April 21, 192 1] 



NATURE 



227 



is a mere confession of weakness, and every 

 step of the argument on which it is based seems 

 open to challenge. It is pure topography, and 

 as such of little value. The colloidal state does, 

 in fact, touch solution on one hand, and sus- 

 pensions on the other, but it is not a matter 

 simply, or even primarily, of scale. The 

 distinctive quality of the state consists in 

 certain constraints which may fairly be called 

 frictional constraints, from which comes the 

 ■characteristic inertia of colloidal systems noticed 

 by Graham. An ideal suspension in which the 

 relation between particles and fluid medium was 

 one of simple repulsion would be free from such 

 •constraints. Its sole characteristic would be a 

 uniform distribution of the particles — this follows 

 from considerations qf entropy- — so that, if 

 appropriate external restraints operated, the 

 system would manifest an osmotic pressure. 



An ideal suspension of this kind is the ideal gas 

 of colloids, and the distinction between it and the 

 simplest colloidal solution lies in the fact that the 

 particles react with the fluid, the energy asso- 

 •ciated with the reaction being of the type known 

 as surface energy, but modified by the excessive 

 curvature of the surfaces. Each particle acts as 

 a strain centre, the molecules about it being 

 orientated more or less with respect to its centre, 

 and the total effect is an increase in the rigidity 

 and a decrease in the mobility of the fluid — a 

 ■decrease that is in the number of molecules which 

 cross unit area of a plane surface in the interior 

 in unit time. Any constraint which the particles 

 •exert on the molecules of the fluid will therefore 

 tend to increase their own diffusive energy, and 

 the osmotic pressure would be greater than that 

 of an ideal suspension, just as when true solution 

 is exothermic the osmotic pressure is greater than 

 that given by the gas equation. 



The energy peculiar to such systems may be 

 classified as capillary and electrical, namely, a 

 contact potential difference between the particles 

 and the medium. We are ignorant of the quantita- 

 tive relations between the two, but stability is 

 least when the contact potential difference van- 

 ishes — that is to say, at the isoelectric point. This 

 feature is almost always and quite wrongly de- 

 scribed by saying that coagulation occurs at the iso- 

 electric point. Coagulation, of course, occurs over 

 a range which is determined by the magnitude of 

 the forces operating to produce agglutination and 

 precipitation. 



It is obvious that two particles which come 

 within range of each other will or will not 

 agglutinate according as the variation of 

 surface energy with the distance between their 

 centres is positive or negative. If it be negative 

 NO. 2686, VOL. 107] 



there will be a buffer action similar to that which 

 may often be observed between drops of one fluid 

 floating on the surface of another. A finite amount 

 of work must be done to bring about agglutina- 

 tion, and this is an instance of one of the fric- 

 tional constraints characteristic of colloids. Third 

 components are practically always present in 

 minute amount in actual sols condensed on to the 

 particles. They decrease the chances of agglu- 

 tination because they decrease the energy of the 

 interface between particle and fluid, and, there- 

 fore, help to make the variation of the energy 

 with the distance between centres negative. 



We may note in passing that the diffusive 

 energy is concerned only with the distribution of 

 the particles. The size of the particles — that is 

 to say, whether they do or do not agglutinate or 

 completely fuse on "contact" — is determined by 

 the variation of energy mentioned above. A strik- 

 ing example is offered by the system ether-water. 

 If the ether phase be distributed through the 

 water by shaking, the drops are brought into con- 

 tact again by the external aggregating force 

 gravity, and, once in contact, they immediately 

 fuse. If, however, a trace of iodine be added, 

 gravity brings the drops together; but they do 

 not fuse, because of the local influences of the 

 iodine upon the local variation of energy on 

 "contact." 



Having got so far, it does not need much 

 imagination to see that the reason why colloidal 

 particles do not fuse must be essentially the same 

 as the reason why solid faces do not weld when 

 pressed together. 



There would be little difficulty in defining the 

 colloidal state if the relations between the com- 

 ponents were only those mentioned above. It is 

 at the other end of the scale where sols shade into 

 true solutions in a perplexing way, not because 

 of variation in the size of the particles, but because 

 true solution exists side by side with true col- 

 loidal dispersion. 



Broadly, there are twa types to consider : those 

 in which true solution involves, or seems to 

 involve, the entire colloidal component — e.g. 

 silica — and some proteins in water ; and those 

 in which the solute is a salt, one ion of 

 which is highly insoluble, in which case the 

 dispersed phase consists of aggregates of 

 this ion with unionised molecules. Such 

 systems are salts of proteins and of fatty acids in 

 water, and the remarkable feature is that though 

 the "colloidal" ion may grow to such a size as 

 almost to reach the limits of microscopic vision, 

 the electric charge it carries is the area of its 

 surface multiplied by a constant. 



To return now to the delimitation of the col- 



