jg Dr. Herschel’s Obfervations 
provided the encroaching of light had no fhnre in the efTefh 
By an obfervation of the fixth fatellite, the refraction of Sa- 
torn’s atmofphere amounts to nearly the fame quantity ; for this 
fatellite remained about 14 or 15 minutes longer in view than 
it fhould have done; and as it moves about 2$ degrees m that 
- time, and its orbit is larger than that of the leventh, tne 
difference is inconfiderable. It is not my prefent intention to 
enter into a confideration of the amount of thefe retractions, 
otherwife we might perhaps find data enough tofubjeCt them to 
fome calculation. But what has been laid will lufficc to (hew, 
that very probably Saturn has an atmofphere of a confiderablc 
denfity. 
The next inference we may draw from the appearance ct 
the belts on Saturn is, that this planet turns upon an axis which 
is perpendicular to the ring. The arrangement of the belts, 
during the •courfe of fourteen years that I have oblerved them, 
has always followed the direction of the ring, which is what 
I have called being equatorial. I hus, as the ring opened, the 
belts began to advance towards the louth ; and to (hew an in- 
curvature anfwering to the projection of an equatorial line, or 
to a parallel of the fame. When the ring doled up, they 
returned towards the north ; and are now, while the ring paffes 
over the center, exadly ranging with the fhadow of it on the 
body; generally one on each fide, with a white belt clofe to 
it. When I fay, that the belts have always been equatorial, 
I pafs over trifling exceptions, which certainly were owing to 
local caufes. The ftep from equatorial belts to a rotation on ail 
axis is fo eafy, and, in the cafe of Jupiter, fo well afcertained, 
that I fhall not hefitate to take the fame confequence for granted 
here. But, if there could remain a doubt, the obfervations of 
