452 Prof. J. Plateau on the Figures of Equilibrium 



was in no way disturbed by the movement of the needle, which 

 simply came and struck against it. It follows from this that in 

 liquids of the second category the superficial layer has not any 

 greater viscosity than the interior; but I have shown that in reality 

 it has less. I will confine myself here to citing a single fact 

 bearing on this point. If the experiment of a small Boating 

 body is made with a mixture of equal volumes of water and 

 alcohol, the body is simply struck by the needle ; thus the ex- 

 cess of superficial viscosity possessed by the water is completely 

 destroyed by the presence of the alcohol. It therefore follows 

 that the superficial layer of the latter must be less viscous than 

 the interior, or, if I may so express myself, that it possesses a 

 negative excess of viscosity which neutralizes the positive excess 

 belonging to the water. 



Lastly, five liquids of the third category were tried, namely, 

 solutions of Marseilles soap, soft household soap, resin soap, 

 saponine, and albumen, and showed, like those of the first ca- 

 tegory, a superficial viscosity much greater than the interior 

 viscosity. One of them (solution of saponine) yielded in this 

 respect extraordinary results ; its superficial viscosity is ex- 

 tremely strong : the needle placed at 90° from the magnetic 

 meridian and then left free remains in this position, as if the liquid 

 were covered with a solid pellicle; but yet it is impossible to 

 detect by any means the presence of such a pellicle. Solution of 

 albumen shows a similar behaviour, but in a less degree. 



Thus the results obtained by means of the magnetic needle in 

 regard to the fifteen liquids that I have submitted to this kind 

 of trial, fully confirm the consequences drawn from the experi- 

 ments on the hemispherical films ; we may therefore, I think, 

 look upon the following principle as fully established : — 



The superficial layer of liquids has a proper viscosity, indepen- 

 dent of the viscosity of the interior of the mass. In some liquids 

 this superficial viscosity is greater than the internal viscosity, and . 

 often much greater, as in water and, especially, in solution of sapo- 

 nine ; in other liquids, on the contrary, it is less than the internal 

 viscosity, and often much less, as in oil of turpentine, alcohol, §c. 



The idea of a viscosity proper to the superficial layer of liquids 

 had already been put forward by M. Hagen; but he seems to 

 consider that this viscosity is greater in all liquids than the 

 internal viscosity. 



In order to be able to form a definite estimate of the relations 

 between superficial viscosity and tension, we should require to 

 have some accurate means of determining the numerical values 

 of the first of these elements, in the same way as those of the 

 second are determined. I have tried without success to find an 

 accurate method for this purpose ; but I have shown that, in the 



