108 BURTON EDWARD LIVINGSTON 



different cups unequally. It is far better to employ a rotating 

 table which insures the same treatment to all cups of the group 

 and avoids, at a single stroke, most of the difficulties of atmos- 

 pher'c conditions. 



In restandardizing, the cups are operated for one period (10 

 to 48 hours, according to amount of water loss, which should be 

 great enough to read with small percentage of error, say 25 cc. 

 as a minimum loss), and the respective coefficients are calculated 

 as already described. The reading of the cup of known co- 

 efficient is multiplied by its coefficient, and the resulting standard 

 reading is then divided by the actual reading of the cup to be 

 standardized. The quotient is the required coefficient of the 

 last-named cup. If these agree (within 0.04 and 0.05) with the 

 corresponding original coefficients, the operation of the cups is 

 discontinued, and they are dried and wrapped, being now given 

 the newly determined coefficient. If the new coefficient does not 

 thus nearly agree with the old, the operation is continued for a 

 second period. If, now, the second new coefficients agree with 

 the first new coefficients (within 0.04 or 0.05), the average 

 of the two new coefficients is taken as the new coefficient to be 

 used. If agreement for any cup is not satisfactory, the opera- 

 tion is continued until a satisfactory average new coefficient is 

 obtained, or until the cup proves itself incapable of restandard- 

 ization, by showing continued fluctuations in the coefficient 

 value. Four or five periods, at the most, should always suffice. 



Renovation. Cups that prove incapable of restandardjzation 

 or appear too much soiled for further use, and those which have 

 developed a coefficient which is greater than 0.80 or less than 0.60 

 (for 8 cm. cups) and have therefore become suspicious, should be 

 renovated or discarded. Cups may usually be brought back to 

 a condition as satisfactory as that of new ones, by carefully scrap- 

 ing the whole of the porous surface with freshly broken glass 

 and then smoothing by means of sandpaper. After this treat- 

 ment they usually give perfectly satisfactory non-fluctuating 

 coefficients, somewhat different, however from the corresponding 

 original ones. The insoluble cups cannot be so treated, but they 

 may be heated to a glowing condition to remove organic matter, 

 and then washed with acid, etc., to remove salts. 



