the effects of parallax, &c. of certain fixed stars. 335 
temperatures of the external and internal air. The refrac- 
tions have been computed by the internal thermometer. 
Now, at the summer observations of a Aquilae and & Lyrae, 
&c. which take place between sunset and midnight, the 
external thermometer is oftentimes several degrees lower 
than the internal ; the average is between 4 0 and 5 0 . At the 
winter observations, the external thermometer at the hours 
when these stars are observed, averages only about one or 
two degrees lower. Hence, if the refractions were computed 
by the external thermometer, the results as to a Aquilae and 
other stars of considerable zenith distance, would be less in 
favour of parallax. 
But several circumstances induce me to conclude, that the 
true result is to be deduced from the internal thermometer. 
In a multitude of instances, were the external thermometer 
used, great discordances would take place. A great number 
of observations of circumpolar stars, made with a view to 
determine the constant of refraction, have given me nearly 
the same mean refraction as that determined by M. Delambre 
from a great mass of observations of his own, and of M. 
Piazzi, and which was also confirmed by the direct experi- 
ments of M. M. Biot and Arago on the refractive force of 
air, whereas had I computed by the external thermometer, 
the constant of refraction would have been much less. Also 
I have found the mean zenith distance, computed by the 
internal thermometer, when it stood several degrees higher 
than the external, fully equal to that found when the external 
and internal thermometer stood at the same height. This 
has been particularly the case as to the Pole Star below the 
Pole. The circumstances of the results I have obtained by 
