456 PHILOSOPHICAL TRANSACTIONS. [aNNO 1788. 



common air in this experiment was actually converted into nitrous acid, and was 

 sufficient not only to saturate the soap-lees, but also to dissolve some of the 

 mercury. The quantity dissolved however was very small, and not sufficient to 

 diminish sensibly the deflagrating quality of the nitre; so that the proof of the 

 air being converted into nitrous acid was as evident as if no mercury had been 

 dissolved. In this experiment, as well as the former, no indication of the soap- 

 lees becoming saturated was affi3rded by any cessation in the diminution of the 

 air; whereas in my experiments it was very manifest. I do not know what this 

 difference should be owing to, except to Mr. Gilpin's giving much stronger elec- 

 trical sparks than I did. In his experiments the metallic knob which received 

 the spark, and conveyed it to the bent tube, was usually placed at about 2-^ inches 

 from the conductor, so that the spark jumped through 2-^ inches of air, in pass- 

 ing from the conductor to the knob, besides from l-i- to 2-|- inches of air in the 

 tube; whereas in my experiments, I believe, the knob was never placed at the 

 distance of more than 1^ inch from the conductor, and the quantity of air in 

 the tube was much less; but the conductor and electrical machine were the same. 



Except this, the only difference that I know of in the manner of conducting 

 the experiment is, first, that Mr. Gilpin usually continued working the machine 

 for -1- an hour at a time, whereas I seldom worked it more than 10 minutes; and 

 2dly, that in Mr. Gilpin's experiments the common air in the reservoir bore a 

 less proportion to the dephlogisticated air than in mine; in consequence of which 

 it was necessary for him frequently to introduce common air. On this account, 

 the proportion of the 2 airs in the bent tube would be considerably different at 

 different times; but on the whole, the common air absorbed bore a greater pro- 

 portion to the dephlogisticated than in mine. 



Though the whole quantity of air absorbed in these experiments is known with 

 considerable precision, yet it is impossible to determine, with any accuracy, how 

 much of each kind was absorbed, on account of our uncertainty about the na- 

 ture of the air which remained at the end of the experiment. But if in the last 

 experiment we suppose that the air absorbed spontaneously between the 15th and 

 1 Qth of March was entirely dephlogisticated, and that what remained at the end 

 of that time was of the purity of common air, it will appear, that 40^0 of de- 

 phlogisticated and 25S8 of common air, which is equivalent to 4480 of pure 

 dephlogisticated air and 21 98 of phlogisticated air, were absorbed at the time the 

 electrification was stopped, and consequently the dephlogisticated air is f gg of 

 the phlogisticated air; whereas in my first experiment it seemed to be -f|4, and 

 in my last -f4-|^. But the quantity of acid produced, and consequently I suppose 

 the saturation of the soap-lees, depends only on the quantity of phlogisticated 

 air absorbed; and the effect of the greater or less quantity of dephlogisticated 

 air is only to make the nitre produced more or less phlogisticated. Now in this 



