j7+ D/\ BLAG DEN’S WJlory of 
“ hanging before in the (hade, and expofed it to the riling fun 
“ about eleven in the forenoon, to fee whether, when that lu- 
minary was lo low, it would produce any effed’t upon the 
44 in Hr undent. But to my great furprife, upon looking at it 
44 about noon, I found that the mercury had entirely fublided 
‘ 4 into the ball, though it was Handing as high as -6i°at ele- 
44 veil o’clock, and the fcale reached down to 238° below o. I 
44 could not perceive or think, that the air had changed fo fud- 
44 denly to fuch an extraordinary degree of cold. 1 therefore 
44 brought my thermometer, the only one I had left, within 
44 doors, and held it before the fire, when it quickly afeended 
44 to the ulual height in a warm room. Upon being carried 
44 out again into the open air, and placed in the lhade, it funk, 
44 as in the forenoon, to -6i°. Afterwards I expofed it once 
44 more to the fun-lhinb ;• but the fun having already begun to 
44 difappear behind the horizon, the quickfilver did not iubfide 
44 into the bulb as before. I then returned with the inftrument 
44 into the room, and held it in my hands before the fire, upon 
44 which the quickfilver fell back into the ball, where it left a 
44 vacuum or hollow bubble about the fize of a pepper-corn. 
44 When I inclined the thermometer, this bubble ran round the 
44 ball ; and after a few minutes the quickfilver role again to 
44 its former height. 
44 1 repeated thefe experiments feveral times atNebuloslock”[a 
fettlement about ten miles diftant from Sombio] 44 the lame af- 
44 ternoonand the following day, by carrying the thermometer 
44 out of the cold fometimes to the fire, and fome.tim.es into a > 
46 warm hut when the fame tiling happened, that the quick- * 
44 filver always fublided into the ball. .After my return to Tor- 
44 nea in April, I attempted to perform them again ; but when 
44 the cold was only a few degrees below the freezing point of 
3 44 water, 
