7 ^0 Col. roy’s Experiments for 
reafon to apprehend, that the thermometer fuffcred not 
only diredl but reflected heat. 
The fuperior barometer was fhaded with a parafol 
from the Sun, while its correfponding detached thermo- 
meter was expofed to his rays : wherefore, in the obfer- 
vations of high temperatures, the top of the column of 
air is ufually four or five degrees hotter than the quick- 
iilver in the barometer Handing in the fame air; and the 
mean heat of the column often exceeds very confidcra- 
bly the mean heat of the quickfilver in the tubes. 
In many of the coldeft of Mr. de luc’s obfervations, 
as well as in thofe of mean temperatures of about 50° or 
55 0 , the fuperior barometer is often the liotteft of the 
two, even when the furrounding atmofphere at the top is 
colder than at the bottom. This circumftance is ealily 
accounted for: Wood is known to be a bad conduc- 
tor of heat, to receive it flowly, and retain it long : that 
barometer, which was moved about from place to place 
upon the mountain, with a very fhort interval between 
the obfervations (as is fufficiently evident from the great 
number of ftations it pafled through in a limited time) 
mufi have acquired and retained heat fuperior to that of 
the atmofphere, and communicated it to the tube with 
which it was in contadl. Some difference would no 
doubt arife from this caufe, if the temperatures of the 
quickfilver 
