in the length of the pendulum vibrating seconds , 349 
On the 23d July I began to observe coincidences, in the 
manner described in my Paper on the length of the seconds 
pendulum. Two series, each of ten intervals were taken 
each day; these are given at large in the Appendix, the 
results were as follow : 
Vibrations of the pendulum at Unst. 
The clock making 86450,63 vibrations in a mean solar day. 
Date. 
Barometer. 
Thermom. 
Vibrations 
in 24 hours. 
Correction 
for 
Temp. 
| Vibrations 
in 24 hours, 
I a t 62 d egr es. 
July 23 
24 
2 5 
26 
27 
28 
P. M. 
P. M. 
A. M. 
P. M. 
A. M. 
P. M. 
A. M. 
Thes 
A. M. 
P. M. 
A.M. 
P. M. 
30,00 
3 °> 3 ° 
29,90 
29,82 
29,84 
29,72 
29,95 
capement 
29,95 
30,00 
3°> 1 5 
30,20 
<5 
58,4 
59>3 
57>3 
59’7 
S 7>7 
59.0 
57’8 
was oiled 
56,8 
57,2 
54>3 
58.0 
86093,78 
86093,14 
86093,33 
86092,45 
86093,12 
86092,24 
86092,37 
without stoppir 
86091,69 
86091,62 
86092,5 I 
86091,57 
1,52 
1,14 
1,99 
0,97 
1,82 
1,27 
1,78 
tg the cloc 
2,20 
2,03 
3,26 
1,69 
86092,26 
86092,00 
86091,34 
86091.48 
86691,30 
86090,97 
86090.59 
k. 
86089.49 
86089.59 
86089,25 
86089,88 
Mean 
29,98 
00 
tC. 
86090,74 
The numbers in the above Table are deduced from the rate 
of the clock (gaining 50 s , 63) between the 22d and 28th of July. 
.For any other interval and rate, the mean of the vibrations 
during such interval is taken, and the difference between 
the corresponding rate and 50 s , 63 is added to, or subtracted 
from such mean number of vibrations accordingly as the rate 
of the clock has increased or diminished. The same method 
is pursued in all the subsequent experiments. In this manner 
the results contained in the next following table under the 
head of “ computed vibrations in a mean solar day” were 
obtained. 
