20 <A. C. Twining on Occultations by the Moon and Planets. 
extent, favors the idea of an existing atmospheric limit in the 
instance of our own planet. 
Thus our present enquiry opens a remote possibility of bring- 
ing astronomical observation to bear conclusively upon the most = 
ea ental question respecting the constitution of material 
, pens also a possibility much less remote of determ- 
8 partially another point, and that of great interest in stellar 
astronomy,—I mean the actual angular magnitude of certain 
frac o leave a space, dno, 2. 
destitute of the star’s light; and how- 
ever nae the angle of refraction it is 
sufficient at the great distance of the ; 
earth to create a dark interval and a 
conseque arance of the » 
But let the planet move forward rela- ? a a 
touch the atmospheric boundary. Im- 
mediately the ray, a o’, will he re- 
fracted in parallelism to nd and so will 
: e star 
Sey 
bg 
‘what preceded, in consequence of be- 
ing more dispersed by passing ehroigh 
ere, 
e 
He 
Ee | 
8 
3 
® 
7 
3 
ts 
ty 
at the surface of another planet only 
be when the earth ha a advanced that 
amount in its apparent motion eo 
_ But if the atosphere should ys Sears pre would be no angular 
break, dno, and of course no disa In this even n/o’, ae 
other ra Sirsa bates -wé ald be flected, alth ough Tes and less, until the de- 
i e imperceptible. rama with the first percep “thes en of the star 
and progressivel persion of light its Piading would 
would not be rete a eye had r received m 
Pp 
e in the uncommon reagsitge <n sy = of my horizon. The instan . 
an fifteen seconds} “and even this could | 
