30 Prof. Forses’s Experiments on the Electricity of Tourmaline 
sulated on a thin stick of gum-lac, and then presented to the 
electrometer, resinous electricity was found of considerable in- 
tensity, shewing that a decomposition of electricity had actually 
taken place at the second surface of the glass, the resinously 
electric pole of the tourmaline forming the coating of its first 
surface, thus attracting the vitreous electricity of the second 
surface, and disengaging the resinous. Hence it is easy to see, 
that if the tourmaline remains sufficiently long warm to prevent 
the recombination of the electricities of its two poles, until the 
disengaged electricity at the second surface of the glass shall 
have been carried off by the air or otherwise, that recombination 
will be prevented, and the electric state will become compara- 
tively permanent. 
The use of Coutoms’s electrometer, in the manner I have al- 
ready described, affords an easy and general method of compar- 
ing the electric intensities of different crystals. For by measur- 
ing the maximum deviation produced by any specimen, we ob- 
tain, wholly independent of the exact temperature, a measure of 
its electric power, a measure independent of time, and, as expe- 
rience shews, little if at ail affected by the precise heat to which 
the crystal has at first been raised, at least within moderate 
limits. That the experiment admits, even with the most ordi- 
nary attention to collateral circumstances, of considerable accu- 
racy, I have proved by repeating the measures of the intensity 
of a particular specimen several times in succession. It will at 
once occur, that a source of fallacy must be guarded against in 
the loss of the electricity with which the disk of the electrometer 
is charged, which, as it is constantly diminishing by the contact 
of air, would give the intensities last measured in a series of com- 
parative experiments too small. In favourable circumstances, 
and by allowing the disk to remain some time charged before the 
series is commenced, it is surprising how little this error amounts 
