closely compressed, and it would, under the same circumstances, be baked and dry throughout its whole depth. No air could enter for the deposit of dew, and, from its compact condition, all of the moisture that it contains would move, by capillary attraction, from particle to particle, to supply the evaporation at the surface, while the crust thus formed on the surface would prevent the free admission of air, even if the lower soil were loose and porous.

It is the same in the field. A heavy clay soil, saturated with water, dries up to a condition that will not admit of the circulation of air. Even if the thin surface soil, containing much vegetable matter, is loose enough, it is soon heated to such a depth that the little moisture it receives during the cooler parts of the day is dried out by the midday sun, while the compact subsoil is impervious to all atmospheric influence. Plants Plants grow well enough during the weeks that separate the rains of early spring from the heat of midsummer, but when the drought sets in, the roots being only in the surface soil, for roots will not enter a cold, saturated subsoil,- vigorous vegetation ceases, and we accuse Providence of having sent us a scourge for our sins. As well blame Providence for our loss if we neglected to plow, and harrow, and plant at seed time, as for loss from neglect to drain away the water that places us at the mercy of the drought. If we underdrain the land, even without the use of the subsoil plow, but better with it, the early growth will be less precarious and more uniform, and the roots of our crops will push down into the subsoil, where they will find, all through the dri est summer, enough moisture for their uses. For the first year or two, of course, we could only hope to modify our evils, but in time we should find that, if we keep the surface of our underdraining ground well stirred, a six weeks' drought, that lays the whole country-side bare, has little power to diminish our crops.

CHAPTER VI.

LIVE-STOCK.

LIVE-STOCK is more or less important to the farmer, according to the circumstances under which his business is carried on. In extensive grain-growing regions, where the policy is simply to raise the largest possible crops, rather by extent of cultivation than by excessive production per acre, and where it is intended either to trust to luck for fertility of land, or deliberately to exhaust and abandon it, live-stock forms no important part of the farm interest, it being necessary to keep only such teams as are required for plowing and harvesting. In other extensive regions, where the chief, almost the entire, business of the farmer is confined to the grazing of large flocks and herds on natural pastures, he cares for little else than live-stock; but, at the same time, his animals live almost in a state of nature, require scarcely any attention beyond the annual branding and the annual selection of droves for market, and he needs to know almost nothing concerning their management, as understood by skilful husbandmen. Live-stock becomes an important element in the economy of the farm only when our object is to raise fine animals, to raise beef for market, or wool, or dairy products, or poultry, as a means for converting the production of the land into a marketable form.

Mixed farming requires close attention and a knowledge of means, methods, and results, that can only be acquired through practice. Everything raised on the farm should be consumed on the farm, if possible. Skill in feeding stock economically is one necessary requirement in a successful farmer. Care in handling stock is another, and the judicious selection of kind, age, and number, is perhaps the most difficult of all. One farm is adapted to cattle, another to sheep, a third to neither, and a fourth to them all. To make a wise selection as to the farm, the kind of stock to keep, and the proper number, supplemented with convenient buildings, careful management, and economical

feeding, requires a quality of judgment that would conduct almost any other branch of productive industry. To be a successful stock-raiser, one must read, think, calculate, and work. It is no easy task, but requires constant application. The slothful or negligent never succeed at this business. As the country becomes older and more settled, the quality of all kinds of stock becomes improved, for men learn that the best are always in demand, while the poorest are hard to sell. There is a great future for the American stock-raiser; and the progress in that line during the past half century is only an example of what may be expected in the years to come. Nature has placed all the requirements for success in this line within easy reach of the farmer, and those who neglect these opportunities will surely repent when too late. Horses, cattle, sheep, swine, and poultry, should all reach the highest types of perfection here in the United States. The choice of breeds must be a matter of judg ment with the farmer, and no definite rules can be laid down. Suffice it to say, that, be the number more or less, they should be selected with discretion, cared for attentively, and fed economically.

If these rules fail to bring success, the cause must be looked for elsewhere. For the purpose of showing the importance of stock-raising, I quote from the last report of the Department of Agriculture, upon that subject. These tables should be studied with care.

There has been a feeling for a number of years that more accurate data should be obtained in regard to the number of the range cattle in the various States and Territories. It is probable that no accurate census of the range cattle has ever been secured, and nearly all the estimates, on account of the inherent difficulties of the case, have varied widely from one another, and probably from the true figures. In order to clear up this question somewhat, an effort was made during the year 1888 to obtain reliable data from the Western States and Territories. Accordingly, trusted agents of the Bureau, well acquainted with the range-cattle industry, were sent into the field to gather the most accurate figures possible from the cattleowners' organizations and from other sources of information. The estimates of the Statistical Division of this Department

have, as a rule, been taken as approximately correct for the number of cattle in the States; but in some cases these estimates have been revised in accordance with more recent information received from the agents of this Bureau. The population since 1880 has been estimated on a basis of a 2 per cent annual increase, in addition to the immigration.

Taking our figures from these sources, we obtain the following table :

Table showing population, total number of cattle, and number of cattle per 1000 of population (estimated since 1880) in the United States and Territories.

This table shows some interesting facts. At the first approximately accurate census of cattle, in 1850, there were 767 cattle to the 1000 of population. This number increased in 1860 to 815, showing a large stock of cattle on hand. In 1870, partly from the effects of the war, and partly from an underestimate by the census of that year, we find the number of cattle reduced to 618 per 1000 of population. In 1880 the number per 1000 increases to the extent of 120, and reaches 738. In 1881 there is an increase of 6 per 1000; from 1881 to 1882, the increase is 14 per 1000; from 1882 to 1883, it is 15 per 1000, being the

largest apparent increase in any one year; from 1883 to 1884, the increase is 14 per 1000; and from 1884 to 1885 it is 13 per 1000, reaching the highest point since 1860, or 800 cattle per 1000 population.

Since 1885 there has been, according to these estimates, a steady decrease in the relative number of cattle. From 1885 to 1886, this was 6 per 1000; from 1886 to 1887, it was 11 per 1000; from 1887 to 1888, it was 12 per 1000; and from 1888 to 1889, it was 13 per 1000. The total decrease in cattle, per 1000 population, from 1885 to 1889, amounted to 42, and the proportion was then as 758 to 1000.

A somewhat clearer presentation of the beef supply is obtained by considering the other cattle by themselves. These figures will be found in the table which is given below:

Table showing the total number of milch cows and of other cattle, and the number of each per 1000 of population.

One of the remarkable facts brought out by this table is that, since 1870, the proportion of milch cows to population has been practically constant. In 1850 there were 275 per 1000, and in 1860, 273 per 1000. In 1870 this number decreases to 232, or about 15 per cent, and increased in the ten years from 1870