460 Proceedings of Royal Society of Edinburgh. [sess. 
In some experiments with cupric ferrocyanide also, very little 
decrease of conductivity was observed either with potassium 
chloride or potassium sulphate. It may be that this film is per- 
meable with respect to these ions ; cupric ferrocyanide might, 
nevertheless, give large effects with other ions. 
4. The tables presented in the paper contain the results of all 
experiments which were carried out to the end, and it will be 
noticed that in most cases the values obtained in any one series 
group themselves around two values, a set of high values and a 
set of low values, each set being fairly concordant. The high 
sets are distinguished in the tables by an asterisk, and have not 
been included in the calculation of the mean value. This 
peculiarity is particularly noticeable in the chromic hydroxide 
results; it is not confined to the measurements made in any 
particular cell, or with one particular solution. The only explana- 
tion which seems at all probable is that there is more than 
one modification of the colloid film ; that, as would be expected, 
they differ in their effect on the conductivity of the solutions, 
and that on some occasions the one modification, and on others 
the other modification, is obtained, the precise conditions which 
determine the formation of the one or the other being so far 
unrecognised. 
This explanation is supported by several observations which 
have been made ; e.g., in the tables for ammonium sulphate with 
aluminium hydroxide films, it is noticeable that with different 
concentrations of the ammonium sulphate and of the precipitants, 
different results are obtained, though each set shows none of 
the sub-grouping referred to. At the one concentration the 
one modification may be obtained; at the other concentration 
the other one. In some cases, also, an apparently final con- 
ductivity was obtained, which remained constant for a day or 
more, and then the conductivity began to decrease very slowly 
over a period of several days, and attained a much smaller, really 
constant value. A slow transformation of the film from a less 
stable to a more stable modification appears to afford the most 
probable explanation of this behaviour. In this connection it 
should be mentioned that the solutions of chromic sulphate and 
of chromic chloride were prepared by reduction of chromic acid 
