118 EXPERIMENT STATION. [Jan. 



The provision as to clean , dry mercury weighed on analytical 

 balances should be carefully observed. The scale equivalent 

 in mercury of the ordinary bottles is stated below, and that 

 of any other percentage can be readily calculated : — 



T7- Capacity, in Per Grams of Mercury 



^™°- Cent. at 20° C. 



Cream bottles, 50.00 135.4710 



Cream bottles 30.00 81.2826 



Milk bottles 10.00 27.0942 



Skim milk bottles .50 1.8547 



A number of quick methods, that are reasonably sensitive, 

 are employed to cull out the questionable bottles. For such a 

 purpose they are extremely valuable, but they should never be 

 considered official. This idea was incorporated into a section. 



Section 3. Optional methods. The mercury and cork, alcohol and 

 burette, and alcohol and brass plunger methods may be emploj^ed for the 

 rapid testing of Babcock bottles, but the accuracy of all questionable bot- 

 tles shall be determined by the official method. 



[b) Pipettes and Acid Measures. — With Babcock pipettes 

 and acid measures, as with other volumetric apparatus of simi- 

 lar character, the delivery is, or should be, the intent of the 

 graduation. There has been consideral)le discussion on this 

 point, but the recognized practice should not be set aside and 

 an exception made in this case. Relative to pipettes, Director 

 Stratton ^ wrote as follows : — 



The basis of test is the actual volume of water delivered by the pipette 

 when used in the maimer specified under Regulations for Testing. ^ 



He also went on to say that, while he could not state in ab- 

 solute terms the accuracy of such pipettes when used for milk, 

 in his opinion the error would not exceed .1 cubic centimeter. 

 In other words, a 17.6 cubic centimeter pipette would deliver 

 in milk approximately 17.5 cubic centimeters, — what has 

 usually been assumed. Probably this difference with milk is 

 largely due to viscosity, though other factors enter in. Calibra- 



1 Correspondence. ^ Circular No, 9, third edition. 



