BAKING POWDERS. 583 



Professor Weber's analyses are rather superficial and incomplete, 

 probably being made under conditions that did not admit of thorough 

 quantitative work. He has overlooked entirely, for instance, the fact 

 of the presence of phosphoric acid in many alum powders. 



PROFESSOR CORNWALL'S REPORT ON BAKING-POWDERS. 



Following are the results of an examination of a large number of bak- 

 ing-powders by Prof. H. B. Cornwall, together with his description of 

 the methods of analysis employed, and his observations and conclu 

 sions. 1 



METHODS OF ANALYSIS. 



The analysis was directed toward determining the "strength" of the powders, or 

 their yield of carbonic-acid gas, and their composition, so far as to indicate the 

 nature of the chief active constituents. No great importance was attached to the 

 amount of starch or other legitimate "filling," which only has au effect on the 

 strength of the powder, nor was it possible to examine so large a number of samples 

 minutely as to the residues left by them. Especial attention was therefore paid to 

 the presence of possible objectionable constituents of the residues,aud to ingredients 

 that might render the use of the powders injurious. 



Carbonic-acid gas. This was determined with great care by boiling 1 gram (15.43 

 grains) of the powder with 125 to 130 cubic centimeters (about 4J fluid ounces) of dis- 

 tilled water in a roomy flask, connected with a Classen condensing, drying, and 

 absorbing apparatus (Classen, Quantitative Cheinische Analyse, 1885), the carbon di- 

 oxide being absorbed in soda-lime tubes, of which there were two, having their fur- 

 ther ends charged with carefully-dried chloride of calcium. The contents of the flask 

 were boiled, with proper use of a slow current of air, for one and one-half to one and 

 three-quarters hours, and the current of air was kept up for half an hour after remov- 

 ing the flame, so that the whole operation lasted from two to two and one-half hours. 

 Only in this way was the carbonic-acid gas with certainty to be expelled from the 

 somewhat viscid, starchy water solution and completely carried over into the ab- 

 sorption tubes. 



Tested before the analyses were begun, with a sample of probably one of the best 

 commercial bicarbonates of soda in the market, the absorption apparatus yielded 

 51.38 per cent, of carbon dioxide; measurement of the gas (see below) indicating 51.44 

 per cent. 



Tested in the course of the series of analyses by decomposing Iceland spar (crystal- 

 lized carbonate of lime) with citric acid in the presence of starch, in the proportion 

 used in the average of good cream of tartar baking-powders, the absorption apparatus 

 showed 43.83 per cent., theory requiripg, for absolutely pure carbonate of lime, 44 

 per cent. 



As a check analysis, when it could be properly done, the gas evolved from the pow- 

 der by 10 cubic centimeters of a mixture of one volume of hydrochloric acid, specific 

 gravity 1.2, with two volumes of water, in a Scheibler's evolution bottle, was collected 

 over mercury and measured, correction being made for atmospheric pressure, tem- 

 perature, and moisture, and also an allowance for the carbon dioxide retained by acid 

 of the strength used, as determined by tests with the Iceland spar. Enough baking- 

 powder was taken to give 90 to 110 cubic centimeters of gas. The results by measure- 

 ment averaged 0.12 per cent, below those by absorption and weighing of the gas, 

 probably on account of the difficulty of liberating the gas, even by violent shaking, 

 from the somewhat viscid liquid produced by the action of the strongly acid solution 

 on the starch of flour. The greatest difference by the two methods was 0.29 per cent. 



Whenever a sample showed a rather low percentage of carbonic-acid gas, or left a 



1 Report of the Dairy Commissipaer of New Jersey, 1888, p. 82. 



