The acid measure holds 17.5 c.c. and it should be filled to the graduation. Then this acid should be added to the test bottle. The bottle should be held at an angle and slowly rotated so that the acid will rinse down any milk remaining in the neck of the bottle. Immediately mix the acid and milk by whirling the body of the bottle in a circle five or six inches in diameter. The mixture should not be allowed to go into the neck of the bottle while mixing. Continue shaking for about a minute after all the curd has disappeared. One should avoid pointing the neck of the bottle toward any person in the mixing operation. The acid unites with all the milk substances except the fat and generates much heat.

311. Centrifuging.—There are two machines in common use for centrifuging, one that runs by mechanical power and the other smaller and runs by hand. If the machine and atmosphere are very cold, the apparatus can be warmed by placing hot water in it. This is not necessary in a steam machine. In a factory where there are a number of samples to test, a power machine is usually employed. In this machine there are pockets or cups in which to set the test bottles. The machine or disk must be balanced by placing bottles in opposite pockets. These pockets are hinged so that when standing still the bottle is in an upright position and when the centrifuge is running, it is in a horizontal position. The machine should then be covered and started running. It should be run at the speed indicated. After five minutes, stop the machine and fill the bottles with boiling water up to the neck. This can be done without taking the bottles out of the machine. A pipette or slender-spouted vessel may be used to add the water. Whirl the bottles two minutes, then add more boiling water to bring the fat column into the graduated part of the neck of the bottle. Then whirl one minute. The test should be read at once or the bottles kept at a temperature of 130° to 140° F. until ready to read.

312. Reading the test.—To read the test, subtract the reading at the bottom of the fat column from that at the highest point. The curved meniscus which always forms at the top of the fat column should be included in the reading. Duplicate samples should not vary more than O.2 of 1 per cent. Standard Babcock test bottles and pipettes should always be used. In some states the agricultural experiment stations examine all glassware and mark it to make certain that it conforms to the requirements of the state law. In New York, glassware found to be correct is branded "S. B.," which means State Brand. In some states a person must have a license to test milk or cream, when it is paid for on the fat test. Such a person must pass an examination to show that he understands the test before a license, will be granted. The license may be revoked if the work is not honestly performed.

313. Testing whey for fat.—Because of the small amount of fat in whey, it is difficult to obtain a representative sample. The best way, if the entire amount cannot be placed in a vat and stirred, is to catch a little of the whey at intervals as it is being drawn from the vat. The sample to be tested is measured with the pipette the same as the milk and placed in the skimmed-milk test bottle. The same acid is used to test whey as to test milk but because there are not so many solids to destroy, not so much is used. If as much acid is used with whey as with milk, it will burn the fat and so interfere with the reading of the test. Just enough acid is added to destroy the milk substances except the fat, or enough to turn the contents of the test bottle dark brown. This usually requires filling the acid measure one-quarter of an inch under the graduation. The remainder of the test is the same as for whole milk.

314. Testing cheese for fat.—The sample of cheese to test for fat is obtained by removing the sample with a cheese-trier. This sample is called a "plug." Different plugs from the same cheese will test various percentages of fat so that it is difficult to secure a representative sample. The usual practice is to take three plugs, one near the center, another near the outside and the third between the first two. The plugs should be put into glass-stoppered bottles to prevent the evaporation of moisture. These plugs are then chopped up very fine. It is of course impossible to measure the cheese as with milk and whey, but it is weighed ([Fig. 70]). If the cheese is soft it can be stirred with a spatula until well mixed. A soft cheese usually sticks to the neck of the test bottle. After being weighed, it can be dissolved in a little sodium hydroxide and poured into the bottle. Different amounts may be used, commonly 4½ or 6 grams, but 6 grams is to be preferred. This is placed in the Babcock cream bottle since there will usually be more fat than can be read in a milk bottle. After the material has been placed in the test bottle, about two-thirds of an acid-measure of warm water is added to assist in dissolving the cheese.

Fig. 70.—Apparatus necessary to test cheese for fat.

The acid is added the same as with the milk. If all the cheese particles are not destroyed, and therefore do not disappear, a little more acid will complete the solution. Centrifuging is performed as with the milk.

315. Reading the test.—In a cream-test bottle the neck is so much wider that there is a much larger meniscus. In order to obtain an accurate result, the meniscus should be removed. This is done by carefully adding a substance called glymol, which is a mineral oil colored red. Usually about one-quarter of an inch of glymol is added to the fat column. This should not mix with the fat. The bottles should be placed in a hot water bath 135° to 140° F. for four minutes before reading. The temperature at reading should be 135° to 140° F. The reading is then taken from the bottom of the fat column to the line between this and the glymol. The bottle is graduated for 18 grams of material, but as only a part of 18 grams of cheese was used for the test, the reading should be multiplied by the part of 18 grams used. For example, suppose 6 grams of cheese were used and the test read 12 per cent fat. Since 6 is one third of 18, the actual percentage of fat is 3 times 12, or 36 per cent.