OF ARTS AND SCIENCES. 



107 



The variation in this case consists mainly in the amount of fat. 

 This variation, as will be seen, amounts to 1.21 per cent. The 

 averages for both the above milks were found by ascertaining the total 

 weight of each product given by tlie teat, adding the four weights 

 together and diviiiing by the total weight of milk yielded. 



It is evident from these experiments that each quarter of the udder 

 yields a milk that may differ considerably from that given by any of 

 the other quarters. This, however, has only a scientific interest, as in 

 practice the four quarters are drawn simultaneously and the average 

 result is used. 



Some further experiments were tried in reference to the influence 

 of breed and feed on the quality of the milk. These are far too few in 

 number to base any opinion on ; but, so far as I have been able to carry 

 them, they show the need of extended experiments on this subject, — 

 experiments which should be carried over the space of several years, 

 with analysis at least once a week, careful records being kept of tem- 

 perature of the air, state of the weather, and general condition of the 

 cow ; and a sufficient number of cows of each breed should be em- 

 ployed, that the individual equation of the cow should be eliminated 

 so far as possible. To be comparaljle, these analyses should be either 

 made by one person, or, if made by several, they should be made l)y the 

 same method, and these persons should compare their working by 

 making several simultaneous analyses of the same sample of milk. 



Analysis No. XXIV., Ayrshire cow " Geor<rie ; " calved July 7, 

 1876; food, pasture, fodder corn, and six quarts of shorts. Evening 

 milk, Aug. 7, 1876. 



