TRANSACTIONS OF SECTION A. 



329 



horizontal for a short distance, then turning upwards and passing on into the 

 form of curve usually depicted. 



The weight of salt adsorbed per gramme of silica at this critical point where 

 the curve becomes horizontal (i.e., at the point where the amount of the 

 adsorbed layer is independent of the concentration) is found in the case of the 

 chlorides to vary with the temperature, getting less as the temperature is 

 raised, but in the case of those sulphates and nitrates which have been examined 

 it is found to be independent of the temperature. Typical curves are shown in 

 figs. 1 and 2. 



By measuring the area of the surface of a gramme of the silica employed 

 the weight of salt adsoi'bed per sq. cm. is known; with this object 'quartz 

 wool ' is best to use. If the density of the material of the adsorbed layer is 



1% 



2% 



3% A% 



Fig. 2. 



b% 



6% 



7% 



known it is then possible to calculate its thickness ; thus it has been found, 

 by assuming the density of the salt in the adsorbed layer to be that in the 

 solid state, that the thickness of the critical layer in the case of the sulphates 

 and nitrates is of the order 10 ix (10 -6 cm.), while in the case of the chlorides 

 the thickness of the critical layer approximates to a limit of the same order as 

 the temperature is lowered. 



In some cases, particularly at higher temperatures, the adsorbtion curve, 

 after becoming horizontal, turns actually downwards before passing on into 

 its final form. So that at a certain strength of solution, increasing the concen- 

 tration causes the silica to give up some of its adsorbed salt ; that is to say, 

 the layer at this point has become unstable for stronger concentrations. 



This instability will recall the observations by Lord Rayleigh on the thick- 

 ness of oil films on water, which showed that certain thicknesses of film were 

 never formed. The experiments of Keinold and Eiicker on soap films also show 

 that certain thicknesses of film were unstable. 



At the high temperatures in the case of the chlorides the curve shows that 

 the adsorption is zero for quite considerable values of the concentrations. The 

 form of the curve as it approached the concentration axis would look as if the 

 curve crossed to the negative side of the axis and would suggest thus, that for 

 small concentrations the force is repulsive, so that the pure water is adsorbed 

 rather tlwn the salt. 1 



1 This has since been verified experimentally. 



