on the' Production of Dephlogflicated Air, log 
Experiment N° 1 5. 
The (ilk made ufe of in the laid experiment having been 
frequently ufed in the foregoing experiments, I was defirous of 
feeing the efFedt of making ufe of frefh filk ; and alio of vary- 
ing the proportion between the quantity of filk, the quantity 
of water, and the fize of the globe ; accordingly, at 6 o’clock 
P.M. upon the 13th of June, I filled a fmall globe, about 3 
inches in diameter, or (to afcertain its fize more exactly) which 
contained juft 20 cubic inches, with frefh fpring water, and 
17 grains of raw filk, wound in a fmgle thread, which had 
never been put into water, or otherwife ufed, fince it came out 
of the hands of the filk-winder. 
At the end of four days, viz. the 14th, 15th, 1 6th, and 
17th of June, this globe had only furnilhed | of a cubic inch 
of air, which, proved with nitrous air, gave \a + m=z 1,32, 
or 68 ; confequently was much worfe than common air. 
Up on the 1 8th, it began to produce good air, and during fix 
hours of funfhine it furnifhed 1 cubic inches, which, 
proved with nitrous air, gave \a + 3^= 1,15* or 285. 
The two following days (viz. the 19th and 20th of June) 
it furnifhed i T ~- cu hic inch of air, which, proved with ni- 
trous air, gave ia + yi~ 1,37? 01*263; after which i? totally 
ceafed to yield air, though expofed for feveral days in the fun’s 
rays. 
Total quantity of air produced z-Pio cubic inches; mean 
quality 1^ + 3^= 1,46, or 254. 
By this experiment it appears, that raw filk, when ufed for 
the firft time, does not immediately difpofe the water to yield 
pure air ; on the contrary, that it phlogifticates the air yielded 
by 
