tot 
of the 'Equatorial Instrument. 
From the preceding table of observations of the declination 
circle it appears, that these divisions were very little, if any, 
inferior in accuracy to those on the equatorial circle. And, 
finally it appears, that the errors, and probability of error, 
were as follows ; viz. in the equatorial circle, from 22 cor- 
responding observations of opposite divisions, 
Obs n . 
In 0, the error amounted to 2" 
therefore the probability 
against this error was 
5. - about i\ 
- 
- = 3 t to 1 
8. 1 
- 
- =2 to 1 
14. - - about o\ 
- 
- =1 to 2 
And in the declination circle, from 19 corresponding ob- 
servations of opposite divisions, 
Obs". 
therefore probability against this error. 
In 1, the error amounted to about 
2" 
= 18 to 1 
3 - 
= 5 to 1 
4 - 
1 
1! 
r-f 
O 
M 
8. 
0— 
2 
= 3 to 2 
Therefore from 41 double observations on both circles, 
Obs". 
In 1, the error amounted to 
2" 
viz. =40 to 1 
8. 
11 
r+ 
O 
12. 
1 
= H to 1 
22. - 
= 1 to 1 
We may conclude, therefore, 
that 
in both circles no error 
of more than 2" need be apprehended from the centre, and 
from the divisions taken together ; and that in general it will 
probably not exceed 1", on condition that the micrometer 
screw head is read off thrice, which in some observations 
