It follows as a natural sequence that all the cans and vessels used for dairy purposes should be absolutely beyond suspicion of contamination. Professor Russell has shown by actual experiment that, even where the vessels are in good condition and fairly well cleaned, the milk has a very different bacterial population when collected in them and in vessels _sterilised by steam_. Two covered cans were taken, one of which had been cleaned in the ordinary way, and the other sterilised by steam for half an hour. Previous to milking the animal was carefully cleaned, and special precautions were taken to avoid raising dust, whilst the first milk, always rife with bacteria, was rejected. Directly after milking bacterial gelatine-plates were respectively prepared from the milk in these two pails, with the following results: In one cubic centimetre of milk taken from the sterilised pail there were 165 bacteria; in that taken from the ordinary pail as many as 4,265 were found. Another experiment illustrates perhaps even more strikingly the effect of cleanly operations in milking upon the initial bacterial contents of milk. The preliminary precautionary measures were carried out by an ordinary workman, and are in no sense so refined as to be beyond the reach of ordinary daily practice. "The milk was received in steamed pails, the udder of the animal, before milking, was thoroughly carded, and then moistened with water, so as to prevent dislodgment of dirt. Care was taken that the barn air was free from dust, and in milking the first few streams of milk were rejected. The milk from a cow treated in this way contained 330 bacteria per cubic centimetre, while that of the mixed herd, taken under the usual conditions, contained 15,500 in the same volume. The experiment was repeated under winter conditions, at which time the mixed milk showed 7,600 bacteria per cubic centimetre, while the carefully secured milk only had 210 in the same volume. In each of these instances the milk secured with greater care remained sweet over twenty-four hours longer than the ordinary milk."