ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 17 
as—at least theoretically—a further increase in the relative volume 
of disperse phase can lead to flattening at the points of contact of 
contiguous spheres and eventually to a polyhedral structure of the 
system. A limit is only reached when the thin films of continuous 
phase which separate the polyhedra of disperse phase are no longer 
capable of cohering. The practical possibility of such systems is 
demonstrated by some of 8. U. Pickering’s emulsions with up to 
99 per cent. of disperse phase. The complete conditions for the 
stability of such films will be discussed further on. but it can be 
said at once that a low inter-facial tension is, if possibly not sufficient, 
certainly necessary. 
Conversely, emulsions with a very low content of disperse phase, 
in which the particles are small and separated by layers of continuous 
phase of some thickness compared with the extent of interface, are 
possible and moderately stable even when the interfacial tension is 
high, viz. of the order of 30 to 40 dyne/em. Such emulsions may be 
produced by agitation, by distilling the two phases together, or by 
‘precipitating’ an alcohol or acetone solution of the disperse phase with 
a large excess of water. The properties of such emulsions with a 
disperse phase consisting of mineral oils, aniline, nitrobenzene, oleic 
acid or castor oil, amounting to one part in a thousand, or less, ef the 
total volume, are throughout those of fine suspensions or of suspen- 
soids. The (negative) electric charge is of the same order as on 
‘suspensoid particles (Lewis) ; is similarly affected by hydrogen and 
hydroxyl ions (Ellis) ; the maximum concentrations at which the 
‘emulsions are stable are of the same order as for suspensoid sols 
(Lewis), and the phases may be separated by filtration through 
suitable septa under considerable pressure (Hatschek). The effect 
of interfacial tension is altogether subordinate to that of the electric 
factors. 
Emulsions containing larger percentages of disperse phase than 
those mentioned, in particles of microscopic or approaching micro- 
scopic size, are stable, or in fact anything more than transient, only 
when the interfacial tension between the phases is low, as has already 
been mentioned. In the emulsions occurring in nature, such as 
milk or latex, this is generally brought about by the protein content 
‘of the continuous phase, while in the emulsions prepared artificially 
the agent which lowers the interfacial tension is very generally a 
soap. The stability of such emulsions is, however, again dependent 
on the phase ratio and is generally not complete unless this approaches 
the figure corresponding to closest packing. Thus, if oil and a dilute 
soap solution are shaken until the whole of the oil is dispersed, the 
resulting emulsion separates into a layer of soap solution containing 
only a very small fraction of oil and a ‘cream’ containing 75 per 
cent., or more, of oil; the latter is stable, provided the walls of the 
containing vessel are wetted only by the continuous phase and no 
oil in bulk is in contact with the cream. The rate of separation 
depends naturally on the difference in density of the phases, the 
‘viscosity of the continuous phase and the degree of dispersion : it is 
obvious, although the case has chiefly a theoretical interest, that a 
‘system stable in all ratios is conceivable if both phases have the 
same density, e.g. castor oil in water with a little alcohol, 
