f the Census) are based on the average increase in the
preceding census decade. The increase between 1910 and 1920, for
instance, is divided by 120, the number of months in the period, and
this average monthly increase is assumed to be the same as that of the
current year and of other years following 1920. Until new figures are
obtained in 1930, the monthly increase will continue to be estimated at
the same rate as the increase from 1910 to 1920, or about 118,000. This
figure will be used in computing the per capita coffee consumption. But
when the 1930 figures are in, it may be found that the estimates were
too low or too high, and the per capita figures for all intervening
years will accordingly be subject to revision. This will not amount to
much, probably five-hundredths of a pound at most; but it is evident
that between 1920 and 1930 all per capita consumption figures issued by
the government are to be considered as provisional to that extent at
least.

In the 1920 _Statistical Abstract_ the government has revised its per
capita coffee and tea figures to conform to actual instead of estimated
population figures between 1910 and 1920, with the result that these
figures are slightly different from those published in previous editions
of the _Abstract_. Figures from 1890 to 1910 have also been slightly
changed, as they were originally computed by using population figures as
of June 1, whereas it is desirable to have computations based on July 1
estimates to make them conform to present per capita figures.

_Reviewing the 1921 Trade in the United States_

According to the latest available foreign trade summaries issued by the
government, the United States bought more coffee in 1921 than in any
previous calendar year of our history, although the total imports did
not quite reach the highest fiscal-year mark. Our purchases passed the
1920 mark by more than 40,000,000 pounds and were higher than those of
two years ago by 3,500,000 pounds.

But this record was made only in actual amounts shipped, as the value of
imported coffee was far below that of immediately preceding years.
Coffee values, however, fell off less than the average values for all
imports, the decrease for coffee being forty-three percent and for the
country's total imports fifty-two percent.

Exports of coffee were somewhat less in quantity than in 1920, and about
the same as in 1919; although the value, like that of imports, was
considerably less t