VS Mr. Faraday on the 



The first source of nitrogen which suggested itself was the 

 atmosphere: the experiment was therefore repeated, very care- 

 fully, in hydrogen gas, but the same results were obtained. 



The next opinion entertained was, that the potash might have 

 been touched accidentally by animal or other substances, which 

 had adhered to it in sufficient quantity to produce the ammonia : 

 the alkali was therefore heated red hot, as a preparatory step, and 

 afterwards allowed to touch nothing but clean glass or metals ; but 

 still the same effects were produced. The zinc used was selected 

 from a compact piece of foil, was well rubbed with tow dipped in 

 alkali, washed in alkaline solution, afterwards boiled repeatedly 

 in distilled water, and dried, not by wiping, but in a hot atmosphere ; 

 and yet the same products were obtained. 



All these precautions, \^dth regard to impurity from fingering, 

 were found to be essentially requisite, in consequence of the deli- 

 cacy of the means afforded by heat and turmeric paper for testing 

 the presence of ammonia, or rather, of matter containing its ele- 

 ments. As a proof of this, it may be mentioned, that some sea 

 sand was heated red hot for half an hour in a crucible, and then 

 poured out on to a copper-plate, and left to cool ; when cold, a 

 portion of it (about 12 grains) was put into a clean glass tube; 

 another equal portion was put into the palm of the hand, and looked 

 at for a few moments, being moved about by a finger, and then in- 

 troduced by platina foil into another tube, care being taken to 

 transfer no animal substance but what had adhered to the grains 

 of sand : the first tube when heated yielded no signs of ammonia to 

 turmeric paper, the second a very decided portion. 



As a precaution, with regard to adhering dirt, the tubes used in 

 precise experiments were not cleaned with a cloth, or tow, but were 

 made from new tube, the tube being previously heated red hot, and 

 air then drawn through it ; and no zinc or potash was used in 

 these experiments, except such as had been previously tried by 

 having portions heated in a tube to ascertain whether when alone 

 they gave ammonia. 



It was then thought probable that the alkali might contain a mi- 

 nute quantity of some nitrous compound, or of a cyanide, intro- 



