CHEMICAL SCIENCE. 217 



ities. 7. The water of certain wells possesses a very disagreeable earthy 

 taste. This taste is derived from alumina, held in solution by carbonic 

 acid. It is observed that those well-waters which contain most of this 

 base have the strongest earthy flavor. 8. It results from these exper- 

 iments that a classification of drinkable waters, based on the relations 

 which exist between the sulphates and the chlorides, must be a defect- 

 ive one ; for this relation varies with respect to the same kind of water 

 within limits of considerable extent, and it is never certain that the 

 water operated on has not met in its course, either above or below the 

 soil, with substances which have altered and changed the proportions 

 in which these salts enter into its composition. L'lnstitut., No. 851. 



QUANTITATIVE SEPARATION OF POTASH AND SODA. 



ROSE has found that the fluo-silicates of potassium and sodium are 

 wholly insoluble in a liquid containing alcohol. When fluo-silicic acid 

 in excess is poured into a solution of a potash or soda salt, and a vol- 

 ume of strong alcohol, equal to that of the whole liquid, is added, the 

 whole of the potash or soda present is precipitated in the form of a 

 fluo-silicate, which is to be washed with strong alcohol, diluted with 

 an equal volume of water. Baryta may be separated from its solutions 

 in a similar manner ; a very small proportion of alcohol, however, suf- 

 fices for its complete precipitation. 



QUANTITATIVE SEPARATION AND DETERMINATION OF PHOSPHORIC ACID. 



REYXOSO has proposed a method of separating phosphoric acid quan- 

 titatively, based upon the fact that phosphate of peroxide of tin is inso- 

 luble in nitric acid, while other phosphates are soluble. A weighed 

 portion of pure tin is to be introduced into a vessel with the phos- 

 phate, nitric acid added in excess, and the whole boiled. When the 

 whole of the tin has been oxidized the mass is to be thrown upon a 

 filter, washed, dried, heated to redness over a spirit lamp, and weighed. 

 The difference between this weight and the weight of peroxide of tin. 

 yielded by the tin empk>3 r ed, gives the weight of the phosphoric acid. 

 The filter must be carefully burned ; a few drops of nitric acid being 

 added to prevent reduction. The flame of the lamp must also be care- 

 fully managed, so as not to mount too high ; and, finally, the mass 

 must be rapidly weighed, as it is very hygroscopic. The process gave 

 good results when applied to the analysis of pyrophosphate of soda, the 

 only compound which Reynoso appears to have examined. Comptes 

 Rendus, 33, 358. 



VOLATILITY OF PHOSPHORIC ACID IN ACID SOLUTIONS. 



MR. J. B. BUNCE, of New Haven, publishes, in Silliman's Journal, 

 the results of some experiments, made to ascertain the extent of loss 

 experienced in analysis in consequence of the volatility of phosphoric 

 acid. .554 grammes of phosphate of soda were dissolved in a gill of 

 water : hydrochloric acid was then added, and the whole evaporated 



19 



