Although the increase in intensive agriculture and the diversification in farming tend to increase the need of farm laborers, the introduction of farm machinery has much more than offset this demand. The tendency of farm laborers to become farm tenants; or, to state it in other words, the tendency of landowners to rent their land rather than to continue to operate it themselves, is not without its influence upon the labor problem. The invention and introduction of farm machinery has accentuated the difficulty of keeping the farm laborer continuously employed. The decrease in the demand for farm labor and the increasing lack of uniformity in the amount required have caused a gradual depletion of the smaller villages and hamlets which were a source of labor supply during harvest and other busy seasons. The problem of keeping labor continuously employed has always been a difficult one on the farm, because of the change of seasons and because of the variations in the weather from day to day. There is a wide difference between those industries which are carried on within doors and farming, which is subject to the caprices of the weather. Natural causes produce tremendous variations in the return for labor. For example, in 1901 there were produced in the aggregate 3,006 million bushels of wheat, maize and oats, while in 1902 there were harvested 4,180 million bushels. Here is an increase of over a thousand million bushels. The same farmers tilled the same soil in the same way as far as natural causes would allow, and yet there was a difference in result amounting to 39 per cent. A variation of one hundred million bushels of wheat from year to year, due to climatic conditions solely, is not at all unusual. The manufacturer also has far greater control of his labor. When it rains, he has a roof over his workmen, and hence the work is not interrupted. When it grows dark, he turns on the light and the work continues. If it gets cold, he lights the fire and still the work continues comfortably. It is not so in agriculture. There is a great variation in the working efficiency of men employed in farming. In a certain locality there were twenty-one days of rain in the thirty-one days of May. The next year between June 5 and September 5 in the same locality there was not half an inch of rainfall at any one time.