on Atmospherical Refraction. 
observations I have been able to make of the sun, to ascertain 
the agreement of the two solstices, are subjoined; these 
appear to confirm the refraction, as deduced from the circum- 
polar stars. 
In the course of my observations I have noticed, that 
applying the correction for the thermometer without, gives 
the most accurate result. The difference is very sensible in 
great zenith distances, from the greater quantity of refraction ; 
and we may reasonably infer, that when the front shutter of 
the Observatory is opened, the horizontal current of the air is 
of the same temperature as without ; although from the short 
time of the shutter being opened, during the observation, it is 
not indicated by the thermometer in the telescope. I have 
therefore constantly used the thermometer without, for the 
correction, when I have opened the front shutter ; but on all 
other occasions, I have applied the correction for the thermo- 
meter within. The instrument is protected from the horizontal 
current of air, when the sloping shutters in the roof only are 
opened; the front shutters being five feet above the graduated 
circle. 
Having formerly proposed certain factors for the thermo- 
metrical correction of the refraction, and now finding them 
relatively the same for the thermometer within and without, 
changing the ratio for each degree of Fahrenheit 5 I 
shall adopt, hereafter, the following formula. Putting for 
the degree of the scale ; then for the thermometer within, 
49°— h° X ,0023 when below the mean ; 4^** — x ,0022 
when above the mean : and for the thermometer without, 
45“ — ii°x 0020 ; will produce the respective factors. When 
/z® is less than tlie mean, these will be positive ; when greater 
MDCCCXiv. Y y 
