Manchester Memoirs, Vol. xliv. [igoo), No. i:5. ii 



at the rate of 270 grms. per horse-power per 12 hours, 

 thus showing an increase cf 60 grms. over and above the 

 quantities obtained in the former experiments. Con- 

 firmatory experiments, in which the above conditions were 

 observed as closely as possible, gave, within the h'mits of 

 experimental error, results agreeing with the first one. 



It was important to try the effect of increasing or 

 decreasing the current of air through the apparatus. So 

 far the air had been supplied to the flame at the rate of 

 about 20 litres per hour. 



Too great an increase in the air supply caused the 

 flame to become unsteady, and appeared to diminish the 

 oxidation taking place in the flame, and the oxides of 

 nitrogen, being also in a state of greater dilution, were 

 more difficult to absorb. On the other hand, if the air 

 supply was deficient, the oxides remained for too long a 

 period under the influence of the flame, and probably 

 suffered partial dissociation and a diminished yield of nitric 

 acid therefore resulted. 



In the light of the preceding experiments, it was 

 deemed interesting to tr}' the effect of varying the current 

 in the secondary circuit, and thus to increase or decrease 

 the temperature of the flame or combustion-chamber as 

 a whole. 



The current used to feed the primary of the trans- 

 former in the former experiments varied between 8 and 10 

 amperes. The high-tension discharge would therefore 

 convey a current of about 02 to 0"2 5 ampere. This was 

 now raised to from 0"3 to 0'38 ampere. After the flame 

 had been in operation for 12 hours, onl)- 71 grms. of acid 

 were found to have been produced. This is equivalent to 

 180 grams per horse-power per 12 hours, or j; the quantity 

 when only 0"2 ampere is used in the flame. This result 

 was confirmed by subsequent experiments. Increasing the 



