512 
PROFESSORS A. W. REINOLD AND A. W. RUCKER ON THE 
films obtained both in 1883 and in 1890-92, when the solution contained 3 per cent, 
of KNOg, does not hold in the case of unsalted solutions. We have already stated 
that we believe the optically measured thicknesses to be correct mean values in all 
cases, but, whether this is so or not, the electrically measured values cannot for the 
reasons just given be real thicknesses. There must, therefore, when no metallic salt 
is present, be a difference between the specific conductivities of a film and the liquid 
of which it is formed. 
This conclusion is borne out by the results of experiments made with soft soap. 
The following table contains the results of measurements on 23 films made in the 
summer of 1889. 
Table IX.—Soft soap. No salt dissolved. 
Thickness in /<./*. as given by the electrical method. 
169, 
166, 
185, 
171, 
189, 
173, 
120 , 
126, 
1-50, 
151, 
150, 
145, 
142, 
143, 
148, 
150, 
188, 
80, 
157, 
165, 
179, 
252, 
232, 
Mean 162. 
Proposition V.—The apparent thickness, as measured hy the dectriccd method, of 
different black films formed of an unsalted solution is very variable. In some 
cases, at all events, this is due to a real variation in thickness. 
The results given above are sufficient proof of the apparent variability of the 
thickness of the film. 
The best proof that this variability is sometimes real, is the fact that we have 
distinctly seen two different shades of black in the same film. Newton recognised 
two different intensities of black in soap bubbles, and Lord Kelvin has on more than 
one occasion asked us whether we have noticed the same phenomenon. So long as 
our experiments were made with a solution containing a relatively large proportion of 
salt our answer was always in the negative. Since we have been working with 
solutions which contain but little salt, we have occasionally noticed traces of two 
different black tints in films containing not more than one per cent, of KNOg. In 
films containing no salt the phenomenon is of frequent occurrence. The difference of 
tint is but slight and may easily be overlooked, but in many cases a definite line of 
demarcation can be observed between the two tints, so that we can certainly state 
that the upper part of the film showed one tint and the lower part the other, which 
may be distinguished as and respectively. The line of separation is very 
irregular in form. That which divides the black and coloured parts of the cylindrical 
