2^2 Obfervat'tons on the Freezing of Wuier^ 



manner, pure nitrous gas until i-i6th or i-20th of it is 

 filled. T then place the fla(k in a cli(h, and intercept its com- 

 jnunication with the atmofphere by means of mercury. Now, 

 if there be any air in the water, its oxygen gas mutt vinite 

 with the nitrous gas, producing more or lefs acid according 

 to the quantity of the oxygen gas. The folution of litmus, 

 therefore, mull acquire a red colour, and that, ceteris panb7iSy 

 more or lefs, and in a (liorter or longer time, according to 

 the quantity of oxygen prefent in the water. When you in- 

 tend to examine different kinds of water at the fame time, 

 after it has ttood two days introduce into it carefully, in fmall 

 quantities, by means of a fyringc, an alkaline foUuion ; and 

 from the quantity of the folution neceflary to reftore the blue 

 colour to the folution of litmus, w hich has been turned red, 

 you can eftimate very accurately the 7v/«/zv£ quantity of oxy^ 

 gen gas, and conlequently that of the common air mixed 

 with the water. 



A proof of this kind was neceflary in order to determine 

 \yith certainty whether water could be entirely freed from 

 oxysren gas, and confequently from common air j for azotic 

 gas is never prefent in water alone. 



Diftilled water, however carefully prepared, and immedi- 

 ately received as it comes oft" in flafks clofcly ftopped, cannot 

 bear the proof by nitrous gas. Water freed from air by 

 means of the air-pump again imbibes it as foou as the aip 

 has free accefs to it : even if the water be boiled in the flafk, 

 when cxpofed for a few moments to the atmofphere it be- 

 comes, in fonie meafure, rtddifh by this proof. 



After a great many fruitlefs attempts, I was never able to 

 free water entirely from air but by the following procels : 



I boil pure water for at leaft two hours in a large bafon } 

 colour it with litmus; warm a large itoppered flafk in it, and, 

 holdin"- it by means of a pair of forceps, fill It entirely with 

 the boilinc: water. I let the fialk remain for at leaft half an 

 hour with its aperture fome inches below the furface of the 

 boilina: water, frequently (triking its (ides wilh the forceps to 

 expel any air which may flii! remain in it. I then invert the 

 fiaflc. place it in an earthen velfel, and immediately introduce 

 nitrous gas : I then intercept all conmuinication with the 

 txternafair by mercury, of which there n)ult be a fiifliicient 

 quantity, on account of the diminution of the volume ot the 

 water, and particularly of the nitrous gas, by cooling; in 

 confequence of which the mercury rifes into the flalk to a 

 certain heii^ht. In this manner, 1 had the fatisfadion to fee 

 that the liquor retains its blue colour for weeks; a certain 

 proof thai water can be entirely freed from air. 



I wag 



