The size of the farm is another of those questions on which there is endless debate and to which no general answer can be given. There are, however, certain rather definite principles which may help in settling an individual problem. The size of the farm is related to the income per acre. If one's ideal or purpose is a gross income of $1,000 or $3,000 or $5,000 a year, he must consider how large a farm will be necessary to bring this return. Assume, for the sake of discussion, it is desired to obtain a gross income of $4,000. In the eastern United States 200 acres of tillable land devoted to general farming may bring this amount. If the land is especially adapted to potatoes, and this crop takes a prominent place in the rotation, 100 acres might be sufficient to return the income named. Likewise a 100-acre retail milk dairy farm may produce a similar result. Forty acres devoted to truck farming or market gardening may be sufficient. There is another way that the size of the farm needed may be estimated. There is a general relation between the gross income and the amount invested. In 1900 the gross income of the farms of the United States was 18 per cent of the total investment, which includes land, buildings, tools, and live stock. The average gross income varied for the different types of farming common to the northern United States from 16 to 19 per cent. This represents, of course, a great deal of very poor farming. The income of prosperous farmers must be somewhat better than this. If we assume that by careful methods the gross income is 25% of the total investment, then an investment of $16,000 will be required to bring a gross income of $4,000. While it is true that the gross income has no necessary relation to net income or profit, yet it is well to remember that a gross income is a necessary antecedent of a net income. The net profit from the production of a bushel of wheat, a dozen of eggs, or a pound of butter is of comparatively small consequence unless a sufficient quantity is produced. A recent investigation by the Cornell station appears to show that with the type of farming now existing in Tompkins and Livingston counties, New York, where the investigation chanced to be made, the larger farms yielded the most profitable returns and that while present conditions exist, the size of farms is likely to increase rather than decrease. The fundamental reason seems to be the substitution of horse-drawn machinery for hand labor. The following table shows the labor income on 586 farms operated by the owners, classified according to size: