422 
MR. BAILY ON THE CORRECTION OF 
If we reject the two results from the cylinder, (which I shall show, in the 
sequel, cannot be depended upon,) we shall have the mean of the rest equal to 
1’748: thus confirming the remark just made, that the factor for this addi- 
tional correction, in pendulums of equal length and of similar construction, 
seems to depend on the form and magnitude of the moving body, and is not 
affected by its weight or density. This result certainly does not accord with 
that deduced by M. Bessel, from his experiments with brass and ivory balls 
of nearly the same size as the present ones ; which result I have already stated 
to be 1*946*. M. Bessel’s experiments appear to have been conducted with 
very great care, and with all that accuracy and all those powerful talents for 
which he is so highly distinguished. At the same time however I would 
remark that I have carefully revised all my own experiments, and have not 
been able to discover any source of error : in fact, the general result is corro- 
borated by the uniformity in the results of the experiments with the other pen- 
dulums. The subject therefore is still open for further elucidation. In all 
M. Bessel’s experiments, he used wires of two different lengths ; one being 
about the length of the seconds pendulum, and the other differing from it the 
exact length of the French toise : or, in round numbers, about 39 inches and 
116 inches. The value of the factor which he has deduced, appears to be that 
which he considers common to both : but it perhaps may be a question whe- 
ther pendulums, differing so much in their lengths, give precisely the same 
value for the factor. 
Third set . — Results with the 2-inch solid Brass Cylinder. 
Flat sides horizontal. 
Flat sides vertical. 
10) Suspended by an 
iron wire. 
13) Suspended by a 
brass rod. 
1 1 ) Round side opposed 
to the line of motion. 
1 2) Flat sides opposed to 
the line of motion. 
Exp. 
ii 
Exp. 
ii 
Exp. 
ii 
Exp. 
n 
65 — 66 
67 — 68 
1-839 
1-880 
77—78 
79—80 
1-905 
1-940 
69—70 
71—72 
1-912 
1-928 
73—74 
75—76 
1-954 
1-946 
Mean = 
1-860 
Mean = 
1-922 
Mean = 
1-920 
Mean = 
1-950 
* See the note, in page 402. 
