relevant interval (R- sleep, in stage-IV sleep, and during sleep the cycles, the episode retards but
wave and peak inspiration, respectively) arousal episode. Five of the 13 studies does not replace the first REM period.
expressed as a percentage of the shortestof enuretics actually showed tachycardia The enuretic episode therefore can in
interval. Average variability was the after the subject had fallen asleep, and no way be considered a "dream equiva-
mean of the minute-by-minute results only three showed bradycardia. Because lent," as Pierce suggested (16).
for the period concerned. of this, further analysis was made, Patients who suffer from other sleep
The data analyzed were restricted to which indicated that the mean heart disorders, such as somnambulism, night-
pre-episode and related post-episode rate in stage-IV sleep of enureticsmares, was and sleep terrors, manifest other
sleep and to later REM-sleep stages. not significantly different from theconstellations
rate of physiological changes
Cycles which occurred much later in during wakefulness, whereas the heart which differ from patterns in normal
the morning were not included, due to rate of normals during such sleep was subjects, which occur both during the
the laborious nature of the calculations. significantly lower (at the 2-percentattacklev- and at other times, and which
No studies were made to determine have not yet been completely studied.
el) than the rate during wakefulness.
whether there were differences inMoreover,
sleep the increase in heart rate Individuals subject to nightmares, for
patterns and physiological function of instance, have relative tachycardia dur-
during the arousal episode over the rate
(i) normal children and (ii) enuretic during stage-IV sleep was greater in ing slow-wave sleep and hyperreactive
enuretics (P < .05), despite the higherheart rates during arousal episodes
children on nights when the latter expe-
rienced no enuretic attacks. base line. The heart rate of enuretic throughout the night. Sleepwalkers
The results showed that the only as- children remained higher (Table 2) evenmake many movements independent of
pect of sleep cycles (Table 1) which ap- during the 5-minute period following the attack of somnambulism itself which
are complex and gestural in nature and
may occur in association with contin-
Table 2. Comparison of heart rates (group means, in beats per minute) for enuretic
uous patients
electroencephalographic slow-wave
(children) and matched control subjects at various stages before, during, and after the slow-
wave arousal episode. Values in parentheses indicate variability (in percentage). activity (14,
19, 30a); they also have
more marked and longer confusional
Subject Presleep Stage IV Slow-wave
episode (peak)
arousal Post-episode REMepisodes
sleep following forced awakening
from slow-wave sleep (14, 19, 27).
Enuretics 94.13' ( 9.51)* 93.05t (11.77)t 128.10t (41.12) 86.75* (15.20) 92.20t (16.17)$ Subjects with a classical sleep dis-
Controls 84.47 (17.03) 72.96 (26.85) 101.67 (35.36) 74.69 (19.06) 78.77 (30.91) order therefore appear to manifest
* P < .05 (Student's t-test). t P < .001. :: P < .01. In all other cases, P > .05. physiological changes which predispose
them to attacks of a given type. These
changes may be enhanced during some
Table 3. Comparison of respiration rates (group means, in respirations per minute) for enuretic
nights,
patients (children) and matched control subjects at various stages before and after the slow-culminating in an attack.
wave arousal episode. Values in parentheses indicate variability (in percentage). All theItcom-
should be noted that very similar
parisons but one are insignificant at the 5-percent level. No data are given for the attacks
period ofmay occur in other subjects dur-
slow-wave sleep arousal, because movement artifacts made respiratory measurements obtained
at that time unreliable. ing intense arousal from virtually any
background level of vigilance. Attacks
Subject Presleep Stage IV Slow-wavearousal Post-episode REM sleepduring wakefulness consist of micturi-
tion during stress (stress incontinence);
Enuretics 21.21 (19.85) 19.76 (19.95)* - 17.96 (22.77) 17.31 (28.90) acute anxiety attacks with confusion,
Controls 21.03 (22.39) 17.91 (28.55) - 16.94 (22.06) 16.42 (3916) automatic behavior, or even fugues, and
* P < .01 (Student's t-test).
with more or less complete retrograde
1076
SCIENCE, VOL. 159

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
amnesia; episodes of overt terror, rage, different from dreaming as the Summary physio-
and so forth. Even when arousal is not logical features of non-REM sleep are
as intense as it is during such attacks from those of REM sleep (43). Such In summary, the classical sleep dis-
but is still supraoptimal, measurable de-mental activity lacks the panoramic orders of nocturnal enuresis, somnam-
grees of impairment of performance are hallucinatory features of true dreamingbulism, the nightmare, and the sleep
observed (38), which can be compared and appears to be much more closely terror occur preferentially during arous-
to the lowered psychomotor functioning related to daytime concerns and to the al from slow-wave sleep and are virtu-
following arousal from slow-wave sleep. environmental situation. Subjects usual- ally never associated with the rapid-eye-
I therefore emphasize the point that ly describe it as "thinking," and they movement dreaming state. Original data
the sleep disorders are best considered may even deny having slept. Recht- are reported here which indicate that
disorders of arousal. schaffen et al. (44) have shown that physiological differences from normal
The conclusions, then, are that the this mental activity is much more com-subjects, of a type predisposing the in-
attacks are essentially disorders of mon than true dreaming in episodes of dividual to a particular attack pattern,
arousal; that the slow-wave sleep arousal talking during non-REM sleep, where are present throughout the night. The
episode is a normal cyclic event setting the effects of arousal are much less of a episode, at least in the case of enuresis,
the stage for them; that preexisting contaminant than they are in full awak-appears to be simply a reinforcement of
constellations of physiological changes enings. these differences to a clinically overt
predispose a subject to a particular type It is therefore conceivaible that, in the level.
of attack during the ,arousal episode; deepest stages of non-REM sleep, when A number of features are common to
and that the postarousal confusional the risk of subsequent recall is minimal, all four sleep disorders. These had been
state explains the many common symp- protective barriers are lowered, allowing shown previously to be attributable to
toms. These conclusions lead to a such mental activity to express emotion- the arousal itself. New data obtained by
more fundamental and perhaps unan- al conflicts which otherwise are repressed, means of evoked potential techniques
swerable question. What, during slow- and that such activity leads to the suggest that these common symptoms of
wave sleep, actually provokes the onset attacks. But the interpretation of any the confusional period that follows non-
of the attacks? This brings us back to data on mental activity that are gath- REM sleep are related to alterations of
consideration of the relationship of the ered following arousal from non-REM cerebral reactivity, at least of the visual
disorders to preceding, and therefore sleep is in fact fraught with the difficulty system.
possibly causative, mental activity. of controlling a possible postsleep gen- The symptoms which distinguish the
esis of this activity during intermediate individual attack types (that is, micturi-
stages of impaired consciousness before tion, prolonged confusional fugues,
Relation of Attacks to Mental Activity the subject is fully awake. overt terror) appear to be based upon
The possibility must even be enter- physiological changes present through-
There is no evidence whatsoever that tained that such disorders arise from a out sleep which are markedly accentu-
REM sleep and its concomitant dream- "psychological void," and that even the ated during arousal from slow-wave
ing play any role in the genesis of these symptoms specific to the attack types sleep. These changes may in some way
attacks. The various theories of the bio- are due to abnormally marked physio- be related to diurnal psychic conflicts.
logical function of REM sleep (13, 39) logical responses during the recurrent But, to date, it has proved impossible to
are therefore considered irrelevant. slow-wave arousal episode (bladder con- demonstrate potentially causal psycho-
Non-REM sleep itself has been traction, producing enuresis; sensoriumlogical activity, dreaming or other forms
shown to have an incidence of dream impairment, leading to somnambulism; of mental activity, or even a psycho-
recall usually varying from 5 to 10 per- cardiorespiratory changes, giving rise tological void in sleep just preceding the
cent (10, 11). But incidences as high asterror). attacks.
54 percent have been reported (40). Im- In fact, a serious and unresolved par- The presence of all-night or even day-
portant factors affecting the degree of adox exists. Although mental activity of time predisposing physiological changes
recall appear to be the abruptness of any one of a number of types (complex and the difficulty in obtaining any solid
awakening (41) and the strictness with hallucinations, illusions, "thinking," and evidence of a preceding psychological
which dreaming is defined (42). Kales so on) might or might not exist in slow- cause explain, no doubt, why the re-
et al. (42), in a recent study directed wave sleep before the attacks, arousal sults of efforts to cure the disorders at
toward explaining these divergent fig- would almost certainly erase any recol- the moment of their occurrence (for
ures, found only a 4-percent recall fol- lection of such activity (as it erases, example, by conditioning procedures in
lowing awakenings from non-REM indeed, most recollection of the attacks nocturnal enuresis) have been far from
sleep in the first half of the night and themselves) and thereby make detailed satisfactory.
the least recall following awakenings study of its nature, or even formal proof I stress the points that the attacks are
from stage-III and stage-IV sleep. of its occurrence, impossible. We may best considered disorders of arousal and
Dreaming is therefore reported least be faced with a biological example of that the slow-wave sleep arousal episode
often by normal subjects following peri- the indeterminacy principle. But cer- which sets the stage for these attacks is
ods when the four sleep disorders are tainly sleep and arousal (as evidenced a normal cyclic event. Indeed it is the
typically most frequent. No studies have by activation of memory traces in REM most intense recurrent arousal that an
been made to determine whether such sleep, physiological amnesia of recent individual regularly experiences.
low percentages of recall for normals events following passage into non-REM The most fruitful possibilities for fu-
hold true for subjects with these sleep sleep or following arousal from non- ture research would appear to be more
disorders. REM, but not REM, sleep) appear high- detailed studies of those physiological
There is some evidence that mental ly relevant to recent research (45) into changes that predispose individuals to
activity occurs which is as qualitatively
memory consolidation and retrieval. certain types of attacks when they un-
8 MARCH 1968
1077

This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
All use subject to https://about.jstor.org/terms
dergo intense arousal or stress; the re-
versal of these changes by psychological
or pharmacological means; and more
refined investigations of the physiolog-
ical and psychological characteristics of
the process of cyclic arousal from non-
REM sleep.
References and Notes
1. F. N. Anderson, Amer. J. Diseases Children
40, 591 (1930).
2. A. Levine, Amer. J. Psychiat. 100, 320
(1943); C. M. Pierce, H. H. Lipcon, J. H.
McLary, H. F. Noble, J. U.S. Armed Forces
7, 208 (1956).
3. M. Hader, Arch. Gen. Psychiat. 13, 296
(1965); R. Strom-Olson, Lancet 1950-11, 133
(1950); S. I. Roland, J. Urol. 71, 216 (1954);
K. J. Bostock, Med. J. Australia 1, 240
(1962).
4. G. Orwell, Such, Such Were the Joys (Har-
court, Brace, New York, 1945).
5. J. A. Sours, P. Frumpkin, R. R. Indermill,
Arch. Gen. Psychiat. 9, 400 (1963); G.
Andr6-Balisaux and R. Gonsette, Acta Neu-
rol. Psychiat. Belg. 56, 270 (1956).
6. C. M. Pierce and H. H. Lipcon, Arch,
Neurol. Psychiat. 76, 310 (1956).
7. E. Jones, On the Nightmare (Hogarth, Lon-
don, 1949), pp. 1-349.
8. S. Freud, in The Basic Writings of Sigmund
Freud (Random House, New York, 1938),
pp. 523-524.
9. E. Aserinsky and N. Kleitman, Science 118,
273 (1953); ----, J. Appl. Physiol. 8, 1
(1955); -, ibid., p. 11.
10. W. C. Dement and N. Kleitman, J. Exp.
Psychol. 53, 339 (1957); -, Electroen-
cephalog. Clin. Neurophysiol. 9, 673 (1957).
11. W. C. Dement and E. A. Wolpert, J. Exp.
Psychol. 55, 543 (1958); E. A. Wolpert and
H. Trosman, Arch. Neurol. Psychiat. 79, 603
(1958); M. Jouyet, F. Michel, D. Mounier,
Rev. Neurol. 103, 189 (1960); R. J. Berger
and I. Oswald, Science 137, 601 (1962); I.
Oswald, Sleeping and Waking (Elsevier, Am-
sterdam, 1962), pp. 120-146.
12. M. Jouvet, Arch. Ital. Biol. 100, 125 (1963);
F. Snyder, Arch. Getz. Psychiat. 8, 381
(1963); -, Amer. J. Psychiat. 122, 377
(1965); W. C. Dement, in New Directions in
Psychology, T. Newcomb, Ed. (Holt, Rine-
hart and Winston, New York, 1965), vol. 2,
p. 135; M. Jouvet, Physiol. Rev. 47, 117
(1967).
13. H. P. Roffwarg, J. N. Musio, W. C. De-
ment, Science 152, 604 (1966).
14. H. Gastaut and R. Broughton, in Recent
Advances in Biological Psychiatry, J. Wortis,
Ed. (Plenum, New York, 1965), p. 197.
15. K. S. Ditman and K. A. Blinn, Amer.. J
Psychiat. 3, 913 (1955).
16. C. M. Pierce, R. M. Whitman, J. W. Maas,
1078
This content downloaded from 205.208.116.24 on Mon, 11 Mar 2019 19:23:28 UTC
M. L. Gay, Arch. Gen. Psychiat. 4, 166
(1961); C. M. Pierce, Can. Psychiat. Ass.
J. 9, 415 (1963).
17. E. Bental, J. Psychosomat. Res. 5, 116 (1961).
18. J. Saint-Laurent, C. Batini, R. Broughton,
H. Gastaut, Rev. Neurol. 108, 106 (1963).
19. H. Gastaut, C. Batini, R. Broughton, J.
Fressy, C. A. Tassinari, in Le Sommeil de
Nuit Normal et Pathologique, H. Fischgold,
Ed. (Masson, Paris, 1965), pp. 215-236;
-, Rev. Neurol. 110, 309 (1964).
20. R. Broughton and H. Gastaut, Electroen-
cephalog. Clin. Neurophysiol. 16, 625 (1964);
H. Gastaut and R. Broughton, ibid., p. 626;
- , J. Saint-Laurent, CO Batini, "Recent
Advances in Bedwetting," C. M. Pierce, Ed.,
unpublished.
21. D. R. Hawkins, J. Scott, G. Thrasher, paper
presented before the annual meeting of the
Association for the Psychophysiological Study
of Sleep, Washington, D.C., 1965; E. Ritvo,
E. Ornitz, F. Gottlieb, A. Poussaint, B.
Maron, S. La Franchi, K. Ditman, K. Blinn,
paper presented before the annual meeting of
the Association for the Psychophysiological
Study of Sleep, Santa Monica, Calif., 1967.
22. S. K. Schiff, J. Nervous Mental Disease 140,
397 (1965).
23. H. Gastaut and R. Broughton, Perceptual
Motor Skills 17, 362 (1963).
24. A. Jacobson, A. Kales, D. Lehmann, J. R.
Zweizig, Science 148, 975 (1965).
25. H. Gastaut, M. Dongier, C. Batini, J. Rhodes,
Rev. Neurol. 107, 277 (1962).
26. D. 0. Hebb, in Brain Mechanisms and Con-
sciousness, E. D. Adrain, F. Bremer, H. H.
Jasper, Eds. (Blackwell, Oxford, 1964), p.
418; W. R. Dement, in Science and Psycho-
analysis, J. Masserman, Ed. (Grune and
Stratton, New York, 1964), pp. 129-162.
27. A. Kales, J. D. Kales, R. D. Walter, A.
Jacobson, M. J. Paulson, paper presented
before the annual meeting of the Associa-
tion for the Psychophysiological Study of
Sleep, Gainesville, Fla., 1966.
28. H. Gudden, Arch. Psychiat. 40, 989 (1905).
29. C. Marc, De La Folie (Bailliere, Paris, 1840),
pp. 658-668.
30. M. de Manac6ine, Sleep: Its Physiology,
Pathology, Hygiene and Psychology (Walter
Scott, London, 1897).
30a. A. Jacobson and A. Kales, Res. Pub. Assoc.
Res, Nervous Mental Disease 45, 424 (1967).
31. W. A. Cobb and G. D. Dawson, J. Physiol.
London 152, 108 (1960); L. Ciganek, Elec-
troencephalog. Clin. Neurophysiol. 13, 165
(1961); M. Ebe et al., Tohoku J. Exp. Med.
77, 352 (1962); H. Gastaut, E. Beck, J. Faid-
herbe, G. Frank, J. Fressy, A. Remond, C.
.Smith, P. Werre, in Progress in Brain Re-
search, vol. 1, Brain Mechanisms, G. Mo-
ruzzi, A. Fessard, H. H. Jasper, Eds. (Else-
vier, Amsterdam, 1963), pp. 374-392; H.
Gastaut and H. R6gis, in "Symposium on the
Analysis of Central Nervous System and
Cardiovascular Data using Computer Meth-
ods," NASA Pub. SP-72, L. D. Proctor and
W. R. Adey, Eds. (1965), pp. 7-34.
All use subject to https://about.jstor.org/terms
32. W. Shakespeare, Macbeth, act V, scene 1.
33. J. T. Walsh and J. P. Cordeau, Exp. Neurol.
11, 80 (1965).
34. D. H. Hubel, J. Physiol. London 147, 226
(1959).
35. E. V. Evarts, Fed. Proc. 19, 828 (1960).
36. J. Fressy, thesis, Univ. of Aix-Marseille
(1963).
37. F. Snyder, J. A. Hobson, D. F. Morrison,
F. Goldfrank, J. Appi. Physiol. 19, 417
(1964); A. Shapiro, D. R. Goodenough, I.
Biederman, I. Sleser, ibid. 18, 778 (1964).
38. G. L. Freeman, Amer. J. Psychol. 45, 17
(1933); D. 0. Hebb, Psychol. Rev. 62, 243
(1955); R. B. Malmo, ibid. 66, 367 (1959).
39. H. S. Ephron and P. Carrington, Psychol.
Rev. 73, 500 (1966); F. Snyder, Amer. J.
Psychiat.. 123, 121 (1966).
40. W. D. Foulkes, J. Abnormal. Soc. Psychol.
65, 14 (1962).
41. D. R. Goodenough, H. B. Lewis, A. Shapiro,
L. Jaret, I. Sleser, J. Personality Soc.
Psychol. 2, 170 (1965).
42. A. Kales, F. S. Hoedemaker, A. Jacobson,
J. D. Kales, M. J. Paulson, T. E. Wilson,
Perceptual Motor Skills 24, 555 (1967).
43. D. R. Goodenough, A. Shapiro, M. Holden,
L. Steinshriber, J. Abnormal Soc. Psychol.
59, 295 (1959); J. Kamiya, in Functions of
Varied Experience, D. W. Fiske and S. R.
Maddi, Eds. (Dorsey, Homewood, Ill., 1961),
p. 145; W. D. Foulkes, Psychol. Bull. 62,
236 (1964); A Rechtschaffen, P. Verdone, J.
Wheaton, Can. Psychiat. Ass. J. 8, 409
(1963).
44. A. Rechtschaffen, D. R. Goodenough, A.
Shapiro, Arch. Gen. Psychiat. 7, 418 (1962).
45. L. Weiskrantz, in Amnesia, C. W. M. Whitty
and 0. L. Zangwill, Eds. (Butterworths, Lon-
don, 1966), p. 1; J. A. Deutsch, Ann. Rev.
Physiol. 24, 259 (1962).
46. The early investigations performed in Mar-
seilles, France, were made during tenure of
a fellowship of the Queen Elizabeth II Can-
adian Research Fund. The studies of evoked
potentials were begun while I was a fellow
of the Medical Research Council of Canada
and completed after I became a member of
the staff of McGill University, Montreal, and
the Montreal Neurological Institute. I am
indebted to Professor Henri Gastaut, Mar-
seilles, for his guidance and stimulation dur-
ing the years in which the first part of this
research was done. Miss Christine Li6nard
and Dr. Sebastien Falchetto helped in the
statistical analysis. The investigations of
evoked potentials were made in collaboration
with Dr. Karl-Heinz Meier-Ewert and Dr.
Mituru Ebe and benefited greatly from the
technical assistance of Mr. Wayne Gibson.
Dr. Frederick Snyder, Dr. Howard Roffwarg,
Dr. Anthony Buffery, and Miss Marie Feltin
kindly read the manuscript and suggested
improvements. I am indebted to Mr. Charles
Hodge for preparing the illustrations and to
Miss Ann-Marie Crosby for the typing. Figure
1 was brought to my attention by J. R. Geigy
of Basel, Switzerland, and provided through
the courtesy of the Detroit Institute of Arts.
SCIENCE, VOL. 159