97 
finite extent of the atmosphere. 
ous that could be chosen on the meridian ; but, for the pur- 
pose of seeing to the greatest advantage when the line of the 
centres is inclined to the horizon, it would be requisite to have 
the power of turning the slit and screen together at right 
angles to any line of direction of the centres. 
The only fixed star sufficiently near to the ecliptic, and 
bright enough to give any prospect of its being seen near the 
sun, is Regulus, which passes between the 20th and 21st of 
August ; but I have not yet had an opportunity of ascertain- 
ing within what distance from the sun this star can be dis- 
cerned. 
In the foregoing remarks, I have perhaps dwelt more upon 
the consideration of the solar atmosphere, than may seem 
necessary to those who have considered the common pheno- 
mena observable in the occultations of Jupiter's satellites by 
the body of the planet. Their approach, instead of being re- 
tarded by refraction, is regular, till they appear in actual 
contact ; showing that there is not that extent of atmosphere 
which Jupiter should attract to himself from an infinitely 
divisible medium filling space. 
Since the mass of Jupiter is full 309 times that of the earth, 
the distance at which his attraction is equal to gravity must 
be as ^309, or about 17.6 times the earth's radius. And 
since his diameter is nearly eleven times greater than that of 
the earth, ^7- = 1.6 times his own radius will be the distance 
from his centre, at which an atmosphere equal to our own 
should occasion a refraction exceeding one degree. To the 
fourth satellite this distance would subtend an angle of about 
3° 87' > so that an increase of density to times our common 
mdcccxxii. O 
1 
