10 



again closing with the palm of the hand. Eeplace the bottle 

 on the table, and take from it for the test 11 c.c. of milk free 

 from air bubbles. Now shake the contents of bottle No. 4, after 

 which return to No. 2, and take a sample, then shake No. 5, 

 then sample No. 3, and so on. Always examine the milk in the 

 pipette. If any bubbles are observed, empty, mix the sample 

 as described above, and refill the pipette.] 



After the milk 1 c.c. of amyl alcohol is added to each bottle. 

 If the operations have been properly performed there should be 

 now three layers of liquid each sharply defined. To secure this 

 the milk and the amyl alcohol should be run down the side of the 

 test bottle and not dropped directly into it. The bottles are now 

 securely closed by the rubber stoppers, which should be dry, soft, 

 and free from cracks and granular matter. Each bottle in turn 

 is then well shaken, the thumb being held on the stopper to pre- 

 vent slipping. As considerable heat is developed a piece of 

 thick flannel, about 4 inches square, with a hole in one corner 

 for the thumb, may be used to protect the hand. When the 

 contents of a test bottle are well mixed turn it upside down to 

 permit the acid in the neck to run into the body of the bottle. 

 Shake till the curd is completely dissolved. 



When less than half a dozen samples are to be tested the bottles 

 may be straightway placed in the centrifuge. If, however, as 

 is usually the case, a larger number of bottles are in use, each 

 should be placed, stopper end up, as soon as it is mixed, in the 

 water bath at a temperature of from 65 C. to 70 C. The water 

 in the bath should be up to the shoulders of the bottles. When 

 a set of bottles has thus been prepared place them in the centri- 

 fuge and whirl for three minutes. In cold weather the 

 body of the centrifuge should be heated, or the samples will 

 cool too quickly to allow complete separation of the fat. The 

 speed of the centrifuge should not be excessive. A few trials 

 will enable the operator to determine the correct speed. When 

 all the places in the centrifuge are not required the bottles 

 should be so distributed that the weight is balanced. 



The results will not be correct if the bottles are allowed to 

 cool down before reading. This is a common source of error, 

 causing differences of from : 05 per cent, to * 1 per cent. As 

 each bottle is read, promptly enter the reading in the test book. 



In testing separated milk or butter-milk, " precision " bottles 

 and sulphuric acid 1-816 should, as already indicated, be used, 

 and the reading ^at the upper curve of the meniscus should 

 be taken. With" these milks the bottles should be shaken 

 much longer, and should be left in the water bath for from five to 

 ten minutes before whirling. In addition, the whirling should 

 be continued for at least nine minutes. Should the centrifuge 

 have no provision for heating the bottles while in motion, they 

 should be taken out at the end of each three minutes and heated 

 in the water bath. Should dirt or curd (which always contains 



