census, which in many of its inquiries seems to verge close on the impossible or the impracticable. Besides, as has lately been pointed out by Westergaard, imperfect data are not necessarily valueless, for the man of science by skilful manipulation and the exercise of the scientific imagination may out of a fragment construct a probable whole; or he can at least say what a thing is not, even when he cannot say what it is. The temperature of the water may enable a ship captain to steer clear of an iceberg whose exact size and position is entirely unknown. Nor must we lay too much stress on the common objection that these imperfect data by mere governmental publication receive the stamp of official approval, and thus attain an authority and a currency which they do not deserve. It is of course true that any census figures beyond a bare enumeration of the people may be so interpreted or misunderstood or manipulated as to furnish a basis for wild theories and absurd, perhaps destructive, proposals. But common sense, in the popular meaning of that term, exercises a restraining influence; for persons are not apt to accept that which is contrary to their ordinary experience unless it rest on very clear and substantial grounds, and nothing is so fatal to a theory as to show that it has been built up on false and mistaken premises. The same figures, however, are accessible to the opponents as well as to the advocates of any social reform; while the educational value of these attempts to throw the light of exact statistical investigation on social phenomena otherwise unknowable is very great, both for the practical statistician and for the large number of persons who in an intelligent democracy try to comprehend and solve social problems.
But besides these general considerations (which I cannot expect will meet with general assent, although they make us in the States view the extent and expense of our census work with complacency), there are certain peculiarities connected with the United States census which it is useful to understand in order to judge of the results. These are briefly: (1) that, owing to the general intelligence of our rural population certain inquiries can be successfully prosecuted here which could not be in Europe; (2) that, owing to the peculiar make-up of our population certain inquiries unimportant elsewhere are very important here (e.g. in respect to the coloured and foreign-born population); (3) that, owing to our federal form of government, certain inquiries are forced into the census which do not rightly belong there (e.g. mortality statistics, local indebtedness, etc.); (4) that, owing to political reasons, certain inquiries have been attached to
