The co-efficients are: fat 100, albumen 47.4, starch 43.l; and he gives the following example in answer to the question: "What is the comparative heat-forming value of the following foods?"

This second food is obviously about one and a half times as heat-producing and work-producing as the first food; the two foods together represent the typical average food of a horse, the first being equivalent to the 12 lb. of hay, the second to the 10 lb. of oats. To find out the number of foot-tons produced by this diet, it is necessary to go back to Frankland's table, and, having found the number of heat units for the article of diet in question in albumen, fat, starch, or sugar, deduct four-fifths, as not more than one-fifth of the whole potential energy of the food will be converted into work. Thus the typical diet of 12 lb. of hay and 10 lb. of oats will have their potential and actual energy expressed as follows: -

Colonel Fred. Smith states that from some careful experiments of his own he found that horses might be. kept in perfect health without loss of body weight on 12 lb. of hay per diem, of which not more than about one-half was digested and assimilated by the system. Assuming that nearly 5 lb. of assimilated food possesses the same digestive co-efficients as those of the 12 lb. of hay given in the table marked "First Food", the horses received:

Presumably the animal referred to as having been kept in health without loss of weight on 12 lb. per day did very little, if any, work, but the writer does not give any information on this point. He, however, remarks that it is difficult to fix the number of foot-tons of daily work which can be performed by a horse without loss of condition and weight, but he adds there are many circumstances which lead him to believe that 3000 foot-tons per day is the quantity.