their minds and leave after a short stay, or that many alien residents upon leaving declare an intention of only a temporary sojourn abroad but do not return to this country as they had intended.

Variations in Comprehensiveness.

In addition to the above distinctions among citizens, immigrant aliens, and nonimmigrant aliens, it is necessary to note that the official statistics of immigration have not always been equally comprehensive and do not have exactly the same connotation throughout the period of a little more than a century for which they are available. The footnotes to the tables contain much of the detail concerning the varying comprehensiveness. For example, the data which are officially published for the annual totals of immigrations cover, prior to the fiscal year beginning July 1, 1867, all recorded arrivals of aliens, without discrimination as to length of intended residence; for the period ending June 30, 1903, they cover immigrants as differentiated from nonimmigrants; for the next three years, "aliens admitted" (though apparently this does not actually include nonimmigrants); and for subsequent years, only immigrants admitted. In brief, to make the data strictly comparable throughout the entire period, it would be necessary to make allowance for the inclusion or non-inclusion of, first, nonimmigrants, or those with announced intention of temporary residence only; and second, would-be immigrants debarred from entry.

Also, there are variances due to the circumstance that in the early period there was no attempt made to record residents of adjoining countries, Mexico and Canada, as they passed over our boundaries. In fact, the early records of the number of Europeans entering the United States via Canada are quite imperfect. For the period from July, 1885, to October, 1893, the statistics entirely omit such arrivals through Canada. Subsequent to October, 1893, the records include foreigners arriving at Canadian ports en route to this country, and more recently, they also include residents of Canada and Mexico who pass our boundary lines for a stay of six months or longer, although they are not counted as immigrants unless a stay of a year or more is intended.

The above-mentioned discrepancies in the official statistics of immigration, while significant for some purposes, are not important enough when we are primarily interested in cyclical fluctuations to United States Immigration Commission, Statistical Review of Immigration: 18201910, p. 4.

prevent us from treating the series as reasonably homogeneous throughout the entire period for which the official statistics are available.

Fiscal and Other Non-calendar Immigration Years."

Through most of the period included in our immigration records, the year covered by the officially published annual statistics does not coincide with the calendar year. For the years 1820 to 1831, inclusive, the annual immigration statistics refer to the twelve months ending September 30th of the given year; for 1833 to 1842, inclusive, the immigration and calendar years coincide; for 1844 to 1850, the immigration year again terminates September 30th; for 1851 to 1856, the year ends December 31st; and beginning with 1858 and continuing until the present time, the official immigration year ends June 30th.

We shall use the term fiscal years for twelve-month periods which end on June 30th. To illustrate, the phrase "in the years 1863 to 1892 (fiscal)," means from July 1, 1867, to June 30, 1892, inclusive. Non-calendar years not ending on June 30th will be appropriately indicated.

PERTINENT FEATURES OF MIGRATION TO AND FROM THE UNITED STATES

Violence of the Major Fluctuations.

The significance of the facts revealed by the subsequent analysis of the quarterly and monthly statistics of migration will be clearer if we first make a preliminary survey of the larger movements in the flow of immigration. In Chart I, we have a curve representing the volume of immigration in each year in a period of slightly over a century, beginning with the year ending September 30, 1820.7 The picture is one of successive waves, the crest of the major waves occurring, respectively, in 1854 (calendar year) and 1873, 1882, 1892, 1907, 1914, and 1921 (fiscal years). In each case the following decline corresponds approximately to a period of industrial depression in this country. We shall return in later chapters to a closer scrutiny of these relations.

While the general sweep of the curve is upward until checked by the Great War and the restrictive conditions of the post-war period,

❝See Table 1.

?For the extent to which these data are not strictly comparable throughout the entire period, see the earlier section in this chapter entitled, "Variations in Comprehensiveness."

the fluctuations in volume are relatively so large that the trend, particularly for anything short of very long periods, is to a large extent obscured by the violence of the fluctuations. Moreover, on closer analysis, it is found that some major elements in migration have been declining while others were increasing in number. In all, the magnitude of the major fluctuations in immigration has led us in some instances, particularly where short periods are under consideration, to analyze the data without attempting to eliminate

CHART 1

FLUCTUATIONS IN THE NUMBER OF IMMIGRANTS, BY YEARS:

1,250,000

1,000,000

1820-1924.

1820-29 30-39 40-49 50-59 60-69 70-79 80-89 '90-'99 1900-09 10-19-24 Numerical data in Table 1.

whatever trend may be present; and for long-period studies, in order to bring out clearly the current alternations in prosperity and depression, trends have been computed by the flexible method of the moving average, which tends to eliminate the effect of the larger swings such as the general decline from the early eighties to the late nineties as well as the general upward trend of immigration. The best data for comparing migration and industrial conditions apply to the years subsequent to 1890 and particularly to the period from 1907 to 1923. Obviously, however, the direction and degree of a significant trend throughout this period is largely a matter of conjecture.

TABLE 1.-OFFICIALLY RECORDED IMMIGRATION INTO THE UNITED STATES:

■ Compiled from the U. S. Bureau of Immigration, Annual Report of the Commissioner General of Immigration, Fiscal Year Ended June 30, 1924, p. 122.

bFor 1820 to 1867, these statistics pertain to "Aliens Arriving," including that class of arrivals later designated as nonimmigrants.

In these periods the available statistics cover other than twelve-month periods. In the fifteen months from October 1, 1831, to December 31, 1832, 60, 482 alien arrivals are recorded; in the nine months from January 1 to September 30, 1843, 52,496; in the three months from October 1 to December 31, 1850, 59,976; and in the six months from January 1 to June 30, 1857, 112,123.

dFor the fiscal years 1868 to 1903, inclusive, these statistics are designated in the official publications as including "Immigrants Arriving;" for the years 1904 to 1906, inclusive, "Aliens Admitted;" and from 1907 to date, "Immigrant Aliens Admitted." However, it would appear from other data given in the reports of the Commissioner General of Immigration that in all years after 1867, the statistics given in the above table do not include nonimmigrants.

The numbers of immigrants as compiled by the U. S. Bureau of Statistics for these years are: 1892, 623,084; 1893, 502,917; 1894, 314,467; 1895, 279,948. Monthly Summary of Commerce and Finance, June, 1903, p. 4364.

Marked Seasonal Variation.

An examination of the quarterly and monthly data on immigration reveals a marked, and, on the whole, a regular seasonal variation. This is clearly evidenced by Chart 2. The upper section of

CHART 2

THE MARKED SEASONAL MOVEMENT IN IMMIGRATION: 1885-1924

Ratio Scale

MALE IMMIGRANTS BY QUARTERS

200,000

100,000

50,000

20,000

1885 1886 1887 1888 1889 1890 1891|1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904

this chart presents the quarterly data for 1885 to 1904 (calendar years) showing invariably a relatively large immigration in the second quarter comprising April, May, and June. In the lower section of the chart, covering the period 1905 to 1924 by months, a similar marked seasonal variation appears prior to the Great War. During and immediately after the war, the seasonal is somewhat distorted and subordinated. After 1921, the influence of the per centum limit law, permitting twenty per cent of the admissible quota to enter in any one month, beginning in July, has caused the