SION.
Adhesion may take place either be tween two solids, as two hemispheres of glass, which, according to an experiment of Desaguliers, adhere to each other with a force eqoal to 19ounces on a surface of contact one-tenth of an inch in diameter ; Or between solids and fluids, as the sus tension of water in capillary tubes; or, lastly, between two fluids, as oil and wa ter. About the same time Mr. Hauksbee proved, experimentally, the error which Bernoulli had fallen into, in attributing the adhesion of surfaces and capillary at traction to the pressure of the atmo sphere. Nevertheless, in 1772, M. M. La Grange and Cigna, takingfcir granted a natural repulsion between water and oily substances, imagined, if there was an ad hesion between water and oil, or tallow, that it must be occasioned by a cause dif ferent from attraction : and having ascer tained the reality of the adhesion, they concluded that it was occasioned by the pressure of the air, and that Dr. Taylor's method was not well founded.
Such was the state of opinions on the subject, when, in 1773, Guy-ton Morveau made his celebrated experiments on ad esion, in presence o f the Non Academy, demonstrating, as indeed Haukshee had done before him, not only that water as. cends between two parallel plates of tal low, separated from each other 1.8 of a line, but also that the atmospheric pres sure is not in the least degree the cause of the phenomenon, which is solely attribut able to attraction ; in proof of this, a po lished disk of glass, Su lines in diameter, was suspended to the arm bf a balance, and brought into contact with the surface of mercury: the counterpoise required to separate it was equivalent to 9 gros and a few grains, and upon moving the apparatus into the receiver of an air pump, and forming as perfect a vacuum as possible, precisely the same counter poise was required as before.
In the prosecution of his on this subject, he observed, that the same -disk of glass, which, when in contact with pure water, adhered to it with a force equal to 258 grains, required a counter poise of only 210, in order to separate it from a solution of potash, notwithstanding the superior density of this last. This
inequality of effects on equal diameters, and in an inverse order to that of the •respective specific gravities of the two fluids, appeared not only to be decisive in favour of Dr. Taylor's method, but to en courage the hope of applying it to the calculation of chemical affinities.
In order to verify this proposition, plates of the different metals in their highest state of purity were procured, perfectly round, an inch in diameter, of the same thickness, well polished, and furnished with a small ring in the centre of each, so as to keep them suspended precisely parallel to the plane of the hori zon. Each of these plates was in turn suspended to the arm of an assay balance, and exactly counterpoised by weights placed in the scale attached to the oppo site arm ; the plate, thus balanced, was applied to the surface of some mercury in a cup, about two lines beneath it, by sliding the plate over the mercury, as in the silvering of mirrors, so as to exclude every bubble of weights were then successively added, till the adhesion be tween the plate and the mercury was broken. Fresh mercury was used for each experiment. The following is the table of results Gold adheres to mercury with a force equal to . . . 446 grains. Silver . 429 Tin 418 Lead 397 Bismuth 172 Zinc 204 Copper 142 Antimony (regulus) . . 126 Iron 115 Cobalt 8 The striking differences in the above table show that the pressure of the at mosphere has no share in them, since in this respect the circumstances of each were precisely nor do they de pend on the respective specific gravities ; for if so, silver should rank after lead, cobalt before zinc, and iron before tin.— The only order which agrees with the above is that of the chemical affinity of these metals, or the respective degrees of their solubility in mercury. It is highly probable, therefore, that at least the prin cipal part of the adhesive force thus found by experiment is owing to chemical affi nity ; and that the above numerical series, 446, 429, 418, 397, &c. is an approxima tion towards the ratio of the relative af finities of gold, silver, tin, lead, &c. for mercury.