Date from the Census Fertility Reports are used to investigate social and regional variations in family size in the Republic of Ireland. Although Ireland is noted for its high level of fertility, average family size declined by approximately 10% between 1946 and 1971. There are distinct socioeconomic variations in family size in that Roman Catholic family size is greater than that of non-Catholics and the middle classes have the smallest families within each of the religious groups. There are also marked regional differences in family size. In 1946, there was a clear division between the west coast counties with the largest families and the east coast conties with the smaller families. By 1961 and 1971 this pattern had become less clearly demarcated and areas with relatively large families were found in most parts of the country. It is suggested that regional variations in family size can be influenced by a variety of factors, such as differences in religious affiliation, occupation, infant mortality and the degree of selection through emigration. Of these factors, it is found that in both 1946 and 1971 variations in occupation were most strongly related to family size and it appears that economic and occupation changes within Ireland have brought about the changing regional pattern of family size.

Summary From a survey conducted in the Irish Republic, data on ideal family size are given. Irish wives have high family size preferences, the overall mean ideal family size being 4.3 children. The Irish data are compared with American and western European; they show that the ideals of wives in Ireland are significantly higher than in these other countries. The concept of ideal family size appears to possess validity in its own right, and is not solely a rationalization of actual fertility experience.

Among a sample of 601 married men, whose wives are in the reproductive ages, religious affliation is strongly related to opinions about the permissibility of abortion under varied circumstances and Canadian abortion law. Education and other influences are altered when the effects of religion are included but some differentials persist within religious groupings. These suggest that secular trends towards increasing levels of education may tend to narrow religious-based differentials in abortion attitudes. An association of permissive attitudes towards abortion with use of highly effective contraceptive methods and with prior use of abortion is found.

Menstrual regulation (MR) was introduced as one of the components of the village-based family planning and meternal and child care programme initiated by the Cholera Research Laboratory, Dacca, in 70 villages in Matlab thana in October 1977. Social and demographic characteristics of the MR clients are compared with those of the users of other contraceptive methods offered by the programme. Reasons for pregnancy termination and the high continuation rates of contraceptive use after MR suggest that, in particular, the younger women are becoming increasingly conscious of controlling their childbearing and spacing their pregnancies.

This paper examines the current use of contraception among a sample of 6646 married women of reproductive age interviewed during 1972–73 in metropolitan and country areas of the state of New South Wales, Australia. The analysis shows that contraception is widely practised in all segments of the population, though to a lesser extent among the Catholic women and those born is southern Europe. In conjunction with now easily available induced abortion, this has resulted in a precipitous drop in the annual number of marital births.

This paper examines the continuation from 1971 to 1975 of the fertility decline in Sri Lanka that commenced in 1960. New trends in contractive practice, especially of female sterilizations in 1974–76, are discussed, and change in the pattern of childbearing is identified. The fertlity decline is compared with that in Taiwan.

Infand mortality for the period 1950—70 has been analysied in a North-Swedish isolate, a population characterized by a high incidence of schizopherenic psychoses and other neuorpsychiatric disorders. Infand mortality is higher than in Sweden generally. As many as 45·9% of the deceased infands in the population sccumbed during the first week of their life, compard to 28·1% in the general Swedish population.

Perinatal mortality among hospital deliveries in Lucknow city in 1976–77 is compared with the results of earlier studies. The effect of biological and socioeconomic factors is considered. High risk cases, for which special care should be provided, are noted.

Data are presented on age at menarche and adult stature in 905 women from the islands of Lewis and Harris, born between 1920 and 1949. Menarcheal age showed no tendency to fall in women born between 1920 and 1935, but thereafter successive year-of-birth groups showed an almost continous decline. There were no significant changes in stature in successive groups.

The changes have been correlated with trenda in affluence, as indicated by unemployment levels. In women born between 1932 and 1941 there was good agreement between average unemployment during various periods of childhood and menarcheal age, but among later-born individuals the relationship broke down. It is suggested that during the earlier period unemployment was associated with some degree of biological deprivation, but that during the later period the introduction of rationing and later, of the Welfare State, buffered the growing child against economic hardship.

Mean household size (used as an estimate of mean sibship size) fell over the period of study, but its pattern of change was different from that of menarcheal age. It is suggested that this factor played at most aminor role in fluencing the results of the investigation.

In an earlier paper (Abelson, 1978) it was suggested that relationships between population density, migration, and the frequency of consanguineous marriage depended on variation in the social class composition of populations. For two valleys in the western Pyrenees, it was found that migration distances are shorter in areas with lower population density and that this could in part be attributed to the effects of social class independently of the occupations of the population (Abelson, 1979). The present paper investigates further the relationships between migration, the frequency of consanguineous marriage and inbreeding in the Roncal valley in Navarre, Spain, for the period 1877–1915. The Roncal valley contains seven villages, each a separate administrative unit. Almost all the total population of 3900 live in these nucleated villages and there are few isolated households. The area of the valley is 420 km2, and the population density 9·28 inhabitants/km2.

The Sherpas of Upper Khumbu (above 10,000 feet), Nepal, and their migrant counterparts at lower altitudes (3000–7000 feet) in Darjeeling District, West Bengal, show no statistically significant difference in age–sex structure or mortality, while there are suggestions of some fertility reduction at the higher altitudes.