VOL. LXX.] PHILOSOPHICAL TRANSACTIONS. 693 



pass, and could be put together by brass screws adapted to the divided extremi- 

 ties. Instead of a water trough, such as is used commonly, I made use of a 

 small round wooden tub, which I found on board the ship, and which I filled 

 with sea-water, fixing to the edge of this tub, by means of a screw, the brass 

 funnel, through which the air was to be let up into the glass tube. 



After having exercised myself during some time in performing the experiment 

 in a water tub much too small for the purpose, I at last acquired a habit of 

 doing it tolerably well. I then began to make my experiments regularly at about 

 11 o'clock; and I found the sea air at the place indicated of a superior purity to 

 any common air I ever met with since the month of June last (when I began to 

 engage in the course of experiments which have afforded me the materials of my 

 work lately published on Vegetables) either in my country retirement, or in 

 London. In 6 different trials made one after another in the short manner 

 described in my book, p. 278, et seq.* I found, that the 2 measures of air 

 (one of common and one of nitrous air) occupied from O.Ql to O.94 ; which 

 difference in the result, though but small in itself, was owing to the disadvantage 

 of not having a vessel deep enough to move the glass tube in with ease, for the 

 purpose of mixing the 2 airs together, and to my not having yet acquiretl the 

 habit of using the portable apparatus. 



I tried also the air of this spot in the manner used by Abbe Fontana, which I 

 have described in my book, p. 155;-f- the result of which trial was, that alter 

 the first measure of nitrous air was let up, the column of both airs occupied 

 1.86, or one measure and JgV of ^ measure; after the 2d measure of nitrous 

 air was let up, the bulk of both airs occupied 202, or 2 measures and -i^ of a 

 measure; and after the 3d measure of nitrous air was let up, the remaining bulk 

 of both airs occupied 2.96, or -jV_ of a measure: so that the remaining bulk of 

 both airs employed in the experiment (viz. 2 measures of common and 3 of 

 nitrous air, each making 100 sub-divisions of the glass tube) amounted to 2 

 measures and -^^ "^ ^ 3d measure, or to 296 sub-divisions, which being sub- 

 tracted from the 5 measures, or from the 500 sub-divisions employed, the 

 remainder was 2.04, which was exactly the quantity of both airs destroyed. 



Give me leave, before I proceed further, to recal to your memory some of the 

 experiments relative to this subject, made in my country retreat in the course of 

 last summer, and related in my book, principally in pp. 155 and 282, together 



• It consists in letting up into the glass tube one measure of common air, and after this one 

 measure of nitrous air, and shaking the tube forcibly in the water trough just at tlie moment the two 

 airs come into contact with each other. — Orig. 



+ It consists in letting up into tlie divided glass tube 2 measures of common air, and afterwards 3 

 measures of nitrous air, one after another, shaking die tube in tlie water trough after each measure 

 of nitrous air, and beginning constantly this motion exactly at the moment the 2 airs come into con- 

 tact with each other, or even before they meet, which is still better. — Orig. 



