566 Mr. S. A. Shorter on Surface Separation from 



A simple calculation showed that these numbers were approxi- 

 mately multiples of 47. Horizontal lines were drawn to cor- 

 respond to multiples of several values in the neighbourhood 

 of 47, and finally 46'8 was chosen, as giving the best fit. 



We can, therefore, state the following important con- 

 clusion : — 



(4) The stationary values are all multiples of a single 



unit value, viz. 46*8 dynes per cm. 

 A further inspection of the results shows that 



(5) The sudden fluctuations which occur are always 



between consecutive stationary values. 



Pliysical Interpretation of the Results. 

 These results may be readily understood if we make the 

 folio wing two assumptions: — 



(1) that the stationary values correspond to uniform 



layers a definite number of molecules thick, e. g. 

 the value 702 corresponds to a surface pellicle 

 15 molecules thick. 



(2) that a partially formed molecular Inyer does not 



contribute appreciably to the surface-elasticity till 

 nearly complete. 

 Let us examine the consequences of these assumptions. A 

 continuous process of surface-separation would give rise to a 

 stepwise mode of increase of the surface-elasticity, such as, 

 apart from certain irregularities, is observed in the diagram. 

 Suppose, however, that owing to some cause molecules are 

 sometimes reabsorbed into the portion of solution adjacent to 

 the surface-layer, and at other times separate out again in 

 the surface-layer, so that comparatively rapid fluctuations in 

 the number of molecules in the surface-layer are super- 

 imposed on the slow general increase. It is easy to see that 

 a small change of this kind will have very different effects 

 at different times. If the latest added molecular layer is 

 nearly complete, so that the surface-elasticity is practically 

 the same as if the layer were complete, a small addition to the 

 surface -layer will have no appreciable effect, while a decrease 

 in the number of molecules in the surface-layer will have a 

 very large effect on the surface-elasticity. If, however, a 

 molecular layer is in the initial stages of formation, an increase 

 will have a large effect and a decrease very little effect. 

 Hence we see that a small change in the number of molecules 

 in the surface -layer may sometimes cause a large change in the 

 value of the surface-elasticity , but this change will always occur 

 betiveen two molecular values and never across such a value. 

 It will be seen that the rapid fluctuations which occur always 

 conform to this rule. , 



