THE ELECTRICAL RESISTANCE OF THIN LIQUID FILMS. 
465 
deflections were, as a rule, within the limits 80 and 250 divisions. In some of the 
experiments the lamp and scale were removed to a distance of 1’6 metres from the 
electrometer, and a sharp image of the slit on the scale obtained by means of a lens. 
In both positions of the scale it was found that the above empirical rule was approxi¬ 
mately true, and that the difference of potential corresponding to a given electrometer 
deflection d was proportional to d (1 — dx O'0001). Similar observations were from 
time to time repeated, and always with the same result. 
According to Maxwell (‘Electricity and Magnetism,’ vol. i., p. 274), the deflections 
of the electrometer, when of moderate magnitude, are proportional to the product 
(A—B)(C—Ap) 
where C is the potential of the needle and A and B those of the quadrants. Now, in 
calibrating the electrometer, the potential of one pair of quadrants (say B) was kept 
constant, and the other gradually raised, and thus our empirical correction is made up 
partly of the conversion of arcs into tangents, and partly of the change due to increasing 
values of A + B. But since a correction which holds for one constant value of B will 
not hold for another, it was necessary to test the validity of our formula for the range 
of potentials studied. For this purpose two resistances, each of 100,000 ohms, were 
included in a circuit, and a piece of carbon paper with a resistance of about 700,000 
ohms interposed between them. The mean electrometer deflections corresponding to 
the equal resistances were 127'8 5 and 127‘55, which differed by only 0‘3 of a division. 
Hence the formula may be regarded as valid, whatever be the position in the circuit of 
the resistance considered. 
Fig. 6. 
Method of observation.— The order of observations of the electrical condition of a 
film was as follows: The film having been formed and adjusted, and the needles fixed 
in their proper places, the copper wires B (fig. 6) were inserted in the mercury cups 
belonging to the particular pair of needles between which the resistance was to be 
measured. The connecting pieces C and D, consisting each of a strip of ebonite with 
