IN two former papers the crop estimates and reports furnished by the Department of Agriculture have been analyzed. They have been carefully compared with the returns furnished by the census and have been found to differ widely from the latter. In order to secure a basis for trustworthy judgments concerning the relative worth of these statistics, it therefore became necessary to review the processes employed in compiling them, by the two offices under investigation, and to point out possible sources of error. Incidentally, it was found expedient to detail the facts of the inquiry made in the autumn of I902 by representatives of the National Board of Trade. From these analyses and investigations the conclusion was reached that, whatever may be the statistical errors of the census, it is certainly much nearer the truth in its recent crop figures than is the Department of Agriculture; and that the figures of the latter office are not only so wide of the mark as to be clearly absurd, but contain many inconsistencies, contradictions, and absolute impossibilities. Worst of all, it was shown, certain features of these statistics give the impression of arbitrary alteration, made possibly with a view to greater correctness, but none the less confusing in the absence of explanation. What is the cause of the unfortunate and discreditable statistical situation outlined in these papers ? To whom is the responsibility for it chargeable ? How can the grouLnd be cleared for better work? And what should that work be? Should the government crop reports continue on their old lines, or should some limitation or alteration of their scope be effected? These questions are of pressing moment. They cannot safely be neglected in view of the great and growing business of our produce exchanges. Yet they can be answered only by a careful study of certain political and quasi-political influences which are not fully understood; and by an analysis of some of the inner

discrepancy was found in the returns for every principal agri? cultural crop in the United States. It, moreover, has been pointed out that the student is apparently compelled to believe in the superior accuracy of the census figures as compared with those of the Department of Agriculture, unless satisfactory evidence going to discredit the work of the census office, and to throw unexpected shadows upon its methods, can ?>e produced. We have seen, too, that during the months immediately suc? ceeding the publication of the reports of the twelfth census, friends of the Agricultural Department sought to cast reflections upon the work of the census. That the Agricultural Department should feel annoyed at the errors in its statistics which seemed to be revealed with the