356 Dr. Herschel's observations of the satellites 
IV. Exterior satellites that are very faint when at their 
greatest elongation, can hardly ever be seen at any other 
time when the orbits are contracted. 
V. The first satellite is probably larger than the second ; 
for though the latter is generally the brightest, it seems to be 
only in consequence of its being farthest from the planet. On 
comparing the limit of its disappearance with the number 302, 
expressing that at which the first satellite generally ceases to 
be visible, we find that the second satellite, upon its own scale, 
should not be lost in the light of the planet till it came within 
the limit of 224, instead of 295. 
VI. Both the satellites are subject to great variations of 
light, not owing to the changeable clearness of the air at diffe- 
rent times ; for by comparing the brightness of one satellite 
with that of the other when they are seen together, the state 
of the air will be of equal clearness to both, and yet their 
comparative brightness has been observed to be very diffe- 
rent: for instance, March 14, 1793, the first satellite was 
brighter than the second, when the distance of the former was 
to that of the latter as 172 to 235; and February 2 6, 1798, 
the first was small, and the second larger when the distance 
of the former was to that of the latter as 175 to 210. 
VII. 1 he variable brightness of the satellites may be owing 
to a rotation upon their axes, whereby they alternately pre- 
sent different parts of their surfaces to our view. These 
variations may also arise from their having atmospheres that 
occasionally hide or expose the dark surface of their bodies, 
as is the case with the sun, Jupiter and Saturn. 
VIII. The real angular distances of the satellites from the 
planet may be determined from the measures that are given in 
