_3. Clarification in separator._ Within recent years the custom has grown of clarifying milk or removing the visible dirt by passing the milk through a centrifugal separator the cream and skim milk being remixed after separation. This process naturally removes the solid impurities as dirt, hairs, epithelial scales and cells, also some of the casein, making what is known as centrifuge slime. This conglomerate mass is incomparably rich in germ life and the natural inference would be that the bacterial content of the milk would be greatly reduced by this procedure. Eckles and Barnes[31] noted a reduction of 37 to 56 per cent. of the bacteria but others have failed to observe such reductions.[32] This condition is explained by the more thorough breaking up of the bacterial masses in the process, thus apparently not reducing them in numbers. It is somewhat surprising that in spite of the elimination of much organic matter and bacteria, such clarified milk sours as rapidly as the untreated product.[33] The mechanical shock of separation ruptures the clusters of fat globules and so delays creaming and also lessens the consistency of cream derived from such milk. This practical disadvantage together with the increased expense of the operation and the failure to materially enhance the keeping quality of the product outweigh the advantage which might come from removal of solid impurities which can be largely accomplished on the farm by efficient straining. _4. Washing the udder._ If a surface is moist, bacteria adherent to it cannot be dislodged by ordinary movements. Thus the air over snow-covered mountains or oceans is relatively devoid of germ life. The method of moistening the udder is applied with success to the hairy coat of the animal thus subserving the double purpose of cleaning the animal and preventing in large measure the continual dislodgment of dust particles. After these parts have been well carded to remove loose hairs and dirt particles, the skin should be thoroughly moistened with clean water and then wiped. It has been urged that this procedure lessens the yield of milk but Eckles[34] concludes from experiments that when the animal becomes accustomed to this treatment, no noticeable change in amount of milk or butter-fat is produced. The effectiveness of this method in reducing the actual amount of dirt and filth introduced into the milk as well as the great diminution in germ life is shown by the instructive experiments of Fraser[35] who found that the actual amount of dirt dislodged from udders of apparently clean animals during the milking process was three and one-half times as much as when the cow's udders were washed. From udders visibly polluted one ounce of such filth was removed in 275 pounds of milk, while from cows whose udders had been washed, the same amount of dirt was distributed through 24,030 pounds.