BAKING POWDERS. 591 



a lamp is placed under it, and the contents gradually heated to boiling, 

 gentle aspiration being made at the same time. The operation is finally 

 finished by the funnel tube being opened, and air, free from CO 2 , drawn 

 through it and through the apparatus, the contents of the flask at the 

 same time being kept at ebullition. This is continued for fifteen min- 

 utes, when the absorption tube is removed from the apparatus, allowed 

 to cool, and weighed. Its increase in weight gives the amount of CO 2 

 absorbed. The determination of the available CO 2 was conducted in a 

 similar manner, with the substitution of pure boiled water in the funnel 

 tube instead of acid. After the sample has all bei'n acted on, the con 

 tents of the flask are just brought to a boil, then the lamp is removed 

 and air is drawn through the liquid for exactly fifteen minutes. The 

 conditions were made as nearly alike as possible for each sample in 

 this estimation, for, different results can readily be obtained by vary- 

 ing them. The above conditions are believed to be as close an ap- 

 proximatioc to those actually obtaining in the use of the powder as can 

 be arrived at in an ordinary chemical analysis. Prolonged boiling of 

 the liquid residue is inadvisable, for in case the ingredients in the pow- 

 der are not accurately proportioned, and a considerable^excess of bi- 

 carbonate is present, long boiling will liberate gas from it after the acid 

 ingredient has all been neutralized, and thus a high result is obtained 

 from a poorly-made sample, while in one where the bicarbonate is not 

 greatly in excess of the proper amount, the above procedure will readily 

 give the full amount available. * 



*In some experiments made to determine the amount of carbonic acid driven off from 

 bicarbonate of soda on heating its water solution, the following results were obtained: 

 (1) Just brought to aboil under the same conditions as in the determinations made 

 above, 6.99 per cent, of the weight of the bicarbonate was obtained ; (2) Boiled 15 

 minutes, 16.17 per cent, was obtained; and, (3) boiled labours, 20.70 per cent., or about 

 the full quantity of acid carbonate. 



ESTIMATION OF STARCH. 



This estimation was made by the well-known method of conversion 

 by heating with acid and the determination of tbe copper oxidereducing 

 power of the resultant solution. While by no means satisfactory, this 

 is probably the best method we have at present for starch estimation. 

 No difficulty was found in its application to all classes of baking-pow- 

 ders, the other ingredients offering no obstacle to its proper perform- 

 ance. To insure agreeing results it is very essential to conduct the con- 

 version under precisely the same conditions in all cases. 



The following is the detailed procedure : 



From 2 to 5 grams are weighed out and transferred to an Erleuineyer 

 flask; to it are added about 150 to 200 cubic centimeters of a solution 

 of hydrochloric acid which has a strength of 4 per cent, of the acid gas. 

 The flask is provided with a cork, perforated for the passage of a con- 

 densing tube about 1 meter in length. The conversion is accomplished 



