The comment is often made that Canada has no war labour policy. Before dealing with this question, it may be well to consider some of the difficulties that attend any effort to determine and administer a national labour policy, especially in time of war. The problems of geography, uneven distribution of population, depressed areas, concentration of industry, and varying standards of living, together with the many revisions of policy necessitated by an ever-changing war need, can only be mentioned, but they constantly intrude on plan and practice. Inertia, easy assumptions that the ways of peace need not be changed materially and that the methods of the last war are good enough in this crisis, put a brake on the formulation of plans and their prompt execution.

There are Canadians who still do not realize, as Mr. Menzies, the Prime Minister of Australia, said, that the British people are engaged in the deadliest war of their history. Some employers and workers do not appreciate the proportions of the war programme or the necessity for many departures from peace-time procedures. Certain firms still feel that they will be able to secure all the labour they require and respond but slowly to appeals to train more workers, to thin out their skilled staffs, and to up-grade labour. Although the Minister of Labour has received a number of voluntary offers from trade unions to work longer than the usual hours at straight time, there are those who hold that the forty-hour week, urged in depression years as a device for spreading work rather than increasing earnings through higher overtime rates, should be maintained.

Measurement of changes of marriage frequency involves peculiar difficulties. To some extent the trend of crude birth and death rates reflects the trend of more refined indexes. A corresponding statement is not even approximately true of crude marriage rates. The crude rate may be either rising or falling, while the true rate remains constant, and vice versa. Though this fact emphasizes the need for accurate analysis, comparatively few computations of true marriage rates exist. One reason for this is probably that the marriage rate has fluctuated widely in the past from year to year in response to economic changes, without showing any very definite secular trend. So the relation between marriage frequency and reproduction rates has not seemed to be highly significant. However, the position is now changing. On the one hand, marriage is less inevitably followed by reproduction. On the other hand, most marriages are still followed by first births, and first births form an ever-increasing proportion of all births. Hence fluctuation in marriage rates is reflected to an increasing degree in fertility rates.

It is now a generally accepted view that the best method of obtaining true marriage rates is to draw up nuptiality tables in accordance with the principle on which life tables are constructed. The true marriage rate is then given by the proportion of women who marry at least once out of a thousand girls alive at 15 years of age (gross nuptiality), or out of a thousand new-born girls (net nuptiality). The present paper presents nuptiality tables for Canada and each of the nine provinces for 1930-2, and for Saskatchewan for 1926 and 1936. It is hoped that similar tables will be calculated for more recent years, when the data become available.