1^ OJf THE THEORY OF CAPILLARY ATTRACTION. 



stiHbe supported. The figure of the surface is a secondury 

 effect, exactly in the same manner as the catenarian form of 

 a siipported chain is a secondary ^Sect ; the cause being the 

 pegs and the force of gravity. 



When, however, I say, that the figure of the surface is 

 <jf no coTisecjuence, I suppose that surface to be at a senbible 

 distance be^ow the orifice of the tube, otherwise the case wil! 

 be very diiferent : for. 

 Corollary 2. Cor. 2d, Suppose, in fig. 5, every thing to remain as 

 before, excepting that the whole tube is made of one kind 

 ©f matter^ {the intensity of attraction for the fluid within it 

 being r) and that the left hand branch terminates at y 5, close 

 to the surface of the fluid. I say that the difference of 

 masses, in ih.& two branches, will, in great measure, depend 

 «n the figure of that part of the surface of thcfltdd, at the 

 orifice 7 5, which is near £0 y and S : and that we should 

 have the greatest difference of masses if it were possible to 

 make the fluid, at this orifice, take any figure as S j:2 <y *, 

 perpendicular at the sides, and having every point of its 

 upper surface distant from y and ^ by a space greater than 

 the radius of the sphere of action of the particles. And the 

 diiference of masses in this case, tvould be to the difference 

 of masses when the surface is horizontal, as y e 5, at the 

 orifice of the tube, as 2 to 1, 



For, in this latter case, when the surface of the fluid is 

 y t \ it is evident, from the reasoning used in the proposi- 

 tion, that the left hand mass is drawn upwards by a force as 

 r, and the right hand mass by a force as 2 r ; whence the 

 difference of the masses is as r. But, if the left hand mass 

 couJd stand at x z, and the column gp become yf, any such 

 column, and consequently the whole left band mass, would 

 be urged neither upwards nor downwards by the branch y 5 

 of the tube: therefore the difference of masses would be 

 entirely occasioned by the other branch, and would be 

 as 2 r. 



Now, tl>6ugh we cannot make the fluid stand at x z, we 



♦ Mr. Haiiy, explaining Abat's experiments after the ideas of la Place^ 

 falls, in consequence, into an erroiir. See * Traite de Physique,' torn. 1, 

 p. 243, Ed. 2. * le petit chaiigemeut de figure lui donne /-to rfe/orc e,' 



may 



