Data on the fertility of mothers and daughters in 10,931 mother–daughter pairs were obtained from family reconstitution studies based on English parish registers between the sixteenth and nineteenth centuries. These data indicate that there is no association between the fecundity of daughters and that of their mothers.

Shortcomings of the illegitimacy ratio as an index of the level of non-marital childbearing in New Zealand are exposed. Trends in the level of bridal pregnancy are analysed in terms of changes in age-specific probabilities of conceiving a child outside marriage, of ‘regularizing’ such a pregnancy before confinement, and of marrying when non-pregnant. Finally, trends in childbearing resulting from non-marital pregnancies are examined using an index which allows all relevant pregnancies to be related to a single population at risk, and which can be partitioned according to the eventual status of confinements as marital or non-marital. A substantial decline in such childbearing during the 1970s is established and attributed mainly to greater use of abortion. Marital births resulting from non-marital conceptions declined especially sharply.

Breast-feeding is an integral part of the reproductive process. This paper reviews the consequences of breast-feeding and bottle-feeding for the child and the mother with emphasis on health parameters and child spacing. Controversies that have arisen in this field are considered. Methodological problems in study design and areas for further research are discussed.

The Gainj of highland Papua New Guinea do not use contraception but have a total fertility rate of only 4·3 live births per woman, one of the lowest ever recorded in a natural fertility setting. From an analysis of cross-sectional demographic and endocrinological data, the causes of low reproductive output have been identified in women of this population as: late menarche and marriage, a long interval between marriage and first birth, a high probability of widowhood at later reproductive ages, low effective fecundability and prolonged lactational amenorrhoea. These are combined with near-universal marriage, a low prevalence of primary sterility and a pattern of onset of secondary sterility similar to that found in other populations. Of all the factors limiting fertility, by far the most important are those involved in birth spacing, especially lactational amenorrhoea.

Data collected as part of the World Fertility Survey programme in the mid-1970s in four Muslim populations (Bangladesh, Java (Indonesia), Jordan and Pakistan) show that the demographic variables age at first marriage, duration of marriage, status of first marriage and experience of child loss explain most of the variations in fertility among these populations. There was no consistent pattern which could explain fertility differentials by selected socioeconomic variables. Fertility differentials by wife–husband education and childhood–current residence were found not to be in the expected direction in most of the populations. Fertility transition has not yet started among the majority of the people in these populations.

The trends of intermarriage, and the attitude to it, of Koreans in Japan were examined. About 47% of total marriages of Koreans during the period 1965–79 comprised a husband and wife of different nationality. The proportion of Korean–Japanese marriages was higher than that of Korean–Korean marriages in recent years. Attitudes of Koreans in Japan are likely to favour intermarriage with Japanese. The results suggest that mixed marriages of Koreans and Japanese will continue to increase and may influence the future fertility of Koreans in Japan.

Lactating women are at reduced risk of conception, but the individual woman may not wish to accept even a reduced level of risk. The risk increases when infant feeding supplementation begins and an alternative form of contraception is then required for the lactating mother. The relationship between lactation and fertility, trends in prolonged lactation, factors affecting these trends, and contraceptive use during lactation are reviewed. Suggestions for advising individuals and formulating programme guidelines are discussed.

This study, based on the Nepal Contraceptive Prevalence Survey carried out in 1981, is the first of its kind. Analysis using a logit-linear model showed that the prevalence of current contraception was directly related to number and sex of living children, and the age of the woman. The prevalence rates are higher among women who did not want more children and women who had discussed family size with their husbands. The prevalence varies according to women's education and work-status. Urban women have a higher prevalence than rural women. Women who have access to the family planning service within a short distance have higher prevalence than those who do not.

Using a unique set of birth registration data from the Demographic Surveillance System of the International Centre for Diarrhoeal Disease Research, Bangladesh, for the period 1974–77, and socioeconomic information collected in the 1974 census, fertility was studied in relation to occupation, size of dwelling, number of cows and number of boats owned. The total fertility rate was found to vary between 6 and 6·5 except in the famine year of 1975. There was no consistent relationship between fertility and education of women. Fertility differentials by occupation showed that the household heads who were farm labourers had relatively lower fertility compared to other occupational groups, except for the year 1977 where the families of service holders were found to have relatively lower fertility. There was a consistent direct relationship between the dwelling size and fertility for each of the years.

Surveys carried out in Lahore, Pakistan, in 1963 and 1980 showed that most women have some knowledge of contraception. The proportion who had ever used contraception rose from 7% of respondents in 1963 to 48% in 1980. The rate of current use in 1980 was 34%. Highest rates of use were found among women aged 30 years or older and among those of higher education and socioeconomic status. The most commonly used methods of family planning were the condom, the pill and abstinence.

Among obstetric patients with at least one previous pregnancy, over 20% reported having at least one induced abortion. Abortion experience was more common among women at lower gravidities and among women with the highest levels of education, suggesting that abortion is most used to delay the first birth, particularly for young women still at school. This is in contrast to the situation in Latin America where abortion is used most often to limit family size.

The proportion of women who reported that they used contraception prior to the pregnancy was low, suggesting that knowledge of attitudes towards and availability of family planning services needs to be improved.

Using retrospective data from Korea, multi-variate log-linear analyses indicate that the sex of a child influences the probability of it being breast-fed. Even for women with a living son, there are greater than average odds that they will not breast-feed an infant girl. The sex of the child did not, however, influence the number of months he/she was breast-fed.

At a time when women's participation in developing countries and their breast-feeding practice are the objects of promotional campaigns, findings on the relationship between the two are required for formulating policies. Empirical findings from the 1973 National Demographic Survey in the Philippines strongly suggest the adverse effects of modernization on both the initiation and duration of breast-feeding. In this study a high level of education and working outside the house are found to be the modernization variables that are most effective in depressing the level and duration of breast-feeding practice, independent of the influences of other socioeconomic and demographic factors.

A new set of abridged life tables for Pakistan is presented. Data from the 1971 Population Growth Survey were first analysed to estimate the degree of completeness of the reporting of male and female deaths; female deaths were substantially more under-reported than male deaths. Age–sex specific mortality schedules were adjusted accordingly.

Life expectancy was around 50 years at birth, but increased by 8–9 years for those surviving the substantial risks of dying in the first year of life. No significant sex differential in mortality could be discerned from the life tables.

Between 1981 and 1982 the Royal College of Obstetricians and Gynaecologists undertook a national survey of late abortion practice in England and Wales. The study was based on a confidential enquiry, specifically authorized by the Chief Medical Officers of the DHSS and the Welsh Office, which used the notification system as the means of identifying cases. All abortions notified as carried out at 20 weeks gestational age or more were selected during the year, together with varying proportions of cases from other gestational groups. Data on avoidable and unavoidable causes of delay were obtained by questionnaire from notifying doctors, and a comparison was made of the subjectively attributed delay with any that was found through objective measures derived from both questionnaire and notification data. This paper presents results on delay associated specifically with abortions carried out at a gestational age of 20 weeks or more to residents of the United Kingdom only.

In recent years both the prevalence and duration of breast-feeding have decreased, especially in urban areas of the developing world (McCann et al., 1981). While the majority of births in these areas continue to be attended by traditional birth attendants, the proportion of mothers giving birth in hospitals is increasing. The potential for hospital personnel and practices to affect infant feeding preferences is likely to increase as the trend towards hospital births continues.

This paper examines secular changes in the influence of maternal age, parity and social class on perinatal mortality in Scotland. Despite changes in reproductive behaviour, improved access to maternity services and more effective perinatal care, the influence of maternal age and social class on perinatal mortality remained unchanged between 1960 and 1982. Although there were no significant changes in the traditional J-shaped association between parity and stillbirths, cross-sectional analysis shows that in the early 1980s the risk of both neonatal and perinatal deaths decreased as parity increased: this finding is consistent with the pattern of risks observed in longitudinal studies and retrospective surveys of reproductive histories. The stability of age, parity and social class effects thus suggests that little if any of the overall decrease in Scottish perinatal mortality can be attributed to a significant narrowing of relative risks in the obstetric population.

This study employs multiple classification and regression techniques to determine the influence of marital, sociodemographic and selected intermediate variables on duration of breast-feeding and post-partum sexual abstinence among a sample of 300 Yoruba women of western Nigeria. Only education and place of residence appeared to be significantly related to the post-partum variables. Breast-feeding was shown to exert a significant positive effect on abstinence. The effect of contraception on the post-partum variables appeared to be a function of socioeconomic status.

Reductions in the duration of the post-partum variables due to relative affluence can result in increases in fertility. Planners may do well to encourage women to go back to the use of these traditional child spacing practices in the absence of large scale and effective use of contraceptives.

Skin colour is measured at the inner upper arm and forehead by reflectance spectrophotometry in 310 Punjabi Hindu spouse pairs. It suggests genetic similarity of the two subcastes Khatri and Arora, of Punjabi Hindus, and that caste formation is essentially a social phenomenon. There is evidence of positive assortative mating with respect to skin colour.

A method is presented for examining the relationship between effective population size and accumulated random inbreeding in human populations. For a set of populations, the inverse of inbreeding is regressed on effective population size using a linear regression model. This procedure allows testing of several hypotheses regarding the common and unique influences on population structure. Deviations from the expected curve suggest demographic or historical change. This method is applied to surname data from nine Irish isolates. The results show that the method is very useful in assessing differential influences on population structure.