ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 53 
sulphate are quite different from those of the crystalline substance 
although the chemical constitution is considered to be the same in 
the two cases. The suggestion has been made? that colloidal barium 
sulphate consists of extremely minute crystals, but this does not 
explain the difference in physical state. The chemist’s idea can only 
be upheld by a considerable modification in the conception of 
molecules as present in solid substances. The Braggs” have estab- 
lished beyond doubt that the fundamental units of a crystal are 
atoms, not molecules. Many chemists*—* have cast doubt on the 
crystal models of the Braggs’, except that of the diamond, simply 
because they are against the generally accepted ideas on valency. 
It has been suggested® that one of the interesting problems of the 
future will be to reconcile the X-ray crystal models with the 
molecular hypothesis. Since the X-ray models have been deduced 
‘from positive evidence whereas no positive evidence has ever 
=_—=_ - 
obtained in favour of the molecular hypothesis as applied to solids, 
the author considers that it is the molecular hypothesis which 
requires broadening. ‘The original definition of a moiecule was the 
smallest particle of a substance which could exist; with the dis- 
covery of chemical polymerisation this definition has been narrowed 
somewhat. The author considers that the difficulty of reconciling 
the X-ray models with the molecular hypothesis disappears if one 
admits a crystal to be a polymeride or single molecule as suggested 
by Langmuir. This idea has far-reaching consequences of very 
great importance in colloid chemistry, since the necessity for 
admitting a high molecular weight as a special characteristic of 
colloids disappears. A single crystal may be considered as a single 
molecule since it consists of atoms bound together by the same 
forces which bind together the atoms in the simple moiecules found 
in gases. The molecular weight of the crystal is then proportional 
to its size. This view is entirely in agreement with the observa- 
tions in brownian movement of particles in different degrees of 
dispersion which were shown by Perrin!’ to be explained by the 
kinetic theory of gases. There are numerous other consequences of 
the idea of considering crystals as polymerides which it is not 
intended to discuss here. 
Regarding the chemical nature of fibres, chemists have divided 
opinions, some considering them to be very complex bodies of high 
molecular weight, others to be physical forms of insoluble substances 
of possibly low molecular weight. These opinions may be taken to 
be identical in the light of the idea discussed above. 
In the case of vegetable fibres, the suggestion has recently been 
made! that the fibre substances, celluloses, are liquid systems ; this 
brings forward the question of reconsidering the definitions of 
liquid and solids. At present, these are somewhat rigid and bodies 
possessing properties of both states of matter are considered to be 
mixtures. The author does not hold this view, but considers that 
there is no distinct line of demarcation between the solid and liquid 
states any more than there is between colloidal—and true—solution. 
The semi-solid state of many colloids does not necessarily imply the 
presence of both solid and liquid phases. It is, therefore, still an 
open question whether the celluloses are mono--or diphasic systems. 
