[ 5 6 ] 
of December, There was then no ice on the river, 
which I could fee. 
I have fincethat time fettled the freezing point of 
this thermometer. I find that the fpace, on this 
thermometer,, contained between the freezing point 
and the top of the wood which covers the ball, is equal 
to 21 divifions of Fahrenheit’s fcale. It, therefore, fol- 
lows that the cold, at one of the clock in the afternoon 
on the 29th of December 1739, was fo great as to fink 
the mercury 21 divifions below the freezing point of 
Fahrenheit’s fcale, within this room, the windows 
being ftiut. 
On the 3 1 ft of December 1767, at 8 o’clock in the 
morning, a Fahrenheit’sthermometer withoutthe win- 
dow of the fame room, where it had remained all night, 
flood at 19*. divifions below the freezing point ; but a 
iimilar Fahrenheit’s thermometer within the room 
flood at 132. divifions below the freezing point ; 
therefore the cold, that morning, was greater without 
the room than within it by 6 divifions of Fahrenheit’s 
fcale. 
On the 7th of January 1768, at 8 o’clock in the 
morning, a Fahrenheit’s thermometer without the 
window of the fame room, where it remained all night, 
flood at 1 9JL divifions below the freezing point ; but 
a fimilar Fahrenheit’s thermometer within the room 
flood at 12 divifions below the freezing point ; there- 
fore the cold, that morning, was greater without the 
room than within it by 8 divifions of Fahrenheit’s 
fcale. 
From what has been faid, I think, we may fafely con- 
clude, that the cold, on the 28th of December 1739 
at one o’clock in the afternoon, was fo great, without 
the 
