Composition of Waters of Land-Drainage and of Rain. 157 



potassium gave rise, upon lengthened boiling, to a liberation of 

 iodine, slight indeed, but still suflicient to impair the delicacy 

 of the results unless great attention was paid to the duration of 

 the operation. 



Iodide of silver is not subject to either of these objections ; 

 it is not affected by hydrochloric acid, neither is iodine liberated 

 from it by sulphuric acid when sulphates are present.* Iodide 

 of lead was tried, but with a less satisfactory result. 



In the foregoing account I have supposed the water to be free 

 from organic matter, which is, however, seldom the case. The 

 presence of organic matter, by acting on the nitrate, would very 

 greatly interfere, unless steps were taken to counteract it. For 

 this purpose, in almost all cases, I have recourse to a solution of 

 permanganate of potash ; two or three drops of which are added 

 to the water in the act of boiling it down. The permanganate 

 effectually destroys the organic matter with the production of 

 peroxide of manganese ; it is added drop by drop, so long as the 

 amethyst tint imparted by it is destroyed ; when this, after some 

 time, remains permanent in the boiling solution, it is known that 

 the whole of the organic matter is oxidated. The excess of 

 permanganate is removed by adding a few grains of carbonate of 

 lead to the boiling solution, by which peroxide of manganese 

 and puce oxide of lead are precipitated. When the latter is 

 sufficiently concentrated, carbonic acid is passed into it as before 

 mentioned ; the liquid is filtered, and the evaporation is com- 

 pleted in the small globular flask. The complete dryness of the 

 product, and the absence, as far as can be, of carbonates, are 

 desirable on account of the necessity that the hydrochloric acid 

 should be as strong as possible, in order to act on the last 

 portions of nitrate. 



I now proceed to give instances of the amount of accuracy 

 which attends this process. 



The following are seven determinations of nitric acid in nitrate 

 of potash repeatedly crystallized. A weak solution of the nitrate 

 was employed, and a given quantity of it carefully measured by 

 a pipette, and evaporated to dryness. The quantity of nitrate em- 

 ployed was 1084 grains, representing 0-585 grains of nitric acid, 

 the results were — Citric Acid. 



1st Experiment 0-582 



2nd „ 0-585 



Snl „ 0-584 



4tli „ 0-581 



5th „ 0-580 



0th „ 0-584 



7th „ 0-583 



* In the instances given further on of the delicacy of this process for nitric acid, 

 -will be found one in which a sulphate was employed without altering the result. 



