IN METALLIC TUBES. 45 



fame procefs, there were found in the tube 3.5 ounce 

 meafures, which though it extinguifhed a candle, was of 

 the ftandard of 1.7 ; fo that feme atmofpherical air mud 

 have got into it. 



One ounce meafure of inflammable air expofed to heat 

 feveral hours in a fiher tube, and left to cool gradually, 

 came out two ounce meafures, of the ftandard of 1.42. 

 The fame quantity of the fame air, after continuing only 

 one hour in the heat, and examined immediately after it 

 was taken from the fire; was only 0.72, and wholly 

 phlogifticated. At another time I kept the fame quan- 

 tity of the air three or four hours in the fame heated tube, 

 and being examined immediately it was only 0.21 wholly 

 phlogifticated, fo that the tranfmiflion of air did not take 

 place while it was hot, but while the tube was cooling, 

 which I thought very extraordinary. 



The tube ot gold was melted by inadvertently heating 

 it too much before I had inade many experiments with it ; 

 and feeing reafon to conclude that its effedl on the air con- 

 fined and heated in it was no other than that of thofe of 

 filver, or copper, I did not renew it. I found, however, 

 that a meafure of inflammable air heated one hour in this 

 tube was fomething more than a meafure, and then ex- 

 tinguiftied a candle. There muft, therefore, have been 

 an addition to the air within from that without, though I 

 neglected to examine it by the teft of nitrous air. 



It was not neceflary to expofe thefe tubes, and the air 

 confined in them, to a red heat, in order to have this ef- 

 fect ; for 1 had a fimilar refult when I only placed them 

 near the fire in a degree of heat little greater than that of 

 boiling water. 



Air contained in clear water, is, as I have obferved, 

 fomething purer than that of the atmofphere ; but 

 when I filled a copper tube with water, and kept it a 

 whole day in the circumftance above mentioned, the air 



within 



