SEASONAL FLUCTUATIONS nN 
satisfactorily determining the trend, we have made no attempt to 
adjust for the growth factor in presenting the statistics of seasonal 
fluctuations in net migration. 
We have noted the characteristic features of the seasonal move- 
ment in the main migratory currents to and from the United States 
before restrictive legislation intervened to modify the seasonal dis- 
tribution; let us now note the corresponding seasonal movements 
in the major occupations in which immigrants engage. 
SEASONAL TENDENCIES IN SELECTED INDUSTRIES 
The great bulk of newly arrived immigrants, as we have previously 
noted, are engaged in manufacturing, coal mining, construction, 
and railway maintenance; hence it becomes desirable to examine 
the seasonal fluctuations in employment in these industries, in order 
that we may determine whether the seasonal variations we have 
observed in migration synchronize closely with changes in employ- 
ment. 
The data available for measuring seasonal variation in these 
several industries are so fragmentary and diverse in nature, that 
we feel impelled to give first some explanation of the nature and 
limitations of the evidence from which our indices are constructed. 
We then proceed to a comparison of these seasonal employment 
indices with the corresponding indices for migration. 
The evidence considered in arriving at our estimates for the 
several industries is shown in Charts 52 and 53; and the numerical 
indices are given in the accompanying tables. The final estimates 
for comparison with fluctuations in migration appear in Charts 54 
and 55. In comparing these charts, the reader should note that 
the scales used in plotting have been varied so as to magnify the 
fluctuations for some series, such as factory employment, so that 
the changes will stand out more clearly. 
Factory Employment (Chart 52, Fig. A). 
In the process of testing for typical seasonal variation in factory 
employment, we computed two indices, both of which appear in 
Chart 52, Fig. A. The first, Curve (a), is based upon our estimates 
of factory employment in Massachusetts, New Jersey, and New 
York. The second, Curve (b), is based upon the Census of Manu- 
factures statistics of factory employment in the United States in 
the census years 1904, 1909, 1914, 1919, and 1921, and upon the 
g. 
7