in the length of the pendulum vibrating seconds. 351 
By the Stars. U 
NST. 
From 
To 
Computed 
Vibrations 
in a mean 
solar day. 
Mean of 
Transits B 
or A coin- 
cidences. 
Correc- 
tion. 
Corrected 
Vibrations 
in a mean 
solar day. 
JMo. of 
Stars 
observed. 
Inter. 
of 
Transit. 
July. 
July. 
h. m. 
23 P. M. 
28 P. M. 
86090,71 
B. 1. 27 
+ .03 
86090,74 
4 
6 
23 P. M. 
28 A M. 
86090,93 
A. 1. 58 
— ,04 
86090,89 
1 
5 
23 P. M. 
25 P. M. 
86090,59 
B. 2. 37 
+ >05 
86090,64 
3 
3 
26 A.M. 
28 P. M. 
-86090,76 
B. 5. 52 
+ .12 
86090,88 
4 
3 
By the Sun. 
24 P. M. 
28 A.M. 
86090,85 
A. 1. 20 
—.03 
86090,82 
2 
4 
24 P. M. 
26 A. M. 
86090,55 
A. 0. 58 
— ,02 
86oqo,3 % 
2 
2 
26 P. M. 
28 A. M. 
86090,90 
B. 6. 11 
+ .12 
86091 ,02 
2 
2 
We have now to consider what authority attaches to each 
result, so that we may employ all the observations in obtain- 
ing a mean, and yet give to each set that degree of weight 
only to which it is entitled. 
The accuracy of any one result will evidently in the first 
place depend on the number of stars observed from which 
the rate of the clock is deduced ; and on this head as may be 
seen by examining the table of transits, there is little proba- 
bility of serious error. 
But the position of the transit intrument with respect to 
the meridian mark, requires the most minute care, and I soon 
discovered that to this, and to the accurate levelling of the 
axis, it was necessary to pay unceasing attention, as a devia- 
tion equal to the diameter of the silkworm's thread in the 
focus of the eye glass, would occasion an error in the time 
of the transit of a star amounting to about three tenths of a 
second. 
