I 
j 
I 
THICKNESS AND ELECTRICAL RESISTANCE OF THIN LIQUID FILMS. 515 
the temperature was 22°'2 C. The same liquid on the same day gave = 103 /r./x.. 
at 18°, and, in the case of another film, /3 = 147 /x./x, at 18°. In this case the 
variations of /3, when the temperature was constant, were greater than the change 
which might otherwise have been attributed to a rise of 4° in temperature. Some¬ 
times the larger value was found at the lower temperature. Thus, in one case, when 
^ = 26°-3 y8 135 y./x. 
t = 21°-0 ^8 = 237 /x./x. 
Sometimes precisely the reverse result was obtained, as 
i = 21°'l ;S= 201 /x./x. 
t 18°'7 /? = 171 /x./x. 
Instances such as these might be multiplied indefinitely, and suffice to show that the 
relatively small variations of temperature which occurred during our experiments are 
not the cause of the fact that a film, of which the true thickness is about 22 /x./x., 
appears when measured electrically to have a thickness of 184 /x./x. 
(2.) The precautions taken to prevent absorption or evaporation by the films were 
precisely the same as those which our previously published results have shown to be 
efficient in the case of liquide glyceriqiie, which is much more sensitive to changes of 
hygrometric state than solutions containing no glycerine. 
The bottom of the vessel in which the films were formed was flooded with the 
liquid. A continuous roll of linen dipped into the liquid, and could be moved so that 
all portions in turn were moistened, thus exposing a large evaporating surface. A 
hair hygrometer was enclosed in the vessel. No observations were made until the 
index of the hygrometer ceased to move. A constant hygrometric state was there¬ 
after maintained. In the case of liquids not containing glycerine, we found that the 
layer of liquid at the bottom of the vessel was sufficient for this purpose, as the results 
were the same, whether the roll of linen was or was not used. The vessel in which the 
films were made was surrounded by a tank of water, the object of which was to check 
variations of temperature. To assure ourselves that the measurements were not 
affected by the diffusion of water vapour, into the interior we made experiments when 
the tank was empty as well as when it was full, and obtained precisely the same 
results in the two cases. Hence, considerable variations in the hygrometric state of 
the air close to the vessel, and in the means taken to secure a constant hygrometric 
state within the vessel produced no effect on the results. 
(3.) As, in order to assist in reducing the thickness of a film, we frequently use 
electric currents which, although they are very weak, pass for some time through the 
film and are relatively strong when it is thick, we thought it very important to prove 
that the change in specific conductivity is not due to the passage of the current. 
Comparative observations were therefore made, in some of which a cuirent from a 
3 u 2 
