214 DR. DAUBENY ON THE ROTATION OF CROPS, ETC. 



rather, that the proportion of base to acid has some reference to the quantity of alkali 

 present, and is therefore dependent in a certain degree upon the manner in which the 

 previous calcination had been conducted. 



Supposing a plant to be rich in alkali, and to contain but little silex, it would seem 

 natural to expect, that the phosphoric acid would be united with three atoms of 



bases. 



When, on the contrary, the proportion of silica was large, a strong heat would 

 cause more of it to unite with the alkali, and hence there might be only enough of the 

 latter remaining to form a bibasic combination with phosphoric acid. Yet even here, 

 if a slighter heat had been applied, it might happen that a tribasic compound would 

 be produced. 



Thus we found, that in all the three crops of potatoes the phosphate gave a yellow 

 precipitate with nitrate of silver, and the same was also the case in the turnips ; but 

 in only one sample of beans, viz. the shifting crop, and in one of barley, which was 

 also the shifting one, did the same hold good. 



On the other hand, in two other samples of ash from the barley, and in two samples 

 of that from the beans, the phosphoric acid seemed, from the precipitate afforded by 

 nitrate of silver, to be united with two atoms only of base. 



It is easy to determine the amount of lime and of magnesia from either of the 

 liquids already operated upon, oxalate of ammonia being added to separate the 

 former, and, after neutralizing with ammonia the acid solution, phosphate of soda 

 throwing down the magnesia. 



In these respects the common methods were adhered to. 



The solution C. was reserved for the determination of the alkalies. 



For this purpose it is necessary to get rid of all the earths and metallic oxides 

 which may be present, which is accomplished by adding barytic water so long as a 

 precipitation takes place. That reagent of course throws down the whole of the sul- 

 phuric and phosphoric acids, the peroxide of iron, most of the magnesia, and most of 

 the lime*. 



The filtered solution may however contain a little magnesia and lime, and pro- 

 bably much barytes. 



To remove these, carbonate of ammonia is added in excess, and the precipitate 

 which is thrown down removed by filtration, after being allowed to stand until it 

 becomes heavy and granular. 



If this be duly performed, the remaining solution can contain only muriate of 

 ammonia and chlorides of the fixed alkalies. 



The former is removed by heat, and the dry chlorides then remaining will repre- 



* As the entire precipitate, excepting what consists of sulphate of barytes, is soluble in muriatic acid, we 

 may estimate the amount of sulphuric acid present, by treating it with the former acid, removing all that is 

 soluble in water by filtration, and lastly weighing the dried residue, from which the weight of sulphuric acid 

 may be readily deduced. 



