Recruitment into psychiatry is perceived as problematic in the UK but the paper by Goldacre and colleagues in this issue of the Journal shows that recruitment of new UK graduates has altered little over 35 years. Sources of psychiatry's malaise should be sought in its changing focus and the quality, rather than quantity, of its recruits.

The mean annual rates of hospital admission for deliberate self-poisoning were calculated for Caucasian and Asian 12–15–year-olds between 1982 and 1990 as 2.31 and 2.47 per 1000 respectively. There was an excess of girls in each group.

During recent years so much has been written about convulsive therapy that it would seem rash to burden still further the very extensive literature on this subject. Nevertheless, in spite of the admirable work that has been done, it cannot be said that the implications of convulsive treatment are fully understood, nor that the fullest possible use has been made of this treatment for research into the pathophysiology of epilepsy. Furthermore, the recent introduction by Cerletti and Bini (1) of a method of treatment with electrically produced epileptic fits has increased still further the scope of research. I have had at my disposal case material provided by the treatment of various forms of mental disorder with cardiazol for more than two years, and with the electrical method in collaboration with Mr. Grey Walter for about six months. This has enabled me to conduct certain investigations of a sort that have received little or no attention hitherto, and to make observations covering various aspects of the physiology and psychopathology of induced epileptic states.

The widespread use of group treatment as one of the modern methods of psychotherapy has provided us with an opportunity to study psycho-social phenomena in a fairly uniform social setting.

The prefrontal region lies in front of the motor and premotor areas of the frontal lobe. In Brodmann's terminology this corresponds to areas 8, 9, 10, 45 and 46 on the convexity; 11, 47 and 13 on the orbital surface; 24, 32 and 12 on the mesial surface. When speaking of selective prefrontal surgery, we mean topectomy or cortical undercutting or partial leucotomy of one or several of these areas.

The finger-prints of mental defectives have been investigated by Bonnevie (1927), Poll (1935) and M⊘ller (1935). It was found that the frequency of whorls was less in the feeble-minded and imbeciles than in the general population but there was no correlation between intelligence quotients and quantitative values of patterns among defectives. Very little investigation has been carried out on the epidermal ridges of palms in mental defectives except in the case of mongoloid imbeciles, where Cummins (1939) and Workman (1939) have investigated the main patterns. Penrose (1949) has measured the angles subtended at triradius t by triradii a and b, and the present writer has counted the epidermal ridges between the a and b triradii.

Thirty elderly demented subjects were assessed in 1985 using a neuropsychological test battery which included tests of parietal lobe function that are allegedly predictive of outcome. Four years later, 29 out of the 30 subjects were followed up. Twelve had died. There were no differences between survivors and deceased in terms of age, pre-morbid intelligence, years of full-time education, or scores on parietal tests. However, proportionally more of the women had died, and those subjects with more global cognitive impairment in 1985 were significantly more likely to have died by 1989. Those who scored lower on an aphasia measure in 1985 were more likely to have died. None of the variables differentiated between survivors and deceased Alzheimer subjects.

For the purposes of this paper the term ‘precipitant’ is defined as a new circumstance associated in time with the onset of the addiction. If the precipitant is merely the ‘final staw’ in a long chain of causal circumstances, no particular value can be attached to its identification. If, on the other hand, the precipitant brings into operation the background causal factors leading to addiction, it has important practical and therapeutic implications. It could be expected that this type of precipitant might persist during the addiction and its disappearance might be associated with remission of the addiction.

The birth records of 78 subjects born to psychotic parents and 72 subjects born to normal parents were studied.

No significant differences in the rates of pregnancy and birth complications (PBCs) were found between the offspring of psychotic parents and normal control parents. There were no differences between offspring born to psychotic mothers compared to psychotic fathers. Neither the onset of the parent's illness, nor the mother's age at delivery, nor the sex of the offspring seemed to influence the rate of PBCs. The offspring of chronic schizophrenic mothers and manic-depressive fathers had lower PBC rates than the offspring of parents of other diagnostic categories. The parents of these two groups, which were of a limited size, did not differ on any variable of significance, excepting the time of their first psychiatric hospital admission.

At the end of the 19th century, when the current psychiatric diagnostic concepts used in most countries were being elaborated, in Germany particularly by Kraepelin, French-speaking psychiatrists, who until then had dominated European psychiatry, continued to develop their own system. This difference in classification is most apparent in the non-organic, non-affective psychoses. Although some of the French names used may be unfamiliar to anglophones, when the new ‘consensus' criteria developed by Professors Pull and Pichot are compared with the DSM-III-R criteria it becomes apparent that the French and American concepts are converging.

Among the theories of making rational choices under uncertainty, the one which originated with an eighteenth-century English clergyman, the Reverend Thomas Bayes, and which uses observed evidence to modify prior judgements concerning different possible ‘causes' of the evidence, is logically nearest to the paradigm of clinical diagnosis. A simple illustration of this theorem is the classical Bertrand Box problem (Fry, 1928), and its wider implications for decision making have been discussed by Birnbaum and Maxwell (1960). An example of the use of the theorem in clinical classification has been given by Maxwell (1961).

As part of an evaluation of the Dangerous and Severe Personality Disorder (DSPD) Programme, we conducted in-depth interviews with 60 participants purposely sampled across four pilot DSPD units. This report is limited to the finding with potential ‘conceptual generalisability’: namely the unanticipated finding of negative and hostile attitudes of participants being managed in hospital units compared with the prison system. The recent UK government policy decision to concentrate this programme in prisons may, in part, reflect the significance of these findings.

Platelet monoamine oxidase (MAO) activity was studied in 33 in-patients with bipolar and schizoaffective disorder who were treated with lithium. Platelet MAO activity was found to increase following 10–41 days of lithium treatment compared to a prior drug free period, and the increase was positively correlated with the duration of lithium treatment. The increase in platelet MAO activity was not correlated with clinical improvement as measured by the Brief Psychiatric Rating Scale (BPRS) and the Global Assessment Scale (GAS). The number of platelets per unit of blood was also significantly correlated with the number of days of lithium treatment. However, the increase in the number of platelets in lithium-treated patients was not correlated with the increase in MAO activity and thus appears not to account for it. These results indicate that studies relating platelet MAO activity to psychiatric diagnosis should be interpreted cautiously if patients are receiving lithium carbonate.