044 ANJ<fUAL REPORT OF THE Off. Doc. 



The divisions on the neck of the bottles should be uniform and 

 the lines run straight across the neck, and not obliquely. When the 

 numbers or lines become indistinct from having the blackened por- 

 tions washed olf, they cao be restored b'^- rubbing over the scale 

 with a lead pencil or with a cloth having a little black paint on it. 

 Each bottle should be numbered. A convenient way is to have the 

 number stamped on a copper ring and slip this over the neck of the 

 bottle. 



The accuracy of the scale on the neck of the bottle can be ap- 

 proximately tested as follows: Fill the bottle to the mark with 

 water, wipe out the neck of the bottle with a piece of filter or blot- 

 ting paper and then measure iato the bottle 2 c. c. of water with 

 an accurate pipette; this should fill the bottle to the 10 per cent, 

 mark. If bottles vary more than 0.2 per cent, in the w'hole length of 

 the scale from to 10 per cent., they should not be used. 



(2.) Pipette for Measuring Milk. — A pipette like that shown in 

 Fig. 9 is the form commonly used. This should hold 17.G c. c. when 

 filled to the mark. This will deliver about 17.5 c. c. of milk or 

 18 grams. It is important that the pipette should be accurate and 

 should hold exactly the amount stated above. Other forms of 

 pipette for measuring either milk or acid are shown in Figs. 11 

 and 12. 



(3.) Measure for Acid. — A cylinder of glass like that shown in 

 Fig. 10, with a lip to pour from and a single mark- at 17.5 c. c. is 

 the form commonly used. It is not necessary that this measure 

 should be completely accurate, since the amount of acid used can 

 be varied a little without affecting the test. In Figs. 13 and 14 are 

 shown automatic pipettes which may be used to advantage where 

 a large number of tests is made daily. This apparatus saves much 

 time but has the disadvantage of being somewhat expensive and 

 readily broken unless carefully handled. The automatic pipettes 

 shown in Figs. 11 and 12 will probably be found to be the most 

 convenient for the majority of those who use the test. 



(4.) Centrifugal Machine. — Various forms of centrifugal machines 

 have been devised for this work (see Figs. 15, 16 and 17). A wheel 

 less than twelve inches in diameter should not be used and it need 

 not exceed twenty inches. A wheel measuring twelve inches in di- 

 ameter should be made to revolve 1,200 times per minute, while, 

 for those of larger diameter, a smaller number of revolutions will 

 suffice, but not less than 700 revolutions per minute should be used 

 for the larger ones. It is better to use a machine in w^hich the mo- 

 tion is transmitted by cog wheels, since, when the motion is trans 

 mitted by belt or friction, there is danger of slipping; and the result 

 is much less motion than is intended and an imperfect separation 



