86 



SCIENCE 



[N. S. Vol. XLV. No. 1152 



can scarcely be conceived without assTiming 

 a force of repulsion instead of attraction be- 

 tween molecules of the liquid. 



But he surely can not mean to question the 

 existence of aaegative surface tension at a sur- 

 face between a liquid and solids, for how other- 

 wise are we to explain the most familiar facts 

 in capillarity. Is it not negative surface ten- 

 sion which causes the water to rise in a capil- 

 lary tube, or against a glass wall, and causes 

 a drop of oil to expand indefinitely over a glass 

 plate? Is it not the greater negative surface 

 tension in the oil-glass surface which causes 

 the film to expand against the contractile force, 

 or positive surface tension of the oil-air sur- 

 face? 



'Not does it appear to be necessary to sup- 

 pose a repulsive force between molecules of the 

 liquid in order to account for the existence of 

 such a negative tension, for if the resultant 

 force of attraction on a particle of liquid near 

 the surface, due to all particles on both sides 

 of the surface lying within the range of sen- 

 sible molecular attraction, is. directed away 

 from the surface and towards the interior of 

 the liquid, the particle will tend toward the 

 interior and we shall have positive surface 

 tension but if the resultant attraction is tow- 

 ard the surface there will be negative sur- 

 face tension. 



In case of an air-liquid surface the attrac- 

 tion of neighboring liquid particles upon a 

 particle in the surface is so much greater than 

 any opposing outward attraction by adjoining 

 air molecules that the first condition holds 

 and the surface tension is positive. While 

 at a glass-oil surface a particle of liquid near 

 the surface may be supposed to be more 

 strongly attracted by the neighboring glass 

 molecules than by the oil molecules in its 

 vicinity, in which case the resultant attraction 

 is toward the glass, the potential energy of a 

 liquid particle is less at the surface than in the 

 interior of the liquid, and the surface tension 

 is negative. 



When liquid comes against liquid the case 

 is complicated by the mobility of particles on 

 both sides of the boundary. It seems probable, 

 however, taking an oil-water surface as an ex- 



ample, that if the resultant attraction on an 

 oil molecule at the surface is directed across 

 the boundary from the oil side toward the 

 water, that a water molecule at the surface 

 being in the same situation with respect to the 

 surroimding molecules will be urged in the 

 same direction. In other words, we can hardly 

 imagine a particle of one sort in the surface as 

 being drawn in one direction by the attrac- 

 tion of all the surrounding particles on both 

 sides of the surface, while a similarly situated 

 particle of the other sort would be drawn in 

 the opposite direction. 



We may assume then that at a surface be- 

 tween two liquids, particles on one side are 

 urged away from the surface, while those on 

 the other side are urged toward it. That is, 

 there are two influences, one tending to con- 

 tract the surface and the other to expand it. 

 If the first is predominant there is positive 

 surface tension, this is the ordinary case where 

 diffusion does not take place, as with water-oil 

 or water-mercury. 



If the second is predominant the surface 

 tends to expand indefinitely, and the limit 

 would seem to be reached only when one 

 liquid is uniformly diffused throughout the 

 other. In this case diffusion is to be expected 

 also from the consideration that if particles 

 in the one liquid are drawn so powerfully tow- 

 ards the other as to force the expansion of the 

 second liquid in opposition to its contractile 

 tendency, it seems probable that they will be 

 drawn actually into the second liquid and thus 

 the integrity of the surface be destroyed. We 

 conclude, therefore, that a positive surface ten- 

 sion is to he expected between all liquids that 

 do not interdijfuse. 



If the particles in a colloid solution are to 

 be regarded as solid, we may expect to find 

 cases where the surface tension is positive 

 and other cases where it is negative. Where 

 it is positive there will be a tendency to floccu- 

 late, for as two colloid particles come together 

 liquid particles move out from between them 

 into the interior of the liquid, and the capillary 

 region surrounding the particles is thus de- 

 creased in volume, and the potential energy of 

 the system is diminished. When, on the other 



