﻿650 Distribution of Particles in Colloidal Suspensions. 



These last experiments show that care should be taken to 

 avoid thinking of the limiting concentration as necessarily 

 giving a saturation value. The same kind of curve is 

 obtained when more dilute suspensions are investigated. 

 The value of K is also different for the two curves. For 

 curve 1 (fig. 3) it is about 29, while for curve 2 it is 55. 

 It would seem that the size of the solid nucleus diminishes 

 to one-half when the suspension is reduced to half strength. 



Fig. 3. 



4: q 



— i¥== ^ a 





%^ 



7* 



-H^ ' 





J—i^ 



£^ 





y-> 



Although Burton's particular assumption of the existence 

 of an electrical repulsion between the particles has been here 

 criticised, it might still seem an open question as to whether 

 some other force between them might not be the operative 

 cause. The chief difficulty applicable to all possible laws of 

 force is that such forces will be exerted in all directions and 

 must act outwards at every surface. Hence, though near 

 the top of the fluid they might be assumed of such a charac- 

 ter as to oppose gravitation, at the bottom they should 

 assist it in producing a change of concentration, which might 

 be expected to be exceedingly large in consequence. Again , 

 at the sides where the forces would act laterally they would 

 produce a change of concentration independent of gravi- 

 tation. What really happens can be seen from fig. 4, 

 which represents the approximate relative values of the 

 numerical concentrations actually observed at various points 

 of a cell \ cm. deep and \ cm. broad. It is obvious from the 

 numbers given in the figure that the only variation of 

 concentration is near the top. It ought to be mentioned that, 

 since the suspension was covered with a microscope cover- 

 glass, the suspension was surrounded by glass on all sides ; 

 the peculiar effect at the top is therefore not due to a 

 transition from liquid to air. 



It would appear from these preliminary determinations 

 that there are in reality three regions in a suspension to- 



