Experiments on Co! el, ^91 
On Monday night, January 22, about twelve o’clock, hav 
lng occalion to take up a little fnow, there was obferveda cone- 
Hon among its parts rather greater than what might have been 
expected in a fubftance, at that time, fo much frozen* This 
circumftance was farther examined by the following experi- 
ment. A pane of glafs was laid on the fur face, of the fnow 
till it had acquired the temperature of 4-3, after which, with 
a bit of parchment equally cold, fome fnow was fc raped from 
the very lurface, and fliaken all over the pane, fo as to cover it 
in mod parts lightly. Upon now lifting the pane, and holding 
it with the fnow undermoft, the whole of it adhered, and it 
required fome fmart raps before the greater part fell away* 
What remained cleaved to the glafs with Hill a greater adhefion. 
The experiments related above afford further reafons againft 
the opinion of the difference of temperature betwixt the fnow 
or hoar-froft and the air depending upon evaporation. It would 
moreover appear, that this phenomenon depends not either upon 
the depofition of hoar-froft. What renders this the more 
probable is, that on laft year there was a much more copious 
depofition at times when the difference of temperature was not 
more remarkable. But allowing that a depofition had been 
found a neceflary circumftance, and always in proportion 
to that difference, the experiments on the capacities of hoar- 
froft and ice feem to fhew, that the fenfible heat which difap- 
pears enters not into the compofition of the hoar-froft ; other- 
wife the capacity of this fubftance for heat, compared to that 
of ice or common fnow, fhould be very different. It muft be 
confeffed, however, that the above mentioned experiment would 
have been more applicable to this reafoning, had it been made 
with hoar-froft given out in colder ftates of the air. 
■F f f a If 
