HOKSE FEEDING. 7 



also cany a rider. Further, the traction is not performed under the 

 .same conditions as with ordinary vehicles. The artillery horse travels 

 over such varied surfaces that the rate of speed can not be calculated 

 even approximately. According to some of the writer's experiments, 

 the coefficient of speed of gun carriages and caissons is 2 per cent on 

 roads, 6 to 8 per cent on fallow lands, and 12 per cent on wet, plowed 

 fields. 



In view of these difficulties, it is evident that the only means of 

 obtaining at all satisfactory results is to estimate the load carried and 

 hauled per horse over the total distance. The writer's observations 

 on this subject were made with 16,000 horses of the (JomjKignie gen- 

 erale des oiiDilhas de Paris^ 17,000 army horses, and about 1,000 

 horses used for hauling heavy wagons. The experiments have 

 extended over a number of years. The horses of the Ooinjjagiiie 

 generate des omnibtis were of nearly uniform size and weight. Those 

 in the army differed in size and weight. In all these tests the weight 

 of the horse has been relied upon as showing whether the ration was 

 satisfactory for the work performed. 



The 20,000 or 30,000 horses experimented upon were maintained in 

 good condition, and performed the required work without any notice- 

 al)le loss of weight, and further the horses still possessed great reserve 

 energy. The numerous weighings which have been made in the 

 progress of these investigations have enabled the author to determine 

 (juite accurately what should be the weight of a good horse in perfect 

 health under various conditions of work or rest, taking into account, 

 of course, age and size. In the present paper it is necessary to omit 

 details, but the table below shows the average weights of different 

 kinds of horses in our experiments: 



Average weight of horses. 



Kilograms. 



Heavy draft horses 700-800 



Light draft horses 500-600 



Fancy horses, reserve cavah-y horses, and horses of the hne. . . 450-510 



Carriage horses and Ught cavalry horses 380-400 



Artillery and train horses 480-495 



Mules 430 



It is on the basis of such data that the rations have been varied, 

 according as the horses gained or lost weight. 



FOOD REQUIREMENTS OF HORSES. 



In connection with the experiments a large luunber of analyses 

 have been made of food, urine, and feces, and the coefficients of 

 digestibility of many feeding stuffs were determined. The object of 

 our experiments has l)een to determine the quantity of protein, fat, 

 carbohydrates, and mineral matter necessary for maintaining a horse 



