Draft-horses, in this connection, include all horses used for slow or comparatively slow and exhausting work, especially farm-horses used for tilling the land. Horses in a wild state graze largely at night, while cattle feed largely in day time. The horse's stomach is relatively small, implying that he should be fed more often than cattle, and less at a time. Horses do not eat so fast as cattle, and do not remasti-cate their food, as do cattle and sheep. Horses may be fed three or four times daily, while cattle and sheep do well when fed but twice. The morning ration of the horse should contain about one-fourth of his total daily ration, and it should be given him some little time before being put to hard work. Another fourth of his food may be fed at noon. On the farm, one hour's nooning instead of two, as of yore, is best. The hour saved will shorten the work-day one hour at night. This will give time for the horse to cool off before the dampness and falling temperature of the evening occur, which tend to produce that disagreeable and dangerous condition which a person feels in the evening whose shirt is saturated with the prespiration of the day. Horses, and especially oxen, if perspiring freely, and worked late, are frequently found in the morning with the hair and skin damp, and in a condition of lassitude which unfits them for entering upon the day's labor with vigor. Ten hours of faithful work per day is quite enough for either man or beast; and such work would better be accomplished by starting early than by continuing late. If the horses are brought to the stable early, they have time to eat hay and to rest, after which one-fourth of the grain ration is fed. Just before the attendant retires, the last fourth of the day's ration may be given. Or, if this is too much trouble, the horses may be left to eat hay for an hour while the teamster is eating supper; after which the legs are cleaned, and then one half of the day's ration of grain may be fed. The hay ration should be fed about the time and in about the same proportion as the grain is fed. How much grain and how much hay should suffice for each horse it is impossible to say; since the size of the horse, his ability to digest and assimilate food, the kind of food consumed, and the work performed vary widely. When horses are put to unusually severe tasks for a month or two, when work is pressing, as they usually are on a farm, it is better to increase their grain than their hay ration. Quite a large part of the energy in food, especially if it is coarse food, is used for preparing the ration for assimilation. Coarse and unconcentrated foods are frequently more expensive per unit of net available energy than concentrated ones. It is not the total energy of the food so much as the available energy over and above that required to masticate and digest it that governs value. Dr. H. P. Armsby's experiments show the following results: "Of the total or 'gross' energy of hay, about 44 per cent was capable of conversion into the kinetic form in the animal, while the remaining 56 per cent was found as potential energy in the excreta. Of the 44 per cent which I have called metabolizable energy, about 63 per cent, equivalent to 27.72 per cent of the gross energy of the hay, was found to be available for the maintenance of the animal, while the remaining 37 per cent of the metabolizable energy, under our conditions of experiment, simply went to increase the heat-production of the animal. This 37 per cent of the metabolizable energy seems to represent the expenditure of energy which is involved in making the remaining 63 per cent available for the actual uses of the organism. The above results represent the average of four experiments only, on a single animal, and of course should be generalized from very cautiously."

The more concentrated the food is, within proper limits, the less per cent of energy is needed to make it available. This fact explains in part why animals cannot be sustained and produce sufficient energy for growth and work on unconcentrated food, difficult to masticate and prepare for assimilation. It is because too great a per cent of the energy of the food is used in its preparation by the stomach, hence the net energy is small.

When horses are doing light work or are idle, not only will less food suffice, but the proportion of rough, cheap food to the concentrated may be increased.

Horses should be fed slightly less on idle days than when employed. Much of the trouble, particularly bowel-complaints, on Mondays is due to over-eating on Sundays. When a record was kept with farm-work horses, it was found that there were more than twice as many indisposed horses on Monday as on any other day. When the Sunday's ration was slightly reduced, the health was equally good on all days. When practicable, work-horses should have some exercise every day. This can most easily be secured by turning them out for a few hours in a paddock, on idle days. If the horses are at severe work, they do better on dry food than on green grasses and clover. If horses are allowed all the green food they desire at night, bowel trouble may occur the next day if the weather is warm and the work hard.

Not infrequently the hay runs short in late spring and grass is fed in the stables as a substitute. This is all very well if the work is light. New hay, while still heating in the mow, is always dangerous. The health of a horse at hard work is governed largely by the food he consumes. A little grass mixed with old, dry hay may be fed safely, but new hay passing through a sweat should never be used. In hot weather, horses at severe work in the fierce sun do best if they are kept in comfortable stables at night and are fed on roughage and concentrates of the previous year's growth. Some horsemen understand this, and will pay more for old than for new oats or hay. So the hay-buyer does not bale hay for the city until it has gone through the "sweat" in the mow or stack.