REPORT ON ELECTRICITY, MAGNETISM, AND HEAT. 9 
each other, or less than of the quantities taking the mean of 
the absolute errors. Under the various chances of inaccuracy 
which belonged to the nature of the observations, this was as 
near a coincidence as could be expected. 
M. Poisson’s calculus also explained the electric spark which 
passes when an electrised body is brought very near another 
body; for it appeared that the thickness or density of the elec¬ 
tric stratum on the nearest points of two spheres would approach 
to infinity, and consequently the tension would become infinite, 
as the spheres approach to contact. Before this contact, there¬ 
fore, a spark passes from one to the other. 
It will be recollected that the numerical coincidences of obser¬ 
vation and calculation are in addition to the explanation of elec¬ 
trical attraction and repulsion, and of all the general phenomena 
of induction which the theory supplies, or rather which force 
some such theory upon us. It cannot be denied, then, that the 
Coulombian theory stands upon strong grounds so far as the 
statical phenomena are concerned. Still, it is always desirable 
that a philosophical theory should be confirmed and verified by 
observations and measures added to those on which it was ori¬ 
ginally founded; and it is to be remembered that the calcula¬ 
tions of Poisson have been confined to spherical conductors. A 
true theory will commonly be upheld by a constant series of con¬ 
firmatory experiments, made by students or professional obser¬ 
vers, whether or not it is thought worth while to publish the 
results of such trials. I do not know that many such mea¬ 
sures of the phenomena of statical electricity have been taken. 
Among the most important of such measures are those made by 
Mr. Snow Harris, which were exhibited before a section of the 
British Association at Cambridge in 1833, and are described in 
the Transactions of the Association for that year, p. 386. These 
experiments have since been more fully detailed in a paper printed 
in the Philosophical Transactions for 1834, Part II. p. 213. His 
invention of a mode of measuring the quantity of electricity by 
what he calls a unit jar , is valuable as offering a new mode of 
verifying, or at least testing, the general laws of electrical action; 
and the considerable amount of the statical forces which were 
brought into action in his experiments inspires more confidence, 
at least at first sight, than the extremely minute quantities em¬ 
ployed by Coulomb. His forces were in many cases measured 
directly in grains. The detailed comparison of the results with 
experience w r ould be a matter of some labour, and indeed of some 
difficulty; for in some cases circumstances which would affect 
the result are not stated, and the electric fluid was distributed 
through conductors so complex in their form that it would not 
