570 PHILOSOPHICAL TRANSACTIONS. [ANNO 1784. 



solution should be caught in a proper receiver, and decomposed by the gradual 

 admission of common air through water ; and the water, which thus becomes 

 impregnated with the acid of the nitrous air, should be added after the process 

 to the water through which the dephlogisticated air has passed. When the solu- 

 tion ceases to give any more nitrous air, the point of the tube of the retort 

 should be raised out of the water ; otherwise, by the condensation of the watery 

 and acid vapours which follow, a partial exhaustion will take place, and the re- 

 ceiving water will rise up into the retort and break it, or at least spoil the expe- 

 riment. A common receiver, such as is used in distilling spirit of nitre, should 

 be applied, with a little water in it, to receive the acid steam ; and it should be 

 kept as cool as can conveniently be done, as these fumes are very volatile. This 

 receiver should remain as long as the fumes are colourless ; but when they ap- 

 pear, in the neck of the retort, of a yellow colour, it is a mark that the mer- 

 curial nitre will immediately produce dephlogisticated air ; the receiver should 

 then be withdrawn, and an apparatus placed to receive the air. The rest of the 

 process has been sufficiently explained in my former letter. 



The phlogisticated nitrous acid, saturated by an alkali, will not crystallize ; 

 and if exposed to evaporation, even in the heat of the air, will become alkaline 

 again ; which shows the weakness of its affinity with alkalis when dissolved in 

 water;* a further proof of which is, that it is expelled from them by all the 

 acids, even by vinegar (which fact has been observed by Mr. Scheele.) I have 

 observed that litmus is no test of the saturation of this acid by alkalis ; for the 

 infusion of litmus added to such a mixture will turn red, when the liquor ap- 

 pears to be highly alkaline, by its turning the infusions of violets, rose leaves, 

 and most other red juices, green. This does not proceed from the infusion of 

 litmus being more sensible to the presence of acids than other tests; for I have 

 lately discovered a test liquor (the preparation of which I mean to publish soon) 

 which is more sensible to the presence of acids than litmus is ; but which turns 

 green in the same solution of phlogisticated nitre that turns litmus red. 



The unavoidable little accidents which have attended these experiments, and 

 which tend to render their results dubious, have prevented me from relying on 

 them as full proofs of the position that no acid enters into the composition of 

 dephlogisticated air; though they give great probability to the supposition. I 

 have therefore explained the whole of the hypothesis and experiments with the 

 diffidence which ought to accompany every attempt to account for the phenomena 

 of nature on other principles than those which are commonly received by philo- 

 sophers in general. And in pursuance of the same motives it is proper to men- 



• I have been informed, that Mr. Cavendish lias also observed this fact ; and that he has mentioned 

 it in a paper lately read before the r. s. ; but I had observed the fact previous to my know ledge of 

 his paper. — Orig. 



