as it reaches the top of the drum it escapes through the cream outlet CO into the collar placed upon the frame and flows through the tube marked cream. The skim milk escapes through the bottom of the bowl and flows out through the tube marked skim milk. The thickness of the cream is regulated by the regulating screw RS. By screwing this upwards thin cream is secured, while if thick cream is desired it is screwed down. The bowl rests on the spindle S. At the lower end of the spindle and in the lower bushing is a series of cog wheels which are so arranged that forty-six turns of the crank per minute will give the proper speed to the bowl. The spindle rests on a small steel ball shown at BF.

The machine used in these trials was the thirty-gallon size, and during a portion of the time was operated by Mr. Geo. R. Meloney, representing the agents in this country.

The milk used was from the Station's herd of full-blood and grade Guernseys. The following presents the details of the trials.

In all work of this sort, slight errors, owing to mechanical loss of milk or cream adhering to vessels, slight unavoidable errors of sampling, analysis, etc., will occur. It will be seen that in these trials, these errors are very small, thus furnishing the best guarantee of the substantial accuracy of the tests.

TESTS OF 30-GALLON VICTORIA SEPARATOR-CONTINUED.

The above results are summarized in the following table in such a manner as to show the proportion of the total butter fat of the milk that was recovered in the skim-milk and cream respectively. In calculating this, the amount of mechanical loss or gain is deducted from the amount of fat in the milk, thus basing the results on the actual quantity of butter-fat found in the products.

The average

In the above eleven trials of the machine 96.78 per cent. of the butter-fat in the milk was recovered in the cream. capacity of the machine was 336.96 pounds of milk per hour.

THE EXTRACTOR-SEPARATOR NO. 2.

This machine makes butter directly from the milk, thereby saving the process known as ripening the cream and leaving but one residue-the extracted milk,-instead of two,-the skimmilk and buttermilk-obtained in the common process of churning. The extractor takes the milk at churning temperature (58 to 62 degrees) when fresh from the cows, or sour milk may be used to equal advantage if it has not become "clabbered," so that the tubes of the machine will become clogged. The machine was invented by Mr. C. A. Johanson of Stockholm, Sweden, and is sold by the Vermont Farm Machine Company, Bellows Falls, Vt., the sole licensee of the American Butter Extractor Company, of Newark, N. J., who manufacture it under patents covering not only the machine itself, but the process involved. The machine can be used either as an extractor to manufacture butter directly, or as a separator, to separate cream from the milk; whence the name extractor-separator.

The general appearance of the machine is shown by the full page cut (Fig. 3).

The mechanical arrangement and construction of the machine is excellent.

HOW THE TESTS WERE MADE.

The machine used in these tests is known as the No. 2 machine, or next to the largest size. The milk used was from the Station herd of Guernsey and grade Guernsey cows, there being available for each trial an amount ranging from 265 to 313 pounds, making the length of the run from 25 to 37 minutes. As far as could be observed the mechanical working of the machine was normal. The speed was fairly well controlled and averaged about 7,200 revolutions per minute, or that prescribed by the company in their instructions for operating.

The milk for each run was placed in a tempering vat and the temperature of 56 to 60 degrees Fahrenheit secured. The butter as it came from the machine was received in a can containing a little water. When the can was full the water and milk were drawn off from the bottom through a fine seive, and the drainings passed a second time through the machine. This was done with all except a small amount of drainings at the close of the run which could not readily be returned and was therefore weighed and analyzed. At the close of each run the butter was washed with water until the latter came off clear, the washings being collected, weighed and analyzed. The milk and skim-milk were carefully sampled. Gravimetric analyses were made of these and the other products.

DETAILS OF THE RUNS.

In all, eight runs were made, the details of which follow. The mechanical losses, while larger than in the separator trials, are not larger than are to be expected in experiments in which sev eral products have to be collected and analyzed.

It will be noticed that in most of the trials the weight of the skim-milk was greater than that of the whole milk. This is due to the addition of water to keep the butter in suspension in the chamber upon the lower side of the bowl.

August 10, 1892.—A run of 27 minutes was made, the temperature being from 56 to 57 degrees and the speed of the bowl from 6,800 to 7,000 revolutions per minute, with the following results: