2 ——— 
ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 19 
two phases, in given conditions, will be the disperse and which the 
continuous phase. 
These views derive strong confirmation from a series of experi- 
ments by G. H. A. Clowes. He prepares an emulsion with oil as 
disperse phase by shaking olive oil with a dilute solution of NaOH 
in water. If a quantity of CaCl, slightly in excess of the equivalent 
of the NaOH is added to this emulsion, it transforms itself sponta- 
neously into one having water as disperse and oil as continuous 
phase. This can be re-converted into the original emulsion by 
adding NaOH until the original OH’ concentration is restored. 
Magnesium, iron, and aluminium have an effect similar to that of 
calcium. Since the oleates of these metals are much more soluble in 
oil than in water, their formation would cause a considerable reduc- 
tion in the interfacial tension between the film and the oil phase, so 
that the film woulc tend to become convex towards the oil and 
concave towards the aqueous phase, 7.e. oil would become the con- 
tinuous phase and water the disperse. The whole process is an 
actual reversal and is therefore sharply distinguished from such 
phenomena as the separation of emulsions into two (no longer 
disperse) phase by the addition of electrolytes or of substances which 
decompose the material of the interfacial film. 
Evidence similar, and complementary, to that afforded by Clowes’s 
experiments is provided by an investigation by A. U. M. Schlaepfer, 
the object of which was to produce emulsions of water dispersed in 
oil by using a finely divided substance more completely wetted by 
oil than by water : soot accomplished the desired result. 
Earlier experiments by Walter Ostwald, made with the intention 
of determining the type of emulsion which would result from the 
agitation of oil with water only, showed that either type was possible 
and that the result depended largely upon secondary factors, such as 
the state of the containing vessel and its previous wetting by one or 
the other phase. Conclusions drawn by this author regarding the 
limiting phase ratio have been proved incorrect both theoretically and 
experimentally. Donnan also found that the state of the vessel intro- 
duced considerable experimental complications. Although the method 
used by these and other authors—viz. agitation of the total volumes of 
both phases—is perhaps as good as any other arbitrary method, it 
neither corresponds to the probable process of formation of natural 
emulsions nor to the technical methods adopted for the production of 
stable emulsions, in which the gradual addition of the disperse phase 
during, and pari »assu with, the process of dispersion is generally 
considered an essential condition of success. 
The theoretical interest of emulsions is considerable and is likely 
to become even greater. Viewed as disperse systems of two liquid 
phases they are the only ones in which the phase ratio is accurately 
known and therefore specially adapted for studying the physical 
properties, eg. the viscosity of such systems. They have also 
acquired very great importance in biology in view of some modern 
hypotheses on the constitution of protoplasm, the possible existence 
of lipoid films and cognate phenomena, such as the action of 
‘antagonistic’ ions and of anesthetics. While at present the con- 
clusions respecting such complicated problems drawn from the study 
