lected sufficient material and is now ready to publish his second volume of four or five hundred more Jersey cows having recorded yields, by authorized tests, of fourteen pounds of butter and upwards in seveu days. No Jersey cow is admitted in this book unless she will make at least two pounds of butter a day, and we have in these two volumes nearly one thousand cows with such records. When any other breed of cattle can get up such a record as this, then, and not until then, will it be necessary to say anything further in defense of the Jersey cow.

MILK SEPARATION BY CENTRIFUGALS, AND MILK TESTS.

By JOHN I. CARTER, Chatham, Chester county, Pa.

I presume intelligent farmers everywhere understand the principles of centrifugal force and its application to the separation of cream from milk by machinery, and probably most of them are familiar with the general construction of the machinery used for this purpose, now called centrifugal cream separators. I shall not, therefore, in this short paper go into any historical account of the various inventions used for this purpose, or of their mechanical construction, or of their introduction to the general public. These phases of the subject have been frequently treated of in papers read before this Board, and it is only necessary to give the present status of these various inventions, of their capabilities and improvements and the practical results of their use in the dairy during the last few years.

Perhaps part of this subject might be left to the manufacturers' circulars, but we know these people are a little given to making rosecolored statements of the construction and work of their own machines and of saying rather disparaging things about the machines of other makers. And you know also there is often quite a difference between the claims put forth in these circulars and the practical work the machines will do when put down to every day use and come under the care of less skilled operators. But we now have had four or five years' experience in the use of these machines and are able to judge pretty accurately of their general usefulness and of the actual capacity of the individual machines and their special advantages.

There are from one hundred to one hundred and fifty creameries in Chester and adjoining counties, almost all of which employ one or more separators of some kind or other, deeming their use economical and essential to the prosperity of the business. These machines will take out of the milk from ten to fifteen per cent. more butter than any other practical method of cream separation; and the convenience and rapidity of the operation is quite an additional advantage. The general quality of the butter has been improved; there is more uniformity in the whole product, and it is easier to fix an equalized standard price for this class of goods.

It is no doubt true that a well-appointed dairy carried on in the oldfashioned way and supplied with milk from well-fed cows might make a better butter than the usual product of our creameries, but it would not be owing to any fault in the manner of separating the cream, but to the fact that the creameries have to use miscellaneous milk, bought once per day, and frequently subjected to quite careless handling by

negligent farmers. Better butter can be made out of such milk as this by centrifugal separators than by any other process, because the cream is got out of it in the quickest possible time, thus avoiding pro longed contamination from the defective milk. The old claim that these machines made a better article of skim milk for cheese and other purposes was perhaps not well taken. It is true that fresh, sweet milk run through a separator will still be sweet, and for a short time pa atable, but the thorough aeration or severe treatment it receives in pass ing through the machines predisposes it to rapid deterioration, and that fact, together with the poverty of the milk from thorough skimming, made the making of a good skim cheese an impossible matter. For drinking purposes the skim milk also soon becomes insipid and unpalatable. Of course, the cream produced by these centrifugals requires different treatment from that raised by the old methods. It must be cooled down quickly and churned in one or two days at most. It also requires a different treatment when used for making ice cream. It must have all the froth taken off of it and be liberally diluted with new milk. By the way, this froth is often considered a great bugbear, but in reality is of little account if you know how to manage it.

All centrifugals make more or less froth, the amount depending more on the temperature and condition of the milk than the style of the machine. In point of fact, there is but little difference in the quality of the work done by the different machines as at present improved, and the prices and capacity of them agree pretty closely. It would be unjust as well as ungracious to recommend one machine above another, as the testimony of the people using them show that some are adapted to one place or circumstance and some to another, and nearly all are equally good.

The following list comprises the kinds of centrifugals in use amongst us, with their capacity for work and the power required to drive them as far as our experience shows:

The twenty-five-inch Danish-Weston costs $500, takes about a four horse-power to drive it, makes 2,000 to 2,400 revolutions per minute, and separates about 1,000 pounds of milk in winter and 1,200 pounds in summer.

The fifteen-inch Danish- Weston costs $325, takes two and a half to three horse-power to drive it, makes 3,500 revolutions per minute, and separates 600 pounds in winter and about 800 pounds in summer. The Home Danish-Weston costs $250, makes 4,000 revolutions per minute, skims 600 pounds per hour, and takes two horse-power to drive it. This machine is so lately put upon the market that the above figures are taken from the manufacturer's circulars and not from our experience.

The DeLaval costs about $275, takes about a two horse-power, makes 8,000 revolutions per minute, and separates 600 pounds in winter and 700 pounds in summer.

An improvement on these machines increases its cost and its capacity about equally.

The Backstrom separator costs $185, takes one and a half horsepower to drive it, and makes 7,000 revolutions per minute, and separates 400 pounds in winter and 600 pounds in summer. This is also a new machine, and part of these figures are those given by agents.

Any of these machines will require skill and some experience to run them successfully, and the cream and butter they produce will take 15 BD. AGR.

more inteligent management than the cream and butter secured in the old fashioned way.

The large Danish-Weston is perhaps the most substanial and reliable machine, but it requires large and heavy foundations, occupies considerable space, and costs a good deal of money. I have running, to-day, about the first large separator ever put up in our country, and after five years of constant work is apparently as good as ever. Our moderate sized creameries, however, now prefer to put in two or more small machines of some kind, partly to provide against a total stoppage of work should anything happen to one machine, and partly because they are more manageable as regards space, power and care. In providing a creamery with separators, it is well to have enough to do the work without too much delay. The separation should be finished by ten o'clock A. M. during the hot weather. When forty or fifty lots of mixed milked is dumped into one receiving vat, to stand till separated, it is soon seriously damaged, and will not make good, long-keeping butter.

Perhaps this is all I need say about separators now; any other information needed will be gladly given upon inquiry.

I had proposed in this connection to say something on the testing of milk for butter value, but owing to a disastrous accident to some of our machinery, was not able to complete some experiments and tests that I hoped would be interesting to report. I am sorry to have to say, therefore, that I know of no reliable test for milk quality that is practical enough for every day use in our dairies or at our cream eries. The simple tests that have been in common use, such as the cream guage glass, the lactometers, pioscopes and other opacity tests, were too uncertain or too indefinite to accept as a basis to assess value upon. When it comes to paying a captious milkman for his milk according to its supposed value for buiter purposes, you must be prepared to make a very clear showing that your figures are all right, if not you will get into infinite trouble, as well as being liable to do injury to others and yourself. The Curtis oil test-churn, now in use in many of our western creameries, looked like a plausible help, but late carefully conducted experiments with it indicate that when used for testing milk the quantity of fat is too small to be accurately measured with the eye or rule.

Probably a chemical analysis by a careful analyst would be the safest test we could make, and as the process is simple, it will not be impracticable to adopt it. It is the simple evaporation of a sample of milk in water-bath ovens, treating the residuum solids to ether to dissolve the fat, and then evaporate the ether. By very delicate scales these remainders can be weighed, and the amount of butter fat accurately determined.

N. F. UNDERWOOD of Wayne. I understood Mr. Carter to state that he found the ordinary or common oil test as applied to milk to be unsatisfactory and unreliable. Does he consider it to be reliable when applied to cream collected by the cream-gathering plan, churn the cream and reduce the butter to oil, is it then reliable?

JOHN I CARTER of Chester. Our churn (for testing samples of milk) is simply a collection of small phials or bottles, perhaps six inches long and five-eighths of an inch in diameter. These are filled half full and then set away and the cream allowed to raise, and then churned. They will only hold a very small quantity of milk, not more than three or four ounces. Now the butter fat in that much

milk must necessarily be very small. The amount of cream is decided by a number of minute lines on the side of the tubes; at best the cream in that amount of milk makes a very thin and indistinct line. When the milk is poor in butter fat of course the line is still more difficult to see and accurately measure by the eye. You cannot tell by the gauge whether it will take nine or ten quarts of milk for one pound of butter. It is a little too fine work to be practical.

Another difficulty with this mode of testing is that butter fat is liable to absorb from five to fifteen per cent. of water, and the amount thus absorbed will depend somewhat upon the character of the cream; consequently we do not find them sufficiently accurate for our pur

pose.

Question. How does the quality of butter from cream separated by the centrifuge compare with that made in the old way?

JOHN I. CARTER. The quality of the butter is not affected; if that made by both processes is equally free from all bad surroundings and odors there should be no difference. Under the old process it was very often the case that the resulting butter was effectually spoiled by the absorption of bad odors while the cream was raising. This was more the fault of the dairyman than of the process. With the centrifuge we, by taking out the cream soon after milking, avoid this danger. As a matter of fact, and taking the average butter as made by the old process of raising the cream in shallow open pans, the centrifuge will make the best and sweetest butter; aside from this there will be no difference. If you take the milk right from the cow and run it through the separator you can readily so take care of the cream that it shall absorb no bad odors, but it is much more difficult to prevent the same trouble with the amount of milk necessary to produce that amount of cream by the old process. In some cases we put the milk through the separator twice each day (morning and evening's milk separately), but once a day is the usual rule, and is ordinarily sufficient if proper care has been taken of the milk after it left the cow and between the milking shed and the creamery. With the separator, as with the shallow system, or the Colley system, all depends upon having good, sweet and pure milk; without this neither will make first-class butter, but from the fact that the milk is exposed to the danger for a much less time, the separator is safest and best.

A gentleman asks an opinion on the "cream gathering plan," by which each patron raises his own cream and it is then sold to the creamery. The plan is open to several objections. The creamery does not control the treatment of the milk so well as it can when the whole milk is taken, and as a conseqence the cream is not as uniform in quality. One batch of bad cream will injure the whole churning from any number of patrons; the farmer has entire control of the milk, and the creamery owner does not know how it is handled. He only gets the cream, and that is by far the most likely to be injured by any bad management or odor. In our neighborhood all of the creameries require that they shall have the whole milk hauled to them within a certain number of hours after it is milked. If the farmer wants the milk we have an arrangement by which he can have the skim milk to haul back. Another trouble is that the cream is more sensitive to injury than the milk, and in hauling it around the country (as in the cream gathering plan) it is much more liable to injury than the whole milk is. From these and other causes the butter from creameries to which the whole milk is hauled should be the best.

Question. What value would you give to whey and skim milk for feeding hogs; will it make them fat?

JOHN I. CARTER. I do not know about making hogs fat on centrifugal milk without some meal with it; we take out all of the cream and there is very little left of fattening value in the milk; it is better for growing stock.

Secretary EDGE. Is it not better to keep more hogs and use some meal, than to feed nothing but the milk? Will you not get better results from the same amount of milk in this way?

JOHN I. CARTER. Certainly, it is best to use some meal; if you feed sweet milk to hogs you must get them fat in about sixty days, or if you feed them much longer they will "break down" and will not do so well. Some farmers take the skim milk home the morning that it is separated and feed it to calves, but they obtain the best results by mixing meal or "shipstuff" with it.

J. P. BARNES of Allentown. With a dairy of from fifteen to twenty cows, would we be warranted in investing in a centrifugal machine for taking out the cream?

JOHN I. CARTER. They are now making hand machines for small dairies, but I do not know with what success; I do not know whether they are practical or not; it does not require much power to separate the cream, but it does require considerable power to maintain the high rate of speed which is absolutely necessary; with a hand machine it would be very difficult to keep the speed regular, and regularity of speed is one of the most essential points in a successful separator; if run too slack all of the cream will not be taken out, and if too fast, some of the milk will go over with the cream.

Question. Can they get sufficient velocity for good work with a hand machine?

JOHN I. CARTER. I should think that they might, but the main difficulty is to obtain the necessary regularity of speed without which the plan will fail; at any rate every dairyman should have a small engine; it is useful to cut fodder and hay and to heat water, &c.; it could readily run a small separator, which might be bought for one hundred and fifty dollars or more, according to size.

Question. Do you know of any reliable mode of testing the milk furnished to creameries by patrons? This matter has given us more trouble than any other; do you know of any reliable plan to overcome the difficulty?

JOHN I. CARTER, I do not; I thought that I was on the right plan when I put each patron's milk in jars, very similar to quart fruit jars and churned twenty samples at the same time and in the same churn, but I found this difficulty-all, the sample jars contained exactly the same amount of each patron's milk; they were all churned by the same operation and at the same temperature; but one sample would give us nice firm and hard butter of an excellent quality, while another would give a product much more soft and oily and not at all satisfactory; in some cases the sample would give up its buttermilk and water and work out clean; in others it was impossible to separate the water and buttermilk fairly; and this spoiled the test, for we could not obtain the weight of the milk and water which was necessarily left in the sample.

JOHN HOFFA of Northumberland. My experience in keeping a small dairy of from fifteen to twenty cows is that I can get the best results from my skim milk by feeding it all to my chickens; I put it in a