400 Dr. Dalton on the Constitution of the Atmosphere, 

 The sixth day 9 measures. 



The seventh day 4 measures. ( '^^'t ^^^^J ^^^^" "'^^'' 

 •' ( the surface. 



The eighth day 7 measures. *! These portions taken up 

 The ninth day ... 9 measures. > consisted nearly one 

 The tenth day ... 7 measures. J half of oxygen. 

 The fifteenth day 2 or .? measures. 

 From these experiments it would appear that by boiling 

 water briskly for three or four minutes, about' half of the at- 

 mospheric air previously in the water escapes along with the 

 steam. But it requires much longer boiling and keeping the 

 atmospheric air as much as possible from the surface of the 

 water to get the rest of the air expelled. It is never all ex- 

 pelled by boiling, except in the construction of a good water 

 hammer. Any one air not chemically combined with water 

 is easily and effectually expelled from it by repeatedly agita- 

 ting the water with another kind of air. 



It also appears that water deprived of its atmospheric air, if 

 kept at rest, acquires the air again slowly, and more so if the 

 surface exposed is small. But if violent agitation of the water, 

 so as to mix the atmospheric air and it intimately together, 

 be used, the full impregnation is effected in one or two minutes, 

 as I have elsewhere shown. 



Trough waters being mentioned above (3.) it may be well 

 to explain some of the circumstances affecting it. The waters 

 I use for the chemical trough is rain-water ; it is preferable 

 to pump water by its freedom from carbonic acid and earthy 

 salts ; it is slightly coloured at first when drawn from the cis- 

 tern, but it soon becomes clarified by standing : my trough 

 contains about nine gallons when in work. I take great care 

 to put nothing in it which can materially affect its purity ; 

 small portions of lime water and of some iron and other salts 

 are the chief impurities which are admitted ; no sulphurets or 

 hydrosulphurets are allowed to enter, and very little of either 

 acids or alkalies. I examine the state of the water occa- 

 sionally ; lately, after it had been more than half a year in the 

 trough, though not very frequently used, I had the curiosity 

 to examine its state before the trough was emptied. The 

 water was neutral by the colour test; it contained about 50 

 grains of saline matter in the gallon ; it was transparent, but 

 slightly milky; prussiate of potash gave sensible blue; oxa- 

 late of ammonia, muriate of barytes, and carbonate of soda 

 produced a white precipitate. The taste was like that of 

 earthy pump water. It had its full share of azotic gas, but 

 rather less than half of its share of oxygen gas; that is, it had 

 about 4 or 5 cubic inches of azote in the gallon, and only 1 

 cubic inch of oxygen. 



