~I~II~I~I~II~II~I~I~II~IIIIIIII~IIIIIIIIIIIIIIIIIIIIIIII~I

Journal of Contemporary Psychotherapy Voh 2, No. 2, pp. 75-86 Winter 1970

Disruption of Affectional Bonds

and Its Effects on Behavior*

Jom BowrmY, M.D.

F AMILY DOCTORS, priests and perceptive laymen have long been aware
that there are few blows to the human spirit so great as the loss of some-
one near and dear. Traditional wisdom knows that we can be crushed by grief
and die of a broken heart, and also that a jilted lover is apt to do things
that are foolish or dangerous to himself and others. It knows, too, that
neither love nor grief are felt for just any other human being but for one,
or a few, particular and individual human beings. The core of what I am
terming an affeetional bond is the attraction that one individual has for
another individual.
Until recent decades science has had little to say about these matters.
Experimental scientists in the physiological or Hullian learning theory tradi-
tions of psychology have never shown interest in affeetional bonds, and have
sometimes talked and acted as though they do not exist. Psychoanalysts, by
contrast, have long recognized the immense importance of affeetional bonds
in the lives and problems of their patients, but have been slow to develop
an adquate scientific framework within which the formation, maintenance
and disruption of bonds can be understood. The void has been filled by
ethologists--starting with Lorenz's (1) classical paper on The Companion
in the Bird's World, progressing through a multitude of experiments on
imprinting .(2, 3) to studies of bonding behavior in sub-human primates (4,
5), and inspiring psychologists to make similar studies of humans (6,7).

~ This paper is released for publication in the Journal through the kind
permission of Oliver & Boyd, Ltd., The Eugenics Society of London, Eng-
land and the editor of Canada's Mental Health.
76 ]OURNAL OF CONTEMPORARY PSYCHOTHERAPY

Before discussing the effects of bond disruption a note about bonding

and its prevalence is in place.
Prevalence of Bonding
The work referred to shows that, even ff not universal in birds and
mammals, strong and persistent bonds between individuals are the rule in
very many species. The types of bond may differ from one species to another.
The most common are those between one or both parents and their offspring,
and those between adults of opposite sex. In mammals, including the pri-
mates, the first and most persistent bond is usually that between mother and
young, a bond which often persists into adult life. As a result of all this
work it is now possible to view the strong and persistent affectional bonds
made by humans from a comparative standpoint.
Affectional bonding is a result of the social behavior of each individual
of a species, differing according to which other individual of his species he
is dealing with, which entails, of course, an ability to recognize individuals.
While each member of a bonded pair tends both to remain in proximity to
the other and to elicit proximity-keeping behavior in the other, individuals
who are not bonded show no such tendencies; indeed, when two individuals
are not bonded, one often strongly resists any approach the other may at-
tempt. Examples are a parent towards the approach of young not its own,
and a male towards the approach of another male.
The essential feature of affeetional bonding is that the two partners
tend to remain in proximity to one another. Should they for any reason be
apart, each will sooner or later seek out the other and so renew proximity.
Any attempt by a third party to separate a bonded pair is strenuously re-
sisted: not infrequently the stronger of the partners attacks the intruder
while the weaker flees or perhaps clings to the stronger partner. Obvious
examples are situations in which an intruder is attempting to remove young
from a mother, e.g., calf from cow, or to detach the female from a bonded
heterosexual pair, e.g., goose from the gander.
Somewhat paradoxically, behavior of an aggressive sort plays key roles
in maintaining affeetional bonds. It takes two distinct forms: first, attacks on
and frightening away of intruders and, secondly, the punishment of an errant
partner, be it wife, husband or child. There is evidence that much aggres-
sive behavior of a puzzling and pathological kind originates in one or other
of these ways.(8)
Affeetional bends and subjective states of strong emotion tend to go
together--as every novelist and playwright knows. Thus, many of the most
intense of all human emotions arise during the formation, maintenance, dis-
ruption, and renewal of affectional bonds--which for that reason are some-
times called emotional bonds. In terms of subieetive experience, the forma-
tion of a bond is described as falling in love; maintaining a bond, as loving
someone; and losing a partner, as grieving over someone. Similarly, threat of
loss arouses anxiety and actual loss, sorrow. Both situations are likely to
DISRUPTION OF AFFECTIONAL BONDS 77

arouse anger. Finally, the unchallenged maintenance of a bond is experi-

enced as a source of security, and the renewal of a bond as a source of joy.
Thus, anyone concerned with the psychology and psychopathology of emo-
tion, whether in animals or man, is soon confronted by problems of affec-
tional bonding: what causes bonds to develop, what they are there for, and
especially problems relating to the conditions affecting the form their de-
velopment takes.
In so far as psychologists and psychoanalysts have attempted to ac-
count for the e~dstence of affectional bonds, the motives of food and sex
have almost always been invoked. Thus, in attempti_ng to explain why a
child becomes attached to his mother, both learning theorists (9, 10) and
psychoanalysts (11) have independently assumed that it is because mother
feeds child. In attempting to understand why adults become attached to one
another, sex has commonly been seen as the obvious and suf/icient explana-
tion. Yet, once the evidence is scrutinized, these explanations are found
wanting. There is now abundant evidence that not only in birds but in
mammals also, young become attached to mother-objects despite not being
fed from that source (12, 13), and that by no means all affeetional bonding
between adults is accompanied by sexual relations. Conversely, sexual rela-
tlons often occur independent of any persisting affectional bonds.
What is now known on the ontogeny of affectional bonds suggests that
they develop because the young creature is born with a strong bias to
approach certain classes of stimuli, notably the familiar, and to avoid other
classes, notably the strange. As regards function, observation of animals in
the wild strongly suggests that the biological function of much, ff not all
bonding, is protection from predators--a function fully as important for the
survival of a population as nutrition or reproduction, but one which has
habitually been overlooked by workers confined in laboratories and to man
living in economically developed societies.
Whether these hypotheses are supported by further work or not, an
individual's capacity to make affectional bonds of a type appropriate to
each phase of his species" life cycle and to his or her own sex, is plainly a
capacity as typical of individuals of mammalian species as are their capac-
ities, for example, to see, hear, eat and digest. And in all likelihood, a capacity
for bonding has as high a survival value to a species as has any of these
other long-studied capacities. It is proving productive to view many of the
psychoneurotic and personality disturbances of humans as being a reflection
of a disturbed capacity for making affectional bonds, due either to faulty
development during childhood or to subsequent derangement.

Disrupted Bonds and Psychiatric Illness

Those who suffer from psychiatric disturbances, whether psychoneurotic,
sociopathic or psychotic, always show impairment of the capacity for affec-
tional bonding, an impairment that is often both severe and long lasting. AI-
78 JOURNAL OF CONTEMPORARY PSYCHOTHERAPY

though in some cases this impairment is dearly secondary to other changes,

in many it is probably primary and derives from faulty development having
occurred during a childhood spent in an atypical family environment. While
disruption of the bonds that tie a child to his parents is not the only form
(adverse in this respect) that the environment can take, it is the form most
reliably recorded and the effects of which we know most about.*
In considering the possible causes of psychiatric disturbance in child-
hood, child psychiatrists were early aware that antecedent conditions of
significantly high incidence are either an absence of opportunity to make
affectional bonds or else long and perhaps repeated disruptions of bonds
once made. (15, 16) However, the view that such conditions are not only
associated with subsequent disturbance but are causal of it, is widely held,
although that conclusion continues to remain debatable.
Studies of the incidence of childhood loss in different samples of psy-
chiatric populations have multiplied in recent years. Due to the samples
and the comparison groups being so differently constituted, criteria of loss
being differently defined, and a host of demographical and statistical hazards,
their interpretation is not easy. Certain findings, however, have been so
consistently reported by independent workers, including reports of a number
of recent and well-controlled studies, that we can be reasonably confident of
them. Two psychiatric syndromes and two sorts of associated symptom are
consistently found to be preceded by a high incidence of disrupted affec-
tional bonds during childhood. The syndromes are psychopathic (or socio-
pathie) personality, and depression; the symptoms, persistent delinquency and
suicide.
Psychopaths
The psychopath (or sociopath) is one who, although not being psy-
chotic or mentally subnormal, persistently engages in: acts against society,
e.g., crime; acts against the family, e.g., neglect, caalelty, sexual promiscuity
or perversion; acts against himself, e.g., addiction, suicide, or attempted sui-
cide, repeatedly abandoning his job. In such people the capacity to make
and maintain affectional bonds is always disordered and not infrequently con-
spicuous by its absence.
More often than not the childhoods of psychopaths are found to have
been grossly dist~trbed by the death, divorce or separation of the parents,
or by other events resulting in disruption of bonds, with an incidence of
such disturbance far higher than is met with in any other comparable group,
whether drawn from the general population or from psychiatric casualties of
other sorts. For example, in a study of over 1,000 consecutive psychiatric out-
patients under the age of sixty, Earle & Earle (17) diagnosed 66 as soeio-
paths and 1,357 as suffering from some other disorder. Taking as their

* There are also valuable studies of the reaction of adults to bereave-

ment and of the relationship of bereavement reactions to mental illness. (14)
D I S R U P T I O N OF A F F E C T I O N A L B O N D S 79

criterion an absence of mother for six months or more before the sixth
birthday, these researchers found an incidence of 41 per cent for the socio-
paths and 5 per cent for the remainder.
When the criterion is made broader, the incidence rises. Thus Craft
and others (18) took as their criterion absence of either mother or father
(or both) before the child's tenth birthday. Of 76 male inmates of the special
hospitals for aggressive psychopaths, no less than 65 per cent had had such
an experience. In a study of several comparison groups Craft shows how
the incidence of this type of childhood experience rises with the degree of
antisocial behavior shown by a group's members.
Amongst others who have reported similar sorts of statistically significant
findings for groups of psychopaths and persistent delinquents are Naess (19),
Greet (20), Brown and Epps (21); and for alcoholics and addicts, Den-
nehy (22). Also, in psychopaths, the incidence of illegitimacy and a shunting
of the child from one ~ome' to another is high.

Suicides
Another psychiatric group which shows a much raised incidence of
childhood loss is that of suicidal patients, both those who attempt it and
those who succeed.* The losses are especially likely to have occurred during
the first five years of life and to have been caused not only by the death
of a parent, but also by other long-lasting causes, notably illegitimacy and
divorce, In these respects suicidal patients tend to resemble sociopaths and,
as will be seen later, to differ from depressives.
Of the many studies reporting a very high incidence of childhood loss
in attempted suicides (23, 24, 25), a recent study by Greer and his asso-
ciates (26) is among the best controlled. A series of 156 attempted suicides
were compared with similar sized samples of non-suicidal psychiatric patients
and of surgical and obstetric patients without a psychiatric history. Both
comparison groups were matched with the attempted suicides for age, sex,
class and other relevant variables. Taking as his criterion of loss the con-
tinuous absence of one or both parents for at least 12 months, Greet finds
that such events have occurred before the fifth birthday three times as often
in the group of attempted suicides as in either of the comparison groups--
an incidence of 26 per cent against 9 per cent for each of the others.
(Table 1)
Furthermore, the losses in the attempted suicide group tended more
often to involve both parents and to have been permanent, whereas in the
other groups they more often concerned only one parent and were temporary,
due to such exigencies as illness or work.
* Athough any group of suicides and attempted suicides will contain
some sociopaths and some depressives, a majority are likely to be diagnosed
as suffering from neurosis or personality disorder (26) and so constitute a
fairly distinctpsychia~e group.
80 ]OUBNAL OF CONTEMPORARY PSYCHOTHERAPY

TABLE 1
Incidence of Loss or Continuous Absence of One or Both
Natural Parents for at Least 12 Months Before the 15th Birthday (26)
Non-Sulcldal
Age at loss Non-Psychiatric Psychiatric Attempted
Patients Patients Suicide
% % %
0-4 years 9 9 26
5-9 year~ IZ 10 fl

10-14 y e a r s 7 7 11

Doubtful 0 2 1

0-14 Z8 28 49

N i56 156 156

In a further study of the same group of attempted suicides (27) it was

found that those who had suffered parental loss before their 15th birthday
differed significantly in certain respects from those who had not. One such
difference, in keeping with other findings, is that those who had suffered
childhood loss were more likely to be diagnosed as sociopaths than were
those who had not suffered a childhood loss (18 per cent against 4 per cent).

Depression
Another condition which is associated with a significantly raised incidence
of childhood loss is depression. The type of loss experienced, however, tends
to be of a kind different to the over-all family disruption typical of the
childhoods of psychopaths and attempted suicides.
First, in the childhoods of depressives, loss is likely to have been due
to the death of a parent rather than to illegitimacy, divorce or separation.
Second, although in depressives the incidence of bereavement tends to be
raised during each quinquenium of childhood, losses tend to have occurred
especially frequently in patients aged ten to fifteen. Third, the incidence of
loss is apt to be most raised for the parent of the opposite sex--mother in
the case of boys, father in the case of girls. Findings of this sort have been
reported by Brown, (28), Munro, (29) Dennehy, (22) and Hill and
Price (30). These studies indicate that loss of a parent by death between
the tenth and fifteenth birthdays tends to have occurred about twice as
frequently in a group of depressives as it has in the population as a whole.

But Is Raised Incidence a Cause?

Thus it now seems reasonably certain that in several groups of psychiatric
patients the incidence of disruption of affeetional bonds during childhood
is significantly raised. While these later studies confirm the earlier findings
DISRUPTION OF AFFECTIONAL BONDS 81

regarding the raised incidence of loss of mother during early childhood,

they also extend them. For several sorts of condition raised incidences of
disrupted bonds are now seen to include bonds to fathers as well as to
mothers and to be found during the years from five to fifteen, as well as
during the first five. Furthermore, in the more extreme conditions, socio-
parlay and suicidal tendencies, not only is an initial loss likely to have
occurred early in life, but it is likely also to have been both a permanent
loss and to have been followed by the child experiencing repeated shifts of
parent figures.
Nevertheless, to demonstrate a raised incidence of some factor is one
thing, but to demonstrate that it plays a causal role is quite another. While
most of those reporting the above findings believe that the raised incidence of
childhood loss bears a causal relation to the subsequent psychiatric disturb-
ance---and there is a wealth of dirdcal records pointing in that direction (8)
--alternative explanations still remain possible. As an example, the raised
incidence of maternal and paternal death in psychiatric patients might be
a result of the parents of patients being older than average at the time of
the patient's birth. Were this so, not only would early death of a parent be
more likely but there might also be greater liability for the offspring to be
born with an adverse genetic loading. Thus, what appears to be an envir-
onmental determinant might turn out to be a genetic one after all.
To test that suggestion is not easy. It requires, first, that the mean ages
of the mothers and/or fathers of psychiatric patients be found, in fact, to be
higher than the mean for the population as a whole; second, that any
raised parental age that may be found, be shown to have an adverse effect
on the genetic endowment of the offspring such that likelihood of psychiat-
ric disability is raised. The first requirement may have already been met:
recent evidence (22) suggests that mean ages of the parents of psychiatric
patients may be above those of the population from which they come. Re-
quirement two, however, is more difllcult to document. It may therefore
be some time before the issue is settled.
Meanwhile those who believe that the relationship between disruption
of affectional bonds during childhood, and disturbance of the capacity to
maintain affectional bonds that are typical of personality disorders of later
fife--those who believe that this relationship is a causal one---point to other
evidence in support of their hypothesis: i.e., the way young humans and
sub-human primates behave when an affeetional bond is broken by separa-
tion or death.
Short-term Effects of Disrupted Bonds
When .a young child finds himself with strangers and without his
familiar parent figures, not only is he intensely distressed at the time, but
his subsequent relationship with his parents is impaired, at least temporarily.
The behavior seen in two-year-olds during and after a short stay in a resi-
dential nursery is the subject of a systematic descriptive and statistical study
82 ] O U R N A L OF C O N T E M P O R A R Y P S Y C H O T H E R A P Y

undertaken at the Tavistock by Heinicke and Westheimer. (31) The relevant

part of this report is where the authors compared the behavior toward their
mother of ten children who had been away in the nursery and are now
home, with that of a comparison group of ten young children who had
remained at home throughout.
In the separated children, two forms of disturbance of affectional be-
havior were seen, neither of which were observed in the comparison group
of non-separated children. One form is that of emotional detachment; the
other, its apparent opposite, i.e., an unrelenting demand to be close to mother.
(i) On first mceting his mother after he has been away from home with
strangers for two or three weeks, a two-year-old child typically remains dis-
tant and detached. Whereas during his early days away a child commonly
cries pathetically for his mother, when at last she returns he seems not to
recognize her or avoids her. Instead of rushing to her and dinging, as he
probably would ff he had been lost in a shop for half an hour, he often
looks right through her and refuses her hand. All the proximity-seeking be-
havior typical of an affectional bond is missing, usually to the mother's
intense distress, and it remains missing--sometimes only for minutes or hours,
but sometimes for days. Resumption of attachment may be sudden but is often
slow and piecemeal. The length of time detachment persists is positively
related with the length of the separation. (Table 2 gives results.)
(ii) When, as is usual, attachment behavior is resumed, a child is
commonly much more clinging than he was before the separation. He dis-
likes his mother leaving him and tends either to cry or to follow her round
the house. How this phase evolves turns a great deal on how his mother
responds. Not infrequently conflict ensues: a child demanding his mother's
constant company, and she refusing it. Such refusal readily evokes hostile
and negative behavior from the child, which is apt to try his mother's pa-
tience even further. Of the ten separated children observed by Heinicke and
Westheimer, six showed strong and persistent hostile behavior to their
mother, and negativism after their return home. No such behavior was seen
in the non-separated children (See table 3).
Clearly, it is still a far ely between showing, on the one hand, that a
child's bonds to his mother--and often to his father also--are thrown into
disequilibrium by a brief separation and, on the other, demonst~aling un-
equivocally that long or repeated separations are causally related to subse-
quent personality disorders. Yet the detached behavior, so typical of young
children after a separation, bears more than a passing resemblance to the
detached behavior of some psychopaths. It would be d~flqcult to distinguish
the aggressively demanding behavior of many a young child recently re-
united with his mother from the aggressively demanding behavior of many
hysterical personalities. Nevertheless, it is useful to postulate that in each
type of case the disturbed behavior of the adult represents a persistence over
the years of deviant patterns of bonding behavior that have become estab-
lished as a result of bond disruptions occurring during childhood. On the
DISRUPTION OF AFFECTIONAL BONDS 83

TABLE 2
Number of Separated and Non-Separated Children Showing
Detachment During First Three Days
After Reunion (or During Equivalent Period) (31)

Separated Not Separated

No Detachment 10
Detachment for One Day Only
Detachment Alternating With
Clinging
Detachment Persistent For
Three Days

10 10

(Degree of Detachment Correlated with Length of Separation:

r = 0.82; P = 0.01)

one hand, it helps to organize data and to orient further research; on the
other, it provides guidelines for the day-to-day management of these kinds
of people.
To advance our knowledge in this field it would obviously be invaluable
to conduct a long series of experiments to investigate the short and long-
term effects on behavior of disrupting an affectional bond, taking into account
the subject's age, nature of the bond, length and frequency of disruptions,
as well as many other variables. It is equally obvious, however, that any such
experiments on human subjects are ruled out on ethical grounds. For these

TABLE 3
Number of Separated and Non-Separated Children Showing
Strong and Persistent Hostility to Mother
After Reunion (or During Equivalent Period) (31)

Separated Not Separated

Little or N o Hostile Behavior

or N e g a t i v i s m to M o t h e r I0
Strong and Persistent Hostile
Behevior and N e g a t i v i s m
to M o t h e r

(P = 0.01) I0 10
84 IOURNAL OF CONTEMPORARY PSYCHOTHERAPY

reasons, it is much to be welcomed that comparable experiments, using sub-

human primates, are now being undertaken. Preliminary findings suggest
that the effects on six-month-old rhesus (monkey) infants of a temporary
loss of the mother (six days) are, both during and after separation, not
unlike those on two-year-old human children. (32) For example, distress
and a lowered level of activity occur during the separation, and an excep-
tions/ly strong tendency to cling to mother after it is over. The monkey-
mother's reactions to this, moreover, are not unlike the human mothers. To
date, however, there is no record of a monkey baby showing detachment,
and this may represent a species difference.
Both in human and in monkey children there are very large individual
variations in the reaction to disruption of a bond. Some of this variation is
probably due to the effects on an infant of events occurring during preg-
nancy and birth. Thus Ucko (33) found that boys, who at birth had been
recorded as having been in an asphyxiated state, are very much more sensi-
tive to environmental change, including separation from mother, than are
boys who at birth had not been asphyxiated. Some other part of this vari-
ance, on the other hand, may well be genetically determined. Indeed, it is
a reasonable hypothesis that a principal way in which genetic factors act to
influence the development of mental health and ill health is by their effect
on bonding behavior: in what degree and form, and in what circumstances
can an individual make and maintain affections/ bonds, and how does he
respond to disruption of bonds? By undertaking studies of this sort it may in
future be possible to bring together environmental and genetic studies of
behaviors/ disorder.

BIBLIOGRAPHY

1. Lorenz, K. Z., Der Kupman in der Umwelt des Vogels, J. Orn. Berl.,
1935, 83. (English translation in Instinctive Behavior, Schiller
(ed.), International Universities Press, New York, 1957.
2. Sluckin, W., Imprinting and Early Learning, Methuen, London, 1964.
3. Bateson, P. P. G., The Characteristics and Context of Imprinting, Biol.
Rev., 1966, 41: 177-220.
4. Sade_D. S., Some Aspects of Parent-Offspring and Sibling Relations
ina" Group of Rhesus Monkeys, With a Discussion of Grooming,
Am. J. Phys. Anthrop., 1965, 23: 1-18.
DISRUPTION OF AFFECTIONAL BONDS 85

5. Hinde, R. A. and Spencer-Booth, Y., The Behaviour of Socially Living

Rhesus Monkeys in Their First Two and a Half Years, Animal
Behaviour, 1967, 15: 169-196.
6. Schaffer, H. T. and Emerson, P., The Development of Social Attach-
ments in Infancy, Mongr. Soc. Res. Child Development, 1964, 29:
1-77.
7. Ainsworth, M. S., Infancy in Uganda: Infant Ca~e and the Growth d
Attachment, Johns Hopkins University Press, Baltimore, Md., 1967.
8. Bowlby, J., Pathological Mourning and Childhood Mourning, J. Ameri-
can Psychoanal. Assn., 1963, 11: 500-541.
9. Dollard, J. and Miller, N. E., Personality and Psychotherapy, McGraw-
Hill, New York, 1950.
10. Sears, R. R., Maceoby, E. E. and Levin, H., Patterns of Child Rearing,
Row, Peterson, Evanston, Ill., 1957.
11. Freud, S., An Outline of Psychoanalysis (in Standard Edition Vol. 23),
Hogarth Press, London, 1964.
12. Harlow, H. F. and Harlow, M. K., The Affectional Systems, Chapter
in: Behaviour of Non-human Primates, Vol. 2, Schrier, Harlow
and Stollnitz (eds.), Academic Press, New York and London, 1965.
13. Cairns, R. B., Attachment Behaviour of Mammals, Psycho. Rev., 1966,
73: 409-426.
14. Parkes, C. M., Bereavement and Mental Illness, Brit., J. Med. Psychol.,
1965, 38: 1-26.
15. Bowlby, J., Maternal Care and Mental Health, World Health Organi-
zation: Monograph Series No. 2, Geneva, 1951.
16. Ainsworth, M. D., The Effects of Maternal Deprivation: A Review of
Findings and Controversy in the Context of Research Strategy,
World Health Organization: Public Health Papers No. 14, Geneva,
1962.
17. Earle, A. M. and Earle, B. V., Early Maternal Deprivation and Later
Psychiatric Illness, Am. J. Orthopsychiat., 1961, 31: 181-186.
18. Craft, M., Stephenson, G. and Granger, C., The Relationship Between
Severity of Personality Disorder and Certain Adverse Childhood
Influences, Brit. J. Psychiat., 1964, 110: 392-396.
19. Naess, S., Mother-Child Separation and Delinquency: Further Evidence,
Brit. J. Criminology, 1962, 2: 361-374.
20. Greer, S., Study of Pm'ental Loss in Neurotics and Psychopaths, Arch.
Gen. Psychiat., 1964, 11: 177-180.
21. Brown, F. and Epps, P., Childhood Bereavement and Subsequent Crime,
Brit., J. Psychiat., 1966, 112: 1043-1048.
22. Dennehy, C. M., Childhood Bereavement and Psychiatric Illness, Brit.
J. Psychiat., 1966, 112: 1049-1069.
23. Brnhn, J. G., Broken Homes Among Attempted Suicides and Psychiatric
Out-Patients: A Comparative Study, J. Ment. Sci., 1962, 108:
772-779.
24. Greer, S., The Relationship Between Parental Loss and Attempted Sui-
cide: A Control Study, Brit. J. Psyehiat., 1964, 110: 698-705.
25. Kessel, N., Self-Poisoning, Brit. Med. J., 1965, 2: 1265-1270, 1336-1340.
86 JOURNAL OF CONTEMPORARY PSYCHOTHERAPY

26. Greer, S., Gunn, J. C. and Koller, K. M., Aetiological Factors in At-
tempted Suicide, Brit. Med. J., 1966, 2: 1352-1355.
27. Greer, S., and Gunn, J. C., Attempted Suicides from Intact and Broken
Parental Homes, Brit., Med. J., 1966, 2: 1355-87.
28. Brown, F., Depression and Childhood Bereavement, J. Ment. Sci.,
1961, 107: 754-777.
29. Munro, A., Parental Deprivation in Depressive Patients, Brit. J. Psy-
ehiat., 1966, 112: 443-458.
30. HII, O. W. and Price, J. S., Childhood Bereavement and Adult Depres-
sion, Brit., J. Psychiat., 1967, 113: 743-751.
31. Heinicke, C. M. and Westheimer, I. J., Brief Separations, International
Universities Press, New York, and Longmans, London, 1966.
32. Spencer-Booth, Y. and Hinde, R. A., The Effects of Separating Rhesus
Monkey Infants From Their Mothers For Six Days, J. Child Psy-
ehol. Psychiat., 1966, 7: 179-198.
33. Ucko, L. E., A Comparative Study of Asphyxiated and Non-Asphyxi-
ated Boys From Birth to Five Years, Developmental Medicine and
Child Neurology, 1965, 7: 643-657.