A PENDULUM FOR THE REDUCTION TO A VACUUM. 
439 
occasion for entering so minutely into the particulars of these experiments, 
since it will be found that the most material inferences deduced therefrom, do 
not depend on nice shades of difference in the results of the experiments, but 
that the cases are marked by broader lines of distinction ; where the probable 
errors of observation and of computation would not make any appreciable dif- 
ference in the results, or in the consequences to be deduced from them. More- 
over, it will be seen that there is a regular march in the results of the several 
sets of experiments, which confirms the general accuracy of the whole : and it 
may be proper to state, once for all, that every value adduced is the result of 
at least four different experiments. 
I believe it has generally been considered, by persons who have paid atten- 
tion to this subject, that, in all funipendulous bodies in motion, the principal 
effect of the air, in adding to the inertia, is exerted on the body attached to 
the wire by which it is suspended ; and that the wire itself (which is generally 
the finest that can be used with safety,) has little or no influence in producing 
any alteration in the time of vibration : and consequently all their experiments 
and investigations have been conducted under this view of the subject. This, 
probably, is not far from the truth in the most usual cases which occur, and 
have been considered : but, as it is desirable that the direct effect of the air on 
each portion of the pendulum should be separately and distinctly ascertained, 
as accurately as possible for all cases that are likely to occur, I instituted some 
new experiments with a view to determine this point. 
In the pursuit of this inquiry I have found the suggestion and recommenda- 
tion of Professor Airy, “ to ascertain the weight of air adhering to each pen- 
“ dulum of experiment,” of very essential service : as it has enabled me not 
only to mark the direct influence of the atmosphere on the pendulum much 
more accurately and distinctly than by merely deducing the value of the factor 
n : but likewise to distinguish its influence on the several parts of the pendu- 
lum. In many of the following experiments the march of the values, indicating 
such influence, appears at first sight very complicated and anomalous : for, in 
some of them, (see the 14th set,) the weight of adhesive air seems to be less 
when the spheres are attached than when the bare rod is used ; and in others, 
(see the 1 9th set,) the weight of adhesive air dragged by a thin disc, appears to 
increase in a most extraordinary manner, merely by changing its position on 
3 l 2 
