MR. J. E. H. GORDON ON MEASUREMENTS OF ELECTRICAL CONSTANTS. 439 
It has been suggested to me that I should give some evidence that— 
(1.) The insulation of the apparatus was perfect. 
(2.) That the plates were good insulators. 
(3.) That their surfaces were not damp. 
With regard to the first point, an examination of the method will show that an 
absolutely perfect insulation is not necessary. 
As long as no electricity leaked to the small plates b and d, a leakage from the large 
plates ace would only produce the same effect as a diminution of the electromotive 
force of the coil current. It would diminish the sensitiveness of the instrument, but 
would not affect the results obtained by it. 
To prevent the electricity which leaked from the large plates leaking to, or acting 
by induction on, the small ones, the wires and supports belonging to the latter were 
completely protected by metal screens well connected to earth. 
The efficiency of this protection was shown by the perfect steadiness of the light 
spot when the earth connection was good, and its perfect obedience to the screw of a. 
The instant, however, that from any cause the earth connection became imperfect, 
the instrument became entirely unmanageable, and the light spot was in continual 
irregular and violent motion. I have, for the continuation of the investigation, lately 
introduced some further improvements, both in the insulation and in the arrangement 
of the metal screens. 
No error is to be anticipated from defective insulation, as with the apparatus in a 
leaky state, instead of incorrect results being obtained, measurement becomes impossible. 
With regard to any possibility of leakage occurring to one of the small plates, I may 
add that, after a few months’ experience of the balance, it would be impossible to 
mistake a deflection due to this cause for the regular deflection caused by induction. 
As to the second point, the only proof that the plates are good insulators is that the 
substances of which they are composed, glass, paraffin, &c., are generally considered 
to be so. 
With respect to the possible existence of a film of moisture, the following is the state 
of the case :—- 
Before putting any plate into the balance it was cleaned and rubbed dry with a 
washdeather. It was then rapidly passed over the flame of a spirit lamp to discharge 
any electrification that it might have acquired from the friction. 
The spirit lamp, of course, deposited a dew on the surface. This, however, was very 
slight, and could be seen to evaporate rapidly in the course of a few seconds. In one 
or two cases the plates were carefully dried by a fire, but no particular difference was 
noticed. 
The exact agreement with each other of the experiments on those substances, of 
which I possessed more than one plate, such as paraffin, ebonite, and particularly 
common plate glass, shows, I think, that this film of moisture can never have produced 
