520 
PROFESSORS A. W. REIROLD AXD A. W. RUCKER OK THE 
Table XIII. 
Electrical 
thickness of ■white. 
Ratio to 
true thickness. 
Air. 
387 
3-99 
Oxygen. 
364 
3-75 
Air. 
424 
4-37 
Air (later experiments; see 
p. 522 . 
434 
4'47 
The ratio obtained when oxygen was used is less than in the other cases, but the 
difference is small when the uncertainty of the colour estimates is allowed for; and 
the agreement is, at all events, sufficiently close to prove that a very great increase in 
the quantity of oxygen in the neighbourhood of this film produces little or no change 
in the specific resistance. 
The nitrogen used was produced by heating a mixture of ammonium chloride and 
potassium nitrite, and was passed through dilute sulphuric acid and caustic potash. 
Whatever the cause may have been, we found that in this gas the film did not thin 
regularly, so as to exhibit pure and unmixed colours. Estimates of the thickness 
were therefore impossible. 
Before proceeding to discuss these results further, it may be well to ask whether 
the large changes in specific conductivity affect the black films only, or whether 
similar phenomena can be detected when the thickness is greater. On this point we 
have made numerous observations, which have led us to the following conclusion :— 
Propositio7i IX.—The specific conductivity of a soap film increases as the thickness 
decreases, hut this mcrease is less in the case of a film formed of a solution to 
tvhich a salt has heen added, a7id cannot he detected in the case of a “ liquide 
glycerique ” such as that ive have emq)loyed. 
To justify this statement we adduce the following facts which are only samples of 
the results of a large number of experiments. Perhaps the most convenient way of 
testing the constancy'of the specific conductivity of a film is to calculate the thickness 
from observations on the electidcal resistance, assuming that the specific resistance is 
the same as that of the liquid in mass, and to compare the number so obtained with 
the thickness given by a direct optical measurement based on the colour. If the 
ratio of the electrical thickness to the optical thickness is unity there is no reason to 
suspect a change in the specific conductivity. If it differs much from unity the 
presumption is in favour of such a change. One film, formed of a solution containing 
one part of oleate of soda to GO of water without the addition of any metallic salt, 
which was observed on July 7, 1892, lasted two hours and a half. During this 
