on the Adhesion between Solids and Liquids. 193 



A still greater evil connected with this method is that one and 

 the same body has to contend in contact with different liquids ; 

 and is it not possible that the results will vary by varying the 

 form and material of the solid that is set in motion on the liquid 

 or within it ? The laws of capillary action would never have been 

 discovered if the phenomena had been studied only by means of 

 one narrow tube of one material ; so it appeared to me that the 

 extensive class of phenomena thus indicated by M. Plateau 

 could not be properly investigated unless a diligent examination 

 were made as to the behaviour of solids of varied shape and ma- 

 terial set in motion upon the surface of a liquid or below it. I 

 held it to be impossible that the varying adhesion between 

 solids and liquids would not produce some variation in the effects 

 now under consideration. 



In carrying out this idea, it first occurred to me to make the 

 axis on which the needle turned moveable, and to fix to it the 

 solid to be tried. But as the function of the needle in this ex- 

 periment is to restore to the position of equilibrium a body dis- 

 turbed therefrom, would not the torsion of a filament be prefer- 

 able to the magnetic force ? Coulomb's balance seemed to be 

 well adapted to this form of experiment. To the filament of this 

 balance is attached in a horizontal position the solid that is to be 

 lowered to the surface of the liquid or to be submerged therein. 

 The liquid is contained in a vessel placed beneath the cover of 

 Coulomb's apparatus. The thread at its lower extremity carries 

 a horizontal index playing within a graduated circle attached to 

 the lower part of the case ; while the aperture at the top is closed 

 with a screw furnished with a finger moving through a small 

 space, by means of which it can arrest the index or leave it free 

 in its course. 



The solid being put into the desired position with respect to 

 the liquid, the filament is equilibrated so that the index may be 

 opposite the arresting finger. This is best done by turning the 

 upper circle of the cover. The index then stops; the filament 

 is twisted a certain number of degrees either to the right or to 

 the left; the arrest is removed, the index turns; but its first 

 movement is a little uncertain, as is that of Plateau's needle. I 

 wait until it has reached 5° from its position of equilibrium,, 

 and then, setting out from some known point, I count the num- 

 ber of degrees through which the index moves in a given time, 

 such as 10, 15, 20, 60 or more beats of a common watch. 



Experiments were made in this way in pairs. A first expe- 

 riment, made by twisting the filament to the right, was soon suc- 

 ceeded by a second experiment, in which the filament was turned 

 to the left. If in both cases the index stop exactly at the position 

 of equilibrium of the filament, the results are identical. If 



