DESCRIPTION OF APPARATUS AND CHEMICALS REQUIRED.

1. Test Bottles. (Fig. 1.) These are of the same form as the bottles used in Short's test, but are made a little smaller and of heavier glass. They should contain up to the neck not less than 40 c. c., and not more than 45 c. c. Each division of the graduated scale upon the neck represents .04 c. c., and in order to facilitate the reading the neck is made of such a diameter that the marks of the scale are about 1 millimeters apart. Five of these divisions are equivalent to one per cent. of fat when 18 gms. of milk are used in the test, it being assumed that the specific gravity of the butter fat, at the temperature at which the reading is made (about 120° F.), is 0.9.

2. Pipette for Measuring Milk. This may be of any form, but that shown in fig. 2, with a rather wide opening at the lower end to allow the milk to run out rapidly is to be preferred. It should contain when filled to the mark, 17.6 c. c. A pipette of this size will deliver a little less than 17.5 c. c. of milk. The quantity of milk required for the test sample of 18 gms. is 17.44 c. c., if the milk has the average specific gravity of 1.032.

3. A Measure for the Acid. A graduate or cylinder of glass (Fig. 3), with a lip to pour from and a single mark at 17.5 c. c., is the best form for general use. In laboratories a large burette, holding 100 c. c. or 200 c. c. with marks at each 17.5 c. c. and having a glass stop-cock, may be used to advantage, but on account of the liability of breakage is not to be recommended in factories or private dairies.

4. A Centrifugal Machine. Figs. 4 and 5 show the construction of the machine used by us. The gear in this apparatus is so proportioned that the wheel which carries the test bottles makes about ten revolutions to one of the crank; with this it is easy to impart from 700 to 800 revolutions per minute to the horizontal wheel. Any arrangement that will do this, either by belt or gear, will answer the purpose. Within the horizontal wheel (a, Fig. 5) are placed sectors (b) made from heavy sheet copper to which are soldered cups. or tubes (c), inclined so as to make an angle of about thirty

degrees with the horizontal, for the support of the test bottles. The horizontal wheel is surrounded by a copper jacket (d, Fig. 5) with a cover. This serves the double purpose of supplying heat for the test by pouring hot water into it, or by heating water directly with a lamp placed beneath, and of arresting the hot acid which would fly off if a bottle should break.

5. A kerosene or gas lamp for heating two or three quarts of water to boiling.

6. Commercial Sulphuric Acid, having a specific gravity of 1.82 or about 90 per cent. pure. If only the pure acid is available, it should be diluted with water to the strength indicated.

MAKING THE TEST.

Sampling the Milk. Every precaution should be taken to have the sample represent as nearly as possible the whole lot of milk from which it is taken. Milk fresh from the cow, while still warm, and before the cream has separated in a layer, may be thoroughly mixed by pouring three or four times from one vessel to another. Samples taken at once from milk mixed in this way are the most satisfactory of any. Milk that has stood until a layer of cream has formed should be poured more times, until the cream is thoroughly broken up and the whole appears homogeneous. No clots of cream should appear upon the surface when the milk is left quiet for a moment. With proper care any milk that has not coagulated or that has not been exposed to the air until the surface of the cream has become dried, may be mixed so that a representative sample may be taken. Milk should not be poured more times than is necessary, as continual mixing in this way is liable to churn the cream, forming little granules of butter that quickly rise to the surface. When this occurs it is impossible to obtain a fair sample. Milk is sometimes churned by being transported long distances in vessels that are not full. When this occurs it is useless to make an examination.

It is impracticable to sample a large amount of sour milk, but a small sample of a pint to a quart may be thoroughly mixed by adding five per cent. of strong ammonia water which will dissolve the curd and permit a uniform mixture being made. When ammonia is added the final results should be increased by five per cent. Samples from sour milk are, however, never as satisfactory as those taken when the milk is in a proper condition.

Measuring the Milk. When the milk has been sufficiently mixed, the milk pipette is filled by placing its lower end in the milk and sucking at the upper end until the milk rises above the mark on the stem; then remove the pipette from the mouth and quickly close the tube at the upper end by firmly pressing the end of the index finger upon it to prevent access of air; so long as this is done the milk cannot