390 Mr. Wilson’s farther 
to the very brim, fo that nothing communicated with the ex- 
ternal air but the fnow itfelf- The apparatus, in this ftate, waa 
let without doors for three hours each time, and then brought 
in to the lobby of the Qbfervatory, where the difti was again 
weighed : but in none of thefe trials did it ever appear, that 
any weight was loft. On the contrary, at the firft weighing, 
which was on Monday night, twelve o’clock, it had gained 
five grains. In the other two trials the increafe of weight was 
fcarce perceivable. 
The temperature of the air in the weft room of the Obferva- 
tory remaining very conftantly for near two days at +27°, a 
' difti of fnow, fimilar to the other expofed there, was found to 
lofe weight very feniibly, and for the moft part at the rate of 
two grains in an hour. Notwithftanding this, the fnow 
thus wafting or evaporating had no power of finking the ther- 
mometer below +27, the temperature of the furrounding air; 
though at one time it was fanned for four minutes by a piece of 
paper fattened to the end of a long ftick. Not to difturb the 
> uniform temperature of this room during thefe experiments, 
care was taken to ftay in it a very fhort time at every vifit, and 
to keep the door and the window- fhutters clofe. 
On Chriftmas-day we had a froft, which in the morning 
made the thermometer in air point to +21 ; and during the 
preceding night there had been a profufe depofition of hoar- 
froft. A pound of this was collected, and its capacity for 
^ heat compared to that of ice, and found equal a.s nearly as 
could be judged. Before making the two mixtures neceffary 
for this experiment, the ice was reduced to a powder, and 
fpread out on a paper befide the hoar-froft till both had acquired ; 
the fame temperature*. 
On 
