fo MIGRATION AND BUSINESS CYCLES 
for aliens of both sexes combined, as shown by the cross-hatched 
bars, and, secondly, for male aliens only, as shown by the circles. 
For both males and all aliens, the net movement is small in 
January, increases in February, reaches high tide in March, April, 
and May, ebbs somewhat in midsummer, recovers to a secondary 
peak in September and October, and then slumps rapidly in Novem- 
ber and December. 
TABLE 55.—PRE-WAR SEASONAL DISTRIBUTION OF THE NET MOVEMENT 
OF MIGRANTS 
(Thousands of persons) 
NET ALIEN MOVEMENT (ARRIVALS NET MOVEMENT OF MALE ALIENS 
LESS DEPARTURES), BOTH SEXES (ARRIVALS LESS DEPARTURES). 
MonNTH Jury, 1907, To JUNE, 1914. Jan. 1, 1910, to Dec. 31, 1913. 
TOTAL IN AVERAGE FOR ToTAL IN | AVERAGE ror) 
7 YEARS GIVEN MONTH 4 YEARS GIVEN MONTH 
January... .... 57.5 8.2 22.1 "55 
February... ... 207.5 29.6 91.1 22.8 
March. ....... 530.1 75.7 253.6 63.4 
April... 593.5 84.8 294.4 73.4 
Vis irih a on 517.2 73.9 | 252.7 63. 
June... 354.5 50.6 | 205.1 51} 
aly i ss 271.6 38.8 105.0 25.9 
August, ....... 312.4 44.6 117.7 29.4 
September. . . .. 429.2 61.3 147.9 37.0 
Déiober. ix.» 428.7 61.2 126.9 31.7 
November. . ... 243.6 34.8 26.6 6.7 
December. . . . . 124.0 17.7 421.1 45.3 
aComputed from statistics in the 1907 to 1914 issues of the monthly bulletin of the U. S. Bureau of 
Immigration and Naturalization. 
dExcess of departures over arrivals. 
The data under consideration represent, strictly speaking, the 
combined effect of seasonal and trend tendencies. The years before 
the war were a period of increasing immigration; and that the net 
immigration in December, for example, is larger than in January is 
due in part to the rising trend. If the influence of the trend factor 
were eliminated, the proportion assigned to the earlier months 
would be relatively increased and that in the latter months de- 
creased, so that the drop at the end of the year would appear even 
more decided than it does in Chart 51. As we shall presently note 
more fully, the crude seasonal distribution as given in Table 55 is 
appropriate for comparisons with changes in employment, and for 
this reason, and also because of technical difficulties in the way of 
220