A PENDULUM FOR THE REDUCTION TO A VACUUM 
443 
of this paper ; to which I added a third, 3 inches in diameter, weighing 29114 
grains, and whose specific gravity I found to be 8’020. The ends of 
the brass and steel rods were screwed into the several spheres: but 
the copper rod was attached by means of an adapting screw. The re¬ 
sults are given in the following Table: where it will be seen that in 
each of the three rods the value of n is diminished by appending either 
of the spheres thereto. The march of these values, however, does not 
appear to be very regular. Indeed, the conducting of the experiments 
when the spheres were attached to the ends of the rods, required great 
attention on account of the slowness of the vibrations, and the conse¬ 
quent frequency of the coincidences with the mean solar clock, with 
which they were compared ; and they may consequently be subject to 
some little uncertainty *. In the case of the brass and steel rods the 
intervals of the coincidences did not exceed eleven seconds : but, on 
the other hand, I sometimes took a mean of several thousand of them, 
for the result. 
Fourteenth set.— Results with the spheres at the ends of the long rods. 
Diameter 
of the 
spheres. 
Copper rod. 
Brass rod. 
Steel rod. 
No. 
n 
Weight 
of adhe¬ 
sive air. 
Weight 
due to 
sphere 
alone. 
No. 
n 
Weight 
of adhe¬ 
sive air. 
Weight 
due to 
sphere 
alone. 
No. 
n 
Weight 
of adhe¬ 
sive air. 
Weight 
due to 
sphere 
alone. 
inches. 
0-00f 
1- 46 
2- 06 
3-03 
42 
45 
46 
47 
2-932 
2-458 
2-234 
1-873 
4-904 
4- 564 
5- 076 
6- 425 
43 
48 
49 
50 
4-083 
2-356 
1-982 
1-933 
1-484 
1-417 
1-973 
4-868 
44 
51 
52 
53 
7-530 
2-344 
1-793 
1-759 
0-479 
0-834 
1-259 
3-670 
0-342 
1-273 
3-251 
0-463 
1-157 
4-0661 
0-607 X 
1-063 
3-480 
* Should it be considered desirable to repeat these experiments with greater accuracy, arrangements 
might he made for that purpose, by altering the rate of the mean solar clock; which I was unwilling 
to disturb during the course of the present experiments. 
f The values in the first line are the same as those given in the preceding Table ; and are here in¬ 
serted in order to show their relative values as compared with the results when the spheres are attached 
to the rods. This plan will he pursued in the subsequent experiments. 
% These two experiments (with the pendulums No. 50 and 51) are very unsatisfactory; and are 
marked as such in my journal. It was consequently my intention to have repeated them : but the 
subject was overlooked till it was too late. I should propose their being rejected altogether. 
