formation of Ammonia^ &c. 19 



duced during its preparation. A carbonate of potash was 

 therefore prepared from pure tartar, rendered caustic by lime 

 calcined immediately preceding its use, the caustic solution sepa- 

 rated by decantation from the carbonate of lime, not allowed to 

 touch a filter or any thing else animal or vegetable, and boiled 

 down in clean flasks; but the potash thus obtained, though it 

 yielded no appearance of ammonia when heated alone, always gave 

 it when heated with zinc. 



The water used in these experiments was distilled, and in cases 

 where it was thought necessary was distilled a second, and even a 

 third time. The experiments of Sir Humphry Davy * shew how 

 tenaciously small portions of nitrogen are held by water, and 

 that, in certain circumstances, the nitrogen may produce am- 

 monia. I am not satisfied that I have been able to avoid this 

 source of error. 



At last, to avoid every possible source of impurity in the potash, 

 a portion of that alkali was prepared from potassium ; and as the 

 experiment made with it includes all the precautions taken to ex- 

 clude nitrogen, I will describe it rather minutely, as illustrative of 

 the way in which the other numerous experiments were made. A 

 piece of new glass tube, about half an inch in diameter, was first 

 wiped clean, and then heated red hot, a current of air passing at 

 the same time through it ; about six inches in length was drawn 

 off at the blow-pipe lamp, and sealed at one extremity. Some 

 distilled water was put into a new glass retort, and heated by a 

 lamp ; when about one-half had distilled over, the beak of the retort 

 was introduced into the tube before-mentioned, and a small portion 

 of water (about fifty grains) condensed into it. A solid compact 

 piece of potassium was then chosen out, and having been wiped 

 with a linen cloth, was laid on a clean glass plate, the exterior 

 to a considerable depth removed by a sharp lancet, and portions 

 taken from the interior by metallic forceps, and dropped succes- 

 sively into the tube containing the water before-mentioned. Of 



• Phil. Trans. 1807. p. 11. v 



C2 



