1880.] Prof. J. Dewar. Studies on the Electric Arc. 89 



the lamp was allowed to work automatically, often for an hour or two 

 together. A continuous stream of dry air was kept circulating through 

 the interior, being afterwards passed through a series of wash bottles 

 containing dilute caustic soda, or directly through strong sulphuric 

 acid, to absorb the oxides of nitrogen. The nitrous acid was estimated 

 in the former case by titration with permanganate of potash, and the 

 total combined nitrogen by the mercury process. 



In this way many experiments were made with a Siemens lamp, both 

 with a long and short arc ; Jablochkoff's candles without any insulating 

 material were also employed with the highest intensity current of 

 a De Meritens machine, in order to have the greatest variety in the 

 character of the discharge. 



The stream of dry air was forced through at varying degrees of 

 speed, and was found to have a decided effect on the quantity of 

 nitrites produced, the more rapid stream giving the largest yield of 

 nitrites. 



The following tables give the quantity produced in a number of 

 different experiments. 



The nitrites were calculated as HN0 2 . 



1. Siemens' machine and lamp. 2. Jablochkoff's candles. 



Nitrates produced in 1 hour 



Siemens' 

 Long arc. I Short arc. 



De Meritens' 

 highest 

 intensity 

 current. 







Mgrms. 



Mgrms. 

 28 



Mgrms. 



1st experiment . . . 





193 



769 



2nd 





804 



97 



723 



3rd 





618 



73 



1,225 



4th 





500 



121 



548 



5th 





622 



90 



955 



6th 





474 



85 



1,006 



7th 





380 





1,257 



8th 





459 





964 



9th 









664 



10th „ 









489 



11th 









693 







509 mean 





930 mean. 



In these experiments, the total nitrogen estimated by the mercury 

 process was almost identically the same as the amount of nitrogen 

 obtained by a very careful dilution of the acid in a large quantity of 

 water and titration with permanganate, proving that the main product 

 was nitrous anhydride, which may be explained by the fact that the 

 quantity of oxygen in presence of nitrogen in the immediate neigh- 



