of ascertaining the Magnitude of small celestial Bodies. 53 
before, will increase the diameter of these disks, but the outside 
rays alone will have a greater effect in reducing it, than when 
the inside rays are left to join with them. 
1 Cjth Experiment. Trial of Estimations. 
I placed two silver globules at a small distance from each 
other upon the post, but without measuring either the globules 
or their distance. When I viewed them with 522,7 they ap- 
peared in the shape of two half moons in an horizontal situa- 
tion. The unenlightened parts of them were also pretty distinctly 
visible. I estimated the vacancy between the cusps of the lunes 
to be i diameter of the largest. 
On measuring the diameters and distance under the micro- 
scope, it appeared that the largest was ,0312 ; a quarter of 
which is ,0078. The distance of the globules from each other 
measured ,0111. The difference in the estimation ,0033 is less- 
than part of an inch. 
The experiment was repeated with a change of the distance 
of the globules from each other. They were then estimated to 
be less than the diameter of the large one asunder, but full 
that of the small one. When they were measured it was found 
that their distance was ,02608, and the diameter of the small 
one was ,0247, which estimation is still more accurate than 
the former. 
20 th Experiment. Use of the Criterion, 
It remained now to be ascertained whether these half moons 
were spurious or real ; for although I could also imperfectly 
perceive the dark part of the disks of the globules, yet a doubt 
would arise whether the two halves were really of equal 
