Some time ago one could have said both truly and succinctly that hospitals are the places where patients are, and from this basic premise it would have been fairly easy to work out a presentable picture of the role of hospitals in research. Now, however, epidemiologists tell us that many mentally sick people are not admitted to hospital, and clinicians stress that if admission occurs it is likely to represent a recurrent episode in a long programme of management in which important therapeutic influences are brought to bear on the patient at work or in other places remote from hospital. It therefore seems reasonable to conclude that certain types or phases of research which might once have been carried out in hospital must now be done elsewhere.

The 1969 disturbances in Belfast were on a scale unprecedented since 1886. By the end of September, there had been 328 injuries requiring hospital treatment, 8 fatalities, and 113 arrests (1). At least 500 houses had been damaged beyond repair and 4,000 people rendered homeless. Some 200 streets were heavily barricaded and guarded, while much of the city was effectively under military occupation.

Since 1955 there has been a steady decline in the number of leucotomy operations, particularly of the open standard type introduced into this country in 1941. It has been considered by most authors that the undesirable side-effects caused this decline but no doubt the introduction of ataractic drugs also played its part (Pippard, 1962). Sykes and Tredgold (1964) discussed in detail the literature up to that time and it was felt unnecessary to repeat that survey here. Suffice it to say that from about 1949 various modifications of the standard operation were devised in the hope of diminishing or eliminating undesirable sequelae. Following the publication of a paper on the late social results of pre-frontal leucotomy by Ström-Olsen and Tow (1949) the late Alexander Kennedy (1949) wrote in the correspondence column of the Lancet—‘The future of this kind of operation (i.e. standard leucotomy) lies in limited and accurately localized sections. The correlation of these with their clinical and neuropathological effects offers a field of study which will occupy us for many years to come’.

Until the last decade there was considerable confusion about the nosology of childhood psychoses. As Kanner (1958) has pointed out, certain psychodynamically orientated writers (Szurek, 1956; Beres, 1956) have eschewed the important operation of differential diagnosis. This has led to the notion of ‘equality of schizophrenias' (Darr and Worden, 1951) and thus to the idea of a single psychosis of childhood. The controversy over this approach has now waned in the face of empirical evidence from aetiological, phenomenological and follow-up studies, and many authors have stressed the importance of age of onset in their typologies or in their attempts at a more comprehensive classification (Kanner and Lesser, 1958; Mahler et al., 1949, 1952; Bender, 1947, 1959; Anthony, 1958, 1962, and Eisenberg, 1967). As this is also central to the present study it merits examination in greater detail.

Although the belief that primitive man might be free from schizophrenia has joined the rest of the golden age myths, there is still a serious question respecting the severity or chronicity of the disease in many tropical peoples. In such peoples typical cases of schizophrenia are proportionately uncommon, and in their place one often meets an acute short-lasting psychosis, which may be indistinguishable from classical schizophrenia in its initial stages but which runs a much shorter course and carries a better prognosis.

In view of the considerable importance attached to sexual adjustment by many psychologists and psychiatrists, it is disappointing that very little work seems in fact to have been done in this field. There are, it is true, surveys of ‘normal’ sexual behaviour, such as those associated with Kinsey, but these are characterized on the whole by the serious limitation of using entirely descriptive statistics; these are useful in a limited sense, but are not very informative. The value of a mean of 2 · 34 for the number of times that members of a given sample indulge in intercourse during the week is doubtful when it is realized that some members of the sample have intercourse once or twice a year, while others have it several times a night; even if one could take the unaided recall of such events very seriously, and even if the rate for a given person were less fluctuating than it probably is, nevertheless when variances are as large as these means have little meaning or relevance. Clearly the important question centres on quite a different problem, namely that of personality traits and other factors giving rise to these very marked individual differences in sexual behaviour and adjustment. Kinsey and his followers have concentrated on the factors of social class and age, perhaps because these are relatively easy to ascertain; it is not so easy to ascertain the psychological factors involved, as many critics of Kinsey's work have pointed out. Nor have psychiatrists, in spite of their professional interest in this field, been more forthcoming; apart from isolated statements about the lower fertility of psychotics, the loss of libido in depression, its excess in manic states, and the widespread sexual troubles of neurotics, little serious and detailed work appears to have been done to relate personality traits, or even descriptive diagnostic labels, to specific types of sexual behaviour. Psychologists have been equally remiss; apart from a few undistinguished and not very meaningful studies reviewed by the writer (Eysenck, 1971) little use has been made of the well-established research methodologies and psychometric tests available to them for use in such a situation.

The difficulties of organizing mental health services in developing countries are made all the greater by inadequacy of information about the extent of illness and disability. Some beginnings have been made in India, particularly under the sponsorship of the All India Institute of Mental Health, Bangalore. The Mental Health Advisory Committee of the Government of India (1966) suggested a probable prevalence of mental illness of 20 per 1,000 population in general, 18 per mille for semi-rural and 14 per mille for rural areas. These figures are much lower than the 72 per 1,000 suggested by Sethi et al. (1967). Ganguli (1968) estimated a prevalence rate of 140 per 1,000 in industrial workers near Delhi. Incidence rates have been much less studied than prevalences (Lin and Standley, 1962). A WHO Expert Committee on mental health convened in 1960 suggested as a working definition of a case of mental illness:

Little has been published about psychiatric out-patient work, and even less about the relationship between in-patient and out-patient psychiatric facilities in service requirements for a given population. The reasons for this are partly general to the whole field of medicine —a preoccupation with severe illness and the hospital bed to which it inevitably leads— and partly particular to psychiatry. The growth of out-patient psychiatric work is relatively recent, and its existence outside teaching hospitals is largely a National Health Service development. With in-patients concentrated in large mental hospitals serving an extensive area, the psychiatrist's connection with any individual general hospital was necessarily limited and the out-patients a small and peripheral part of his work.

At one time the doctor used to treat both the patient and his illness; with the understanding of aetiology comes the possibility of ignoring the patient, saying only ‘take the pills and you will recover’. Since such treatments have become all too easy, both doctors and patients have turned away from them to personality problems and the symptoms of psychogenic disorders, so much more difficult, more time-consuming and more interesting.

The period of the study was from August 15th to the end of September, 1969. Severe rioting occurred in Belfast on the nights of 14th and 15th August, and during the next few weeks rioting, arson, looting and intimidation continued. Eight fatal casualties were recorded and a large number of people were injured, mainly from gunshot wounds.

This paper analyses the clinical picture in 80 children referred to the Park Hospital, Oxford, or to the Newcastle department of child psychiatry, and admitted for intensive assessment of their psychosis. All were seen by two psychiatrists and nearly three quarters of them in Oxford. The diagnostic criteria, and the differentiation by age of onset into infantile psychosis (I.P.) and late onset psychosis (L.O.P.) were discussed in the previous paper (Kolvin, I).

There is considerable dispute over the value of frontal leucotomy in the treatment of mental illness. Uncontrolled studies have been numerous. Many report good results (Pippard, 1955; Scoville, 1960; Sargant and Slater, 1963; Birley, 1964; Sykes and Tredgold, 1964; Post, Rees and Schurr, 1968). Some are more sceptical (Barahal, 1958). Controlled investigations have been few, and their findings unfavourable. Robin (1958) observed no difference between the outcome of 198 patients who had a standard leucotomy for functional psychosis and an equal number of carefully, but retrospectively, matched controls. In another five-year controlled follow-up study involving 183 patients in each group no significant differences in outcome, as measured by the rate of hospital discharge was found. (McKenzie and Kaczanowski, 1964.)

The first paper of this sequence discussed the division of childhood psychoses into those with onset before the age of 3 (infantile psychoses, I.P.) and those with onset after the age of 5 (late onset psychoses, L.O.P.). The second paper described in detail the clinical phenomena in 80 psychotic children so classified at Oxford or Newcastle (see Table I), and showed that this division corresponded to clear-cut distinctions in the clinical pictures. The I.P. group showed gaze avoidance, abnormal pre-occupations, self-isolating patterns of behaviour, echolalia, overactivity. The L.O.P. group showed hallucinations and thought disorder. With these two groups clearly established we now look to see whether anything in their family and background also distinguishes them.

Gross somatic illnesses and psychotic reactions associated with pregnancy, parturition and the puerperium are well known, but it is only in recent years that attention has been paid in any systematic way to milder neurotic reactions post partum. During the last decade reports have appeared from several centres on woman's reactions to pregnancy, studied from various viewpoints. As regards psychological reactions it is obvious that the modern Western woman has to a considerable extent an ambivalent attitude to maternity. It is therefore hardly surprising if this attitude, together with the sometimes pronounced somatic strain of pregnancy, results in a high frequency of emotional disturbances during and after pregnancy.

It has been estimated (Zellweger, 1961) that less than 5 per cent of all cases of mental subnormality have a hereditary metabolic defect. This is a figure which requires confirmation, and is one that is likely to vary in different parts of the world. We have accordingly carried out a systematic clinical and biochemical screening of the cases of mental subnormality (IQ lower than 70) attending the out-patient department of the Mental Hospital, Bangalore. A total of 729 patients (439 males and 290 females) were available for this study.

In recent years there has been a considerable interest in the classification of depressed patients, and in the use of multivariate statistical techniques as tools for classifying, with the hope of either validating traditional classifications or suggesting useful new ones. One technique employed for this purpose has been factor analysis. A number of studies have used this method to explore the overlapping endogenous-reactive or psychotic-neurotic dichotomies. Most have demonstrated a factor or factors reflecting such a dimension which appears to form a consistent, systematic source of variance in ratings of depressed patients (1, 5, 11, 15, 16, 18, 26, 31, 33).

The Department of Health carried out a complete census of the resident patient population of local authority and private psychiatric hospitals in the Republic of Ireland on 31 March, 1963. This paper analyses the Irish census returns and compares them with those of the similar psychiatric hospital census carried out in England and Wales on 31 December, 1963 (Brooke, 1967).

Failure of treatment does not necessarily mean inappropriate treatment; it must first be ascertained that the treatment has been carried out. Increasing attention is being paid to this question in regard to drug therapy with estimations of blood and urine levels. Patients who fail to get better after leucotomy may have been unsuitably chosen; they may also fail to get better because the aim of the operation, namely to sever the thalamo-frontal radiations, has not been achieved. Neuro-anatomical studies have long suggested that not only do many patients remain with these tracts intact following a blind operative approach but that there is also great variability in the size and position of the lesions produced, a consequence of unreliable surface markings, brain fibre elasticity, complicating haemorrhages and progressive gliosis. The blind approach is no ideal surgical method, since it results in a largely fortuitous lesion (Meyer and McLardy, 1948, 1949; Beck, McLardy, and Meyer, 1950; Eie, 1954; Meyer and Beck, 1954). There is even a gross discrepancy between the actual and desired point of insertion of the leucotome (Dax, 1943). Whilst the effects of misplaced cuts have been studied by these workers, one can only guess what proportion of clinical failures can be attributed to misplaced cuts and what proportion of successes follow misplaced cuts (Meyer and Beck, 1954). Appeals for further studies with post-mortem correlates have been made by these authors.

Psychoanalytic literature has a number of references to suicidal acts taking place on the anniversary of the death of a loved person. Anniversaries of significant bereavements have been seen as likely to reactivate a depressive syndrome and as possible precipitants of suicide. Zilboorg (7) suggested that suicides might take their lives on the anniversary of a parent's death as though to expiate the death which they unconsciously perceived as the result of their own wishes. Others have seen the meaning of the suicide's act as a fantasied rebirth or new beginning with a lost love object, and the supposed occurrence of suicides on the anniversary of a death has again been instanced as support (Jackson (3)).

The first part of this study of female monozygous twins discordant for childbirth covered intrapair differences before pregnancy (Kaij and Malmquist, 1970). This part is focused on the influence of motherhood on health, presented as intrapair differences in gynaecological, other somatic, and psychiatric morbidity after pregnancy.