In the present study an attempt has been made to determine the relative importance of several factors on the mortality rates of infancy. In view of the differences in the rates at the several ages under one year and of the differential decline in these rates within recent years, it has been suggested that for any complete consideration of the death rates in infancy, some subdivision of the rate, either of a biological or temporal character, must be effected, and each of these subdivisions examined separately. The infant mortality rate has been treated in three broad categories, namely, the antenatal, neo-natal and post-natal death rates; and it has been shown that each of these depends on different factors.

Considered as death rates per unit of time the neo-natal rate is much higher than at any other period of life. The ante-natal death rate is much lower than this, and the post-natal rate lowest of all.

In relation to these rates, the influence of three, presumably important, factors have been measured. These are:

1. The provision of skilled attendance to mothers in childbed.

2. The health of the mothers.

3. Environmental and social conditions.

The indices used as measures of these three variables and their limitations have been pointed out. The author feels only too conscious of these limitations and, with such vital statistical data as the only indices available of the factors the influence of which on the mortalities of infancy it is desired to measure, it is probable that any conclusions drawn from such an analysis should be accepted with some degree of caution and reserve. Nevertheless. I am firmly convinced that even these (to the clinician and administrator) rather crude results are immeasurably superior to expressions of opinion, unsupported by facts, so frequently found forming the basis of much of the medical literature on the subject. That we have so relatively little exact quantitative knowledge of the influential factors concerned in the growth, nutrition and mortality of infants is surely not for lack of opportunity of collecting the requisite data. An admirable machinery for such a purpose—the various maternity and child welfare centres—is in extensive operation throughout this country; and if the medical officers in charge of these centres could only be persuaded to collect and analyse (or have analysed) periodically the data contained in suitably framed and completed questionnaires, many questions of scientific importance as well as of the efficacy of administrative efforts would soon cease to be subjects of controversy.

Our results show that, of these factors, only the provision of skilled medical assistance to mothers in childbed is of importance in connection with antenatal mortality. The general health of the mother and external environmental conditions have no direct influence on the death rate at this period of life. The magnitude of the correlation leads to the further conclusion that a great part of the still-birth rate is not to be controlled by improvements in even all of these factors.

The neo-natal death rate is related both to variations in external environment and in the obstetrical assistance available for mothers in childbed. At this state of life, again, the health of the mother has no substantial influence on the death rate. External surroundings and the quality of the obstetrical assistance afforded to the mother are of approximately equal importance in determining these rates; but here again the conclusion is advanced that the greater part of the death rate at this stage of life is beyond human efforts of control.

The post-natal death rate seems to offer the greatest scope for administrative measures. In this case the health of the mother would appear to come first in order of importance, environment also is of some importance, whereas the effects of variations in obstetrical services have now ceased to be reflected on the mortality of infancy.

In isolated rural communities there is a part of the post-natal death rate which could be eliminated by the speedy arrival of skilled medical assistance.

Foreign statistics have been collected and analysed to enquire into the causation of still-births. An obvious defect in these was pointed out, namely, the large proportion of deaths due to unknown causes. A comparison of these figures with those collected from several maternity hospitals has shown that hospital figures give a biassed view of the problem. The proportion of deaths due to causes in which the life of the mother as well as that of the foetus is endangered is greatly magnified, and consequently will produce a too optimistic estimate of the amount of amelioration which will result from increased attention to pregnant and parturient women. The statistics which have been given in this section show that in a really random sample of the population only a small proportion of still-births is the result of causes which can be controlled by this method, and that the chief causes of ante-natal death are developmental defects of the foetus and prematurity.

A short study has been made of certain statistico-anatomical features of the foetus and infant. From this it has been inferred that the foetus is physically in a state of extreme instability, that the conditions which kill in utero are rapidly fatal to the foetus, and that this instability is a somewhat more prominent feature in the male than in the female. As infancy advances, visceral variability decreases steadily, so that it would seem probable that some time is necessary for the foetus to adapt itself to its new mode of existence after birth. This state of unstable equilibrium in early life may provide some reason for the higher mortalities at these ages.