I have had occasion to make several studies of this kind. In order to inform myself on the potential of the South in the production of cotton, I undertook a study of the physical geography and climatology of the cotton States and of other cotton-producing countries nearly forty years ago. The results of this research were first given in Cheap Cotton by Free Labor, published in 1861. In that pamphlet and in many treatises following, finally in an address in Atlanta, in 1880, a true forecast or "speculation" or "intellectual examination" will be found of the production of the cotton fiber, the potential of the future and of the cotton-seed-oil industry, then almost unheard of in this country. In 1880 I also entered upon my first "speculation" (not in the market) on the lines of a "contemplation" or forecast of the effect of agricultural machinery applied to our wheat land, coupled with the prospective reduction in the cost of carrying wheat to England, upon the condition of the American farmer and the British landlord. That forecast of prosperity to our farmers in the supply of bread at low cost to our kin beyond the sea has been justified at every point and in every detail. I therefore ventured to review Sir William Crookes's address, and I am well assured that what Mr. Hyde now calls a "somewhat dubious outlook" is subject to no doubt whatever as to our ability to continue our full supply for domestic consumption and export for the next century. Let me now repeat again what I have often said: statistics are good servants, but very bad masters. I long since ceased to put any great reliance upon averages of crops, wages, or products covering wide areas and varying conditions, unless I could find out, _first_, the personal equation of the man who compiled them; _second_, ascertain what he knew himself about the subject of which his statistics or figures were the symbols; and, _third_, unless I could verify these great averages from one or more typical areas of farm land, or from one or more representative factories or workshops, of the conditions of which I could myself obtain personal information. General statistics and averages of farm products and earnings I regard with more suspicion than almost any others because of the immense variation in conditions. I have sometimes almost come to the conclusion that so many of the figures of the United States census are mere statistical rubbish as to throw a doubt on nearly all the schedules. Yet without accurate statistics on many points, many of them yet to be secured, the conduct of our national affairs must become as uncertain as would be the conduct of any great business corporation without a true ledger account and a trial balance. Hence the necessity for a permanent census bureau and for a careful "speculation" or "intellectual" and intelligent examination and "contemplation" or study of the facts about our land by which our future welfare must be governed.