Remarks on the Crude Sodas of Commerce. 49 



termine the quality of a lot of barilla. We select a number 

 of fragments, which may be considered as a fair sample of 

 the whole ; pulverize them finely in an iron mortar, and then 

 weigh out two or three portions of one hundred grains each, 

 which may be put in as many tumblers, together with about 

 two or three ounces of water ; distilled water is best. After it 

 has stood for a few hours, being occasionally stirred with a 

 glass rod, carefully strain off the solution into a clean tum- 

 bler, through bibulous paper. Wash the residuum by adding 

 small quantities of water until it passes through the paper 

 tasteless. Add a solution of litmus until the alkaline solu- 

 tion becomes decidedly blue. Having then filled the phial 

 exactly with the diluted sulphuric acid, in order to observe 

 the change of color more accurately, place the tumbler con- 

 taining the alkaline solution, upon a sheet of clean white, pa- 

 per ; and then pour the acid in slowly and at intervals, agita- 

 ting at the same time with a glass rod, until the litmus be- 

 gins to assume a red color. We must now proceed still 

 more slowly and carefully ; the redness is at first faint and 

 delicate, and is produced by the carbonic acid gas evolved 

 and not by the sulphuric acid. On first adding the sulphu- 

 ric acid, no effervescence is observed ; probably because the 

 first portions of acid combine with that part of the soda that 

 remains in a caustic state. When the whole of the soda is 

 saturated, this will be indicated by the marked deeper red 

 color, and by the acid forming but a mechanical mixture 

 with the solution, as it is added drop by drop, without any 

 chemical action. By weighing the diluted acid remaining in 

 the phial, we can determine how much has been consumed 

 in saturating the alkali ; and, as we already know the num- 

 ber of grains of the diluted acid, required to saturate one 

 hundred grains of pure soda, by the rule of proportion we 

 can at once ascertain the proportion of pure alkali in the ba- 

 rilla. By repeating this operation with the two remaining por- 

 tions, we can make a still closer approximation to the truth. 



But this process, which is found to answer very well in as- 

 saying barilla and kelp, is an insufficient guide for ascertain- 

 ing the quality of the artificial soda, especially when indif- 

 ferently manufactured. This always contains a quantity of 

 the hypo-sulphite and hydro-sulphuret of soda, which are de- 

 composed by, and assist in saturating the sulphuric acid, and 

 thus give a too high return of the quantity of alkali ; espe- 

 cially as these substances are positively injurious in the man- 



Vol. XIV.— No. 1. 7 



