12 PHILOSOPHICAL TRANSACTIONS. [ANNO 1781. 



man, 1.586. I next proceeded to examine the proportion of acid, water, and 

 fixed alkali, in nitre, in the same manner as I had before done that in digestive 

 salt, and found that 100 gr. of perfectly dry nitre, contain 28.48 gr. of acid, 5.2 

 of water, and 66.32 of fixed alkali. 



I shall now compare the result of these experiments with those of Mr. Hom- 

 berg. The specific gravity of the spirit of nitre which Mr. Homberg made use 

 of, was 1.349; an d of this, he says, 1 oz. 2 dr. and 36 gr., that is, 62 1 Troy, 

 are requisite to saturate 1 French oz. (472.5 Troy) of dry salt of tartar; accord- 

 ing to my computation, 6l3 gr. are sufficient; for this specific gravity lies be- 

 tween the tabular specific gravities by observation, 1.362 and 1.337, and is nearly 

 an arithmetical mean between them. The corresponding mathematical specific 

 gravity lies between the tabular quantities 1.315 and 1.286, and is nearly 1.300. 

 Now the proportion of acid and water in this is 2.629 °f ac 'd> ar >d 7-465 of 

 water; for 8.765 — 1.300 = 7.465 water, and 8.765 X .300 = 2.629 of acid; 

 and the sum of both is 10.044. Now, since 10.5 gr. mild vegetable fixed alkali 

 require 3.55 gr. of acid for their saturation, 472.5 will require 159.7; therefore 

 if 10.044 gr. of nitre contain 2.629 gr. acid, the quantity of this spirit of nitre 

 requisite to give 159.7, will be 6l3.2 nearly; and hence the difference between 

 us is only about 8 gr. 



2dly. Mr. Homberg says, he found his salt, when evaporated to dryness, to 

 weigh 186 gr. more than before; whereas, by my experiment, it should weigh 

 but 92.8 gr. more than at first. I shall mention the cause of this difference in 

 treating of tartar vitriolate, for it cannot be entirely attributed to the difference 

 of evaporation. 3dly. Mr. Homberg infers, that 1 oz. (that is, 472.5 Troy gr.) 

 of this spirit of nitre contains 141 gr. Troy of real acid: by my computation 

 it contains but 123.08 gr. of real acid. This difference evidently proceeds from 

 his neglecting the quantity of water that certainly enters into the composition of 

 nitre; for he proceeds on this analogy, 62 1 : 186.6 :: 472.5 : 141. 



The proportion of fixed alkali I have assigned to nitre is fully confirmed by a 

 very curious experiment of Mr. Fontana's, inserted in Rozier's Journal for No- 

 vember 1 778. This ingenious philosopher decomposed 2 oz. of nitre by distilling 

 it in a strong heat for 18 hours. After the distillation there remained in the 

 retort a substance purely alkaline, amounting to 10 French dr. and 12gr. Now 

 1 French oz. = 944 gr. Troy, and the alkaline matter amounts to 607 gr. Troy; 

 and, according to my computation, 944 gr. of nitre should contain 625 of alkali. 

 So small a difference may fairly be attributed to the loss in transferring from one 

 vessel to another, weighing, filtering, evaporating, &c. 



Mr. Lavoisier, in the Paris Memoirs for the year 1776, has given us, after 

 Dr. Priestley, the analysis of the nitrous acid. In 2 oz. French measure 

 (=945gr. troy) of spirit of nitre, whose specific gravity was 1.3 160, he dis- 



