131 
it is notorious that even small reflectors act very imperfectly : 
and he was therefore unprepared for any tolerable action of 
this gigantic speculum. In the day time it was of course 
colder than the air, and, if uncovered before that had sunk to 
its temperature, was covered with dew: when this went off it 
always defined sharply. The huge mass of metal cooled much 
more slowly than the atmosphere; and as the difference in- 
ereased, the performance of the telescope was deteriorated. 
This arose from no change of figure, as he satisfied himself 
by throwing the stars out of focus ; it was probably the result 
of currents in the tube occasioned by this difference of tempe- 
rature. How far it will be possible to obviate this by mecha- 
nical means, remains to be tried ; but it is certain that the in- 
convenience does not increase in a higher ratio than the power 
of the telescope, as he had formerly apprehended. On the 
same nights, it defined quite as well as the three-feet with a 
far lower power ; and therefore, it is reasonable to expect that 
it can be used with advantage much more frequently than he 
once supposed. 
Enormous as is its illuminating power, it might be in- 
creased one-third, by using it with the front view, supposing 
_ it can be properly figured for this oblique action. Without 
that, he fears that in an instrument where the aperture is so 
large compared with the focal length, the definition would be 
imperfect. He verified this by an experiment with the 
three feet, and found that though the light was increased 
quite as much as he expected, yet the perfection of the image 
was utterly destroyed for large stars. There was no exact 
focus, but merely two places where the sections of the cone 
of rays were smallest. One, the least exceptionable, shewed 
a flare in the direction of the slope like a comet’s tail: at the 
other this disappeared, but the star became a sort of curved 
rectangle with rays from its corners. In the Newtonian 
form this speculum a few nights before had defined Z Orionis 
very well with 500; but now y Leonis could not be seen 
