ON COHESION. 475 



direction : thus, two pieces of glass require to be brought together with 

 considerable force, and generally with some friction, before they can begin 

 to cohere ; and a small drop of water, falling lightly on the surface of a 

 pond, may remain for some instants without coming into perfect contact 

 with it ; the same circumstance is also still more observable in spirit of 

 wine a little warmed. 



The first and simplest effect of cohesion is to produce liquidity. That 

 all liquids possess some cohesion, is very obvious, from their tendency to 

 assume a spherical form when they are sufficiently detached from other 

 substances, and from the suspension of a drop from any solid, to which its 

 upper surface adheres with sufficient force. Without cohesion, indeed, a 

 liquid would be only a very fine powder, except that the particles of powders 

 have not the power of moving with perfect freedom on each other, which 

 constitutes fluidity. The apparent weakness of the cohesion of liquids is 

 entirely owing to this mobility, since their form may be changed in any 

 degree without considerably increasing the distances of their particles, and 

 it is only under particular circumstances that the effects of their cohesion 

 can become sensible. 



When a liquid is considered as unlimited in its extent, the repulsion of 

 its particles, situated in all possible directions with regard to each other, 

 may be supposed in all cases precisely to balance the cohesion, which is 

 derived from the actions of particles similarly situated ; and this must also 

 be the state of the internal parts of every detached portion of a liquid, 

 where they are so remote from the surface as to be beyond the minute 

 distance which is the limit of the action of these forces. But the external 

 parts of the drop will not remain in the same kind of equilibrium : they 

 may be considered as a thin coating of a liquid surrounding a substance 

 which resists only in a direction perpendicular to its surface, and does not 

 interfere with the mutual actions of the particles of the liquid. Now since 

 the repulsive force increases as the distance diminishes, it must be exerted 

 more powerfully by the nearest particles, while the cohesion is directed 

 equally towards all the particles within a certain distance, and wherever 

 the surface is curved, the joint cohesive force will be directed to a remoter 

 part of the curve than the repulsive force opposed to it, so that each 

 particle will be urged, by the combination of these forces, towards the 

 concave side of the curve, and the more as the curvature is greater ; 

 hence the coating of the liquid, thus constituted, must exert a force on the 

 parts in contact with it, precisely similar to that of a flexible surface, 

 which is every where stretched by an equal force ; and from this simple 

 principle we may derive all the effects produced by a cohesion of this kind, 

 which, from its being most commonly observed in the ascent of water in 

 capillary tubes, has been denominated capillary attraction. (Plate XXXIX. 

 Fig. 531.) 



It is, therefore, a general law, that the surface of every detached portion 

 of a fluid must every where have such a curvature, as to be able to with- 

 stand the hydrostatical pressure which acts against it ; and hence we may 

 calculate in many cases the properties of the curve which it must form ; 

 but in other cases the exact calculation becomes extremely intricate, and 



