cream. The samples should be kept in a cool place out of the sun. 

 Ground glass stoppered bottles are to be preferred, but owing to the 

 expense they are not generally used. Bottles with tight fitting corks 

 of first quality form the most satisfactory substitute. 



The preservatives most commonly employed for keeping compos- 

 ite samples are corrosive sublimate, bichromate of potash and form- 

 alin. The two former are sold in tablets prepared especially for the 

 purpose, and simply need to be powdered and dissolved in the sam- 

 ple ; formalin (formaldehyde) is sold in 40 per cent, solutions, and 

 the quantity needed for a composite can be easily added by means 

 of a small dropper or pipette. 



Corrosive sublimate tablets are to be preferred to other preserva- 

 tives, as they interfere least with the action of the sulfuric acid in 

 making the test. Where the total solids of milk samples are to be 

 determined in addition to the percentage of fat, formalin can be 

 used to advantage, as the amount necessary does not affect the accu- 

 racy of the test. Six to ten drops are sufficient to preserve a pint 

 sample several weeks if the sample is kept in a cool place. An 

 undue excess of any preservative should be avoided. 



A thorough preparation of the sample is of 

 Sampling vital importance, and is best accomplished by 



and pipetting'. gently rotating and by pouring back and forth 

 from the mixing vessel. Shaking and forcible 

 pouring is not permissable, as it is likely to cause partial churning 

 and is sure to inclose a large amount of air, rendering the sample 

 unfit for pipetting. All cream adhering to the sides and the stopper 

 of the sample bottle should be incorporated, and the resulting mix- 

 ture should not show any solid particles of fat. A small, fine wire 

 sieve is of great help in detecting and correcting a lumpy condition. 



After mixing, the sample should be pipetted immediately as the fat 

 globules rise rapidly toward the surface. The pipette should be 

 first rinsed with the sample, then filled again slowly, taking care to 

 avoid air bubbles, held in a vertical position when lowering the liquid 

 to the mark and read with the entire meniscus or crescent above the 

 line. In transferring to the test bottle, smearing of the neck should 

 be avoided as far as possible and the pipette blown clear. 



Sour milk can be readily brought into a condition suitable for 

 pipetting by the use of a wire sieve to break the clots or by neutral- 

 izing the acidity with a small pinch of powdered caustic alkali or 5 



