relative to Black, Watt, and Cavendish. 125 



a cylindrical form: this we filled with oil of vitriol and fair 

 water, of each almost a like quantity, and casting in half a 

 dozen small nails we stopped the mouth of the glass, which 

 was top-full of liquor, with a flat piece of dia palma provided 

 lor the purpose, that, accommodating itself to the surface of 

 the water, the air might be exquisitely excluded ; and speedily 

 inverting the phial, we put the neck of it into a small wide- 

 mouthed glass that stood ready with more of the same liquor 

 to receive it. As soon as the neck had reached the bottom 

 of the liquor it was dipped into, there appeared at the upper 

 part, which was before the bottom of the phial, a bubble of 

 about the bigness of a pea, which seemed rather to consist of 

 small and recent bubbles produced by the action of the dis- 

 solving liquor upon the iron, than any parcel of the external 

 air that might be suspected to have got in upon the inversion 

 of the glass, especially since we gave time to those little par- 

 ticles of air which were carried down with the nails into the 

 liquor to fly up again. But whence the first bubble was pro- 

 duced is not so material to our experiment, in regard it was 

 so small ; for soon after we perceived the bubbles produced 

 by the action of the menstruum upon the metal, ascending co- 

 piously to the bubble named, and breaking, did soon exceed- 

 ingly increase it, and by degrees depress the water lower and 

 lower, till at length the substance contained in these bubbles 

 possessed the whole cavity of the glass phial, and almost of 

 its neck too, reaching much lower in the neck than the sur- 

 face of the ambient liquor wherewith the open-mouthed glass 

 was by this means almost replenished. And because it might 

 be suspected that the depression of the liquor might proceed 

 from the agitation whereinto the exhaling and imprisoned 

 steams were put by that heat which is wont to result from the 

 action of corrosive salt upon metals, we suffered both the phial 

 and the open-mouthed glass to remain as they were in a 

 window for three or four days and nights together ; but look- 

 ing upon them several times during that while, as well as at 

 the expiration of it, the whole cavity of the glass bubble and 

 most of its neck seemed to be possessed by air, since by its 

 spring it was able for so long to hinder the expelled and am- 

 bient liquor from regaining its former place. And it was re- 

 markable that just before we took the glass bubble out of the 

 other glass, upon the application of a warm hand to the con- 

 vex part of the bubble, the imprisoned substance readily di- 

 lated itself like air, and broke through the liquor in divers 

 bubbles succeeding one another. 



" Having also another time tried the like experiment with 

 a small phial and with nails dissolyed in aquafortis, we found 



