Dependency refers to an individual's attachment to and reliance on others. There are at least two different kinds of dependent behaviour. Person-oriented dependency includes close physical proximity, affectionate contact and verbal interchanges. Task-oriented dependency is concerned with help given for personal activities such as dressing and assistance with jobs around the house (Maccoby and Masters, 1970).

This paper reports some aspects of a survey made of the training experiences of a group of recently appointed consultants in general psychiatry; it is the third such survey made by the author. The original enquiry was a postal one and asked about the training and education received by all consultants who had been appointed for the first time to a post in general psychiatry, with at least 6 N.H.S. sessions, between 1 October 1963 and 30 September 1969 (R.M.P.A., 1969). A second survey, using a questionnaire much modified from the first, was made of consultants appointed between 1 October 1966 and 30 September 1969, using the same criteria of eligibility as the first group (Brook, 1972, 1973). The present enquiry used a postal questionnaire almost identical to that used for the previous group, which was circulated to consultants in general psychiatry appointed between 1 October 1969 and 30 September 1972.

In 1966 a census was carried out of the epileptic population in prisons and borstals in England and Wales (1). This indicated that there are at least 7.1 epileptics/1,000 men in these institutions, a figure higher than that expected from general population estimates. The present paper reports an attempt to explain this higher prevalence using an interview study of a representative sample of epileptic prisoners together with a similar study of non-epileptic prisoners matched for institution and date of reception into prison.

Shortly after the advent of modern psychopharmacological agents in the 1950s it was noted that their clinical use led occasionally to galactorrhoea (Gäde and Heinrich, 1955; Winnik and Tennenbaum, 1955), to menstrual disorders (Polishuk and Kulcsar, 1956), and to false positive pregnancy tests (Foxworth and Lehman, 1957; Marks and Shackcloth, 1966). The drugs were also shown to have a profound effect upon the oestrous cycle (Cranston and Segal, 1960) and breast activity (Sulman and Winnik, 1956; Sulman, 1970) of laboratory animals. Most published reports incriminated the phenothiazines, especially chlorpromazine, and the literature pertaining to this drug was reviewed in detail by De Wied (1967); but similar effects have been noted with several related psychotropic substances (Shader and Di Mascio, 1970). These side effects are generally accepted to result mainly from the action of the drugs on the hypothalamus (De Wied, 1967), but the precise mechanisms involved are not fully understood.

It is a paradox that during the years of psychosocial development each man has to achieve an identity, a role in which he can feel secure, and yet poets and philosophers tell us that man is forever seeking to diminish the isolation which a personal identity imposes. Without the development of a personal identity and a role in life the individual remains a social isolate, a dependent parasite or a social deviant. Yet, once achieved, identity brings a new awareness of separateness from others and a need to merge the self with others, to identify with someone or something else. This would seem to be a main objective of love. It has been said that human love is the supreme act of communication with another, a merging of identity, and poets speak of communicating with nature or with God in like terms. Romantic and sacred literature abound in metaphors and images describing the bliss which such activity offers. If it is true that while we have a need to develop a sense of personal selfhood, an identity and a role, we have a further need to diminish the separateness such awareness imposes, how do we go about it? The contemporary answer is, of course, by communication, by relating to others. The better the communication we have with others, the greater our sense of belongingness, the less the pain of individual isolation. It is not necessary to emphasize that a cardinal feature of all severe mental illness is either a loss or a distortion of the individual's communicative capacity. But I would like to emphasize that the debate as to whether failure of communication is either a cause, or an association, or a consequence of mental illness is not my theme.

The National Health Service has entered a period of substantial change. The impending reform of local government (1974) will coincide with considerable reforms in the administrative structure of the N.H.S. (1). In addition, our subject matter—the mental health services—has been the subject of various Government White Papers and policy decisions (2). These reforms are an attempt to increase the efficiency of the service, and throughout their pages reference is made to the need for scientific management and administrative efficiency.

Mortality rates in psychiatric patients have been reported as higher than those of the general population in Scandinavia (Odegaard, 1952), the United States (Gorwitz et al., 1966; Babigian and Odoroff, 1968), and Scotland (Innes and Millar, 1970). These findings may be related both to a greater prevalence of physical disease amongst psychiatric patients (Kay and Roth, 1955; Culpan et al., 1960; Shepherd et al., 1964; Kay and Bergman, 1966; Eastwood and Trevelyan, 1972) and to a greater frequency of suicide (Stenstedt, 1952; Stenstedt, 1959; Pokorny, 1964).

In England, public intoxication has been a criminal offence ever since an Act of Parliament was passed in the reign of James I, somewhat over three and a half centuries ago. Edwards (1970) compiled some figures on these offences for the last two hundred years, showing that the rate of arrest was much higher in the nineteenth century than at present. For example, in 1878 there were 70 arrests per 10,000 of the population, compared with a comparable figure for 1968 of 16. Nevertheless, the latest available figures from the Home Office (1971) show that 82,961 persons were found guilty of simple or aggravated public drunkenness in 1971. In the U.S.A., with its apparently much larger alcoholism problem, there were nearly one and a half million arrests for this group of offences in 1966 (Pittman, 1969); this accounts for one third of all arrests in that country (Chafetz, 1971).

There is ample evidence that high doses of phenothiazines and other neuroleptics depress the pituitary-gonadal axis in experimental animals (De Wied, 1967), but the effects of these drugs on sex hormones in human subjects are still controversial (Shader and Di Mascio, 1970). Literature concerning growth hormone (GH) is even more controversial, since phenothiazines have been found to inhibit GH release in rodents (Muller et al., 1967) but to increase it in the rhesus monkey (Meyer and Knobil, 1967). In human subjects phenothiazines have been reported to depress both basal levels of GH and the response to hypoglycaemia (Sherman et al., 1971), while others have found that this response is enhanced (Schimmelbusch, Mueller and Scheps, 1971). Studies of prolactin levels are more consistent, showing raised prolactin both in experimental animals and in human subjects following administration of phenothiazines (Apostolakis et al., 1972; Hwang et al., 1971; Kleinberg et al., 1971; Sulman, 1970).

In a number of previous papers in this series (Roth et al., 1972; Gurney et al., 1972; Kerr et al., 1972; Schapira et al., 1972) evidence was adduced to suggest that depressive illnesses and anxiety states could be clinically differentiated from one another. It was shown also that with the aid of appropriate statistical techniques patients exhibiting these disorders could be separated into two distinct groups with relatively little overlap. The possibility that these findings were due to diagnostic preconceptions or other forms of bias was excluded by the results of follow-up studies conducted over an average period of 3.8 years in which the outcome of the entire patient population was evaluated by independent observers. The clinical picture found was consistent over the period of observation, with little diagnostic interchange between the groups. Moreover, the outcome of the depressions was significantly better than that of the anxiety states, and this provided external and independent validation of the view that the groups initially separated with the aid of clinical observation were indeed distinct from one another.