563 



is neutralized so far as that is practicable by averaging the results of 

 trials with the same ration in periods near the begiuuiug and the end 

 and equidistant from the middle period. The amounts of watery foods 

 in the rations per head per day were such as to give quantities of water 

 langiug in one series from 43 to 82 pounds, and in the other from 84 to 

 150 pounds. As the watery food was increased the dry food was dimin- 

 ished so as to make the quantities of digestible nutrients practically 

 constant throughout each series. The total quantity of nutrients was 

 liberal. The digestible protein per 1,000 pounds live weight per day 

 Mas somewhat over 3 pounds in both series, while Wolff's standard calls 

 for only 2.5 pounds. The Calories were about 36,000 in the first and 

 about 33,000 in the second, Wolff's standard calling for 29,700. The con- 

 dition of the cows in respect to fatness at the beginning of the experi- 

 ment is not mentioned in the report, but where such liberal feeding is 

 practiced, they could hardly be expected to have been very lean, and 

 the fact that with the liberal feeding during the experiments they fell 

 off in live weight when the milk yield increased would favor the sup- 

 position that they were not too fat. The results may be briefly sum- 

 marized thus: 



(1) The increase of watery food was without discernible effect upon 

 the comijosition of the milk. 



(2) The quantity of milk increased regularly with the increase of beet 

 diffusion residue (series A), the yield being largest with the largest 

 amount of residue (28 pounds per head per day) with 82 pounds of 

 water in the ration. With the mixture of beet diffusion residue and 

 potato residue (series B) the milk yield likewise increased with the first 

 increase of watery food, which raised the water in the ration fi'om 84 to 

 116 pounds. With the next increase, which brought the water in the 

 ration up to 150 pounds, the milk yield fell off". But in this case the 

 ration was unintentionally changed so as to reduce the protein from 3.2 

 to 2.7 pounds per 1,000 pounds live weight and the energy from 33,740 

 to 32,140 Calories, hence the decrease of milk could not be fairly ascribed 

 to the water in the food. The authors suggest that the very large 

 amounts of water may have increased the nitrogen metabolism and thus 

 tend to decrease the yield of milk or the laying on of flesh. 



(3) There was an almost continual gain in live weight throughout all 

 the series of experiments. This gain was in general less as the water 

 in the food was gTcater. The only exception was in the case in which 

 the largest amount of water of all was fed — period III of series B. This 

 was also the case where the milk yield fell off. That there should be 

 less gain in live weight where the milk yield was larger, is not strange. 



Experiments with different quantities of concentrated foods and different 

 nutritive ratios. — Of these there were two sets, (1) with large propor- 

 tions of x)rotein, and (2) with excess of carbohydrates in the ration. 

 The general plan consisted in taking a ration of attested fitness for 

 milk production as a basis and adding oil meal and barley meal. For 



