Professor H. L. Russell, of the Wisconsin Agricultural Experiment Station, cites in his little volume on _Dairy Bacteriology_ an instructive experiment which brings home very forcibly the importance of such precautions. A cow pastured in a meadow was selected for the experiment, and the milking was done out of doors, so as to eliminate as far as possible any intrusion of disturbing foreign factors into the experiment, such as the access of microbes from the air in the milking-shed. The cow was first partially milked without any precautions whatever being taken, and during the process a small glass dish containing a layer of sterile nutrient gelatine was exposed for one minute beneath the animal's body, in close proximity to the milk-pail. The milking was then interrupted, and before being resumed the udder, flank, and legs of the animal were thoroughly cleansed with water; a second gelatine surface was then exposed in the same place and for the same length of time. The results of these two experiments are very instructive. When the cow was milked without any special precautions being taken, 3,250 bacteria were deposited per minute on an area equal to the surface of a ten-inch milk-pail; after, however, the animal had been cleansed, only 115 bacteria were deposited per minute on the same area. Thus a large number of organisms can, by very simple precautions and very little extra trouble, be effectually prevented from obtaining access to milk. Even in the event of the milk being subsequently Pasteurised, clean milking is of very great importance; but still more imperative is it when it is destined for consumption in its raw, uncooked condition. If we consider how cows become covered with dirt and slime, that obstinately adhere to them when they wade through stagnant ponds and mud, and realise the chance thus afforded for malevolent microbes to exchange their unsavoury surroundings for so satisfactory and nourishing a material as milk, then indeed precautions of cleanliness, however troublesome, will not appear superfluous. That a very real relationship does exist between the bacterial and dirt contents of milk has been clearly shown by actual investigation. A German scientist has made a special study of the subject, and has determined in a large number of milk samples the amount of foreign impurities present per litre, and the accompanying bacterial population per cubic centimetre. The following results may be taken as typical of those obtained: in milk containing 36·8 milligrammes of dirt per quart as many as 12,897,600 bacteria were present per cubic centimetre; in cleaner samples, with 20·7 milligrammes of dirt per quart, the number of bacteria fell to 7,079,820; whilst in a still more satisfactory sample, containing 5·2 milligrammes of dirt per quart, there were 3,338,775 bacteria per cubic centimetre.