I have already stated that all good milk of the cow contained butter enclosed in little round globules held in suspension, or floating in the other substances. As soon as the milk comes to rest after leaving the udder, these round particles, being lighter than the mass of cheesy and watery materials by which they are surrounded, begin to rise and work their way to the surface. The largest globules, being comparatively the lightest, rise first, and form the first layer of cream, which is the best, since it is less filled with caseine. The next smaller, rising a little slower, are more entangled with other substances, and bring more of them to the surface; and the smallest rise the slowest and the last, and come up loaded with foreign substances, and produce an inferior quality of cream and butter. The most delicate cream, as well as the sweetest and most fragrant butter, is that obtained by a first skimming, only a few hours after the milk is set. Of three skimmings, at six, twelve, and eighteen hours after the milk is strained into the pan, that first obtained will make more and richer butter than the second, and that next obtained richer than the third, and so on. The last quart of milk drawn at a milking, for reasons already stated, will make a more delicious and savory butter than the first; and if the last quart or two of a milking is set by itself, and the first cream that rises taken from it after standing only five or six hours, it will produce the richest and highest-flavored butter the cow is capable of giving, under like circumstances as to season and feed. The separation of the butter particles from the others is slower and more difficult in proportion to the thickness and richness of the milk. Hence in winter, on dry feeding, the milk being richer and more buttery, the cream or particles of butter are slower and longer in rising. But, as heat liquefies milk, the difficulty is overcome in part by elevating the temperature. The same effect is produced by mixing a little water into the milk when it is set. It aids the separation, and consequently more cream will rise in the same space of time, from the same amount of rich milk, with a little water in it, than without. Water slightly warm, if in cold weather, will produce the most perceptible effect. The quantity of butter will be greater from milk treated in this way; the quality, slightly deteriorated. It must be apparent, from what has been said, that butter may be produced by agitating the whole body of the milk, and thus breaking up the filmy coatings of the globules, as well as by letting it stand for the cream to rise. This course is preferred by many practical dairymen, and is the general practice in some of the countries most celebrated for superior butter.