THE ELECTRICAL RESISTANCE OF THIN LIQUID FILMS. 
457 
We are unable to place much confidence in the observations of the violet of the 
second and the red of the first order, as they were very few in number. Even in the 
case of the blue of the second order, though the quotients agreed well together, the 
difference between the figures in columns IV. and V. is too large to be satisfactory. 
We have therefore in most of our work taken 4 X 10 -5 centims. as the smallest apparent 
thickness which could be used where great accuracy was required. For the rest, the 
difference between the observations on Newton’s rings and the corrected table rarely 
exceed one per cent., while Newton’s scale in parts differs from both by as much as 
ten per cent, of the thickness. On the whole we think the optical observations alone 
show that the corrected scale above the blue of the second order is accurate to one 
per cent. 
V. Method of Experiment. 
In our previous experiments, described in the paper already referred to, a cylindrical 
film was formed, supported above and below by cylindrical platinum cups, and its 
resistance was measured by Wheatstone’s bridge. This method was unsatisfactory, 
first, because it took no account of any “ polarisation ” which might occur at the 
extremities of the film; and secondly, and this is the more important disadvantage, 
because it compelled us to measure the resistance of the film as a whole, instead of 
selecting any part which might be suitable to our purpose. 
The method since adopted and now to be described is free from the above-mentioned 
disadvantages. It depends upon the measurement of the difference of potential, pro¬ 
duced by the passage of a current, between two horizontal sections of the film. Steel 
sewing needles or gold wires (preferably the latter) are inserted in the film, and the 
difference of potential between them measured by an electrometer, and compared 
with that betw r een two other points in the same circuit separated by a known resist¬ 
ance. Three needles were actually employed, supported horizontally, and about 
15 millims. apart, the upper one being about 5 millims. below the upper cylindrical cup, 
and the lowest 10 millims. from the lower cup. Calling the needles 1, 2 and 3, counting 
from top to bottom, the difference of potential between 1 and 2, between 2 and 3, 
or finally between 1 and 3 could be measured as might be desirable. It is clear that 
this method disposes of any difficulties that may arise (1) from counter-electromotive 
force set up at the electrodes by the passage of the current, and (2) from irregularities in 
the substance of the film itself at or near its supports. Observation shows that such 
irregularities are of frequent occurrence. For example, it happens not unfrequently 
that a film, after thinning continuously for some time, begins to thicken from the 
bottom, the colours rising and being so crowded together that it is difficult to distin¬ 
guish them. Again, white flecks occasionally appear at the top, and more frequently 
black patches, forming themselves gradually into a ring of black, the resistance of 
which is, as we have already shown,* enormously great in comparison with that of a 
* Loc. cit. 
