Joty—The Inner Mechanism of Sedimentation. O27 
attending the passage of water through capillary tubes, is ordinarily 
productive of mutual repulsion between the particles, then the 
free charge on the ion may serve to discharge or so reduce this, 
that conjunction of the particles becomes possible. Once this 
occurs, the tendency to minimum surface energy will effect the 
retention of the particles in close union. Hardy’s recent experi- 
ments‘ confirm the idea that some such phenomena actually occur ; 
for he finds that the ions which effect the discharge are those of 
opposite sign to that revealed in electric endosmose by the particles 
of the colloid. ‘Thus a colloidal particle moving with the current 
may be assumed to be electro-positive. Now such particles are 
coagulated by the electro-negative ions: those moving against the 
current by the electro-positive ions. This is established by the 
fact of the valency law, working out, in the first case, for the valency 
of the acid radicle in solution ; in the second, for the valency of the 
metal in solution. A further and very beautiful confirmation is 
contained in Linder and Picton’s observation that the metallic ion 
may be carried down with the colloid. I find that the view that 
the action consists in each case principally of the neutralisation of 
the inner electric layer harmonises well with the experiments of 
Hardy and others. 
We have so far reviewed the ivory as it stands for the case 
of the very minute colloidal particles. The phenomena of the 
deposition of sediments involve particles rising to dimensions 
much greater than those which are considered in the foregoing 
theory. But, as will be seen, these larger particles, even of the 
most insoluble substances, reveal precipitating effects due to salts 
in solutions which are undoubtedly connected with the valency of 
the ions and also with the electric sign of these. It therefore 
becomes of interest to seek in the first place for some clue as to 
how the valency may affect the rate of precipitation of large 
particles. By large particles I mean particles which are large 
compared with the ionic grain of the liquid. 
In this case, if U be the component of the mean ionic diffusive 
: : : : Ae N 
velocity perpendicular to unit area in the liquid, and 3 be the 
number of ions of the proper sign having a component directed 
1 Proc. Roy. Soc., vol. 66, p. 110. 
