158 MEMORIAL OF JOSEPH HENRY. 



plunged into it, the other end resting on the shelf.' In this con- 

 dition it was found, after a few days, that the mercury had passed 

 through the solid lead, as if it were a siphon, and was lying on the 

 shelf still in a liquid condition. The saucer contained a series of 

 minute crystals of an amalgam of lead and mercury. A similar 

 result was produced when a piece of the same lead wire was coated 

 with varnish, the mercury being transmitted without disturbing the 

 outer surface. 



When a length of wire of five feet was supported vertically, with 

 its lower end immersed in a vessel of mercury, the liquid metal 

 was found to ascend, in the course of a few days, to a height of 

 three feet. These results led me to think that the same property 

 might be possessed by other metals in relation to each other. The 

 first attempt to verify this conjecture was made by placing a small 

 globule of gold on a plate of sheet-iron and submitting it to the 

 heat of an assaying furnace ; but the experiment was unsuccessful, 

 for although the gold was heated much beyond its melting point, it 

 showed no signs of sinking into the pores of the iron. The idea 

 afterward suggested itself that a different result would have been 

 obtained had the two metals been made to adhere to each other, so 

 that no oxide could form between the two surfaces. To verify 

 this a piece of copper, thickly plated with silver, was heated to near 

 the melting point of the metals, when the silver disappeared, and, 

 after the surface was cleaned with diluted sulphuric acid, it pre- 

 sented a uniform surface of copper. This plate was next immersed 

 for a few minutes in a solution of muriate of zinc, by which the 

 surface of copper was removed and the surface of silver again 

 exposed. The fact had long been observed by workmen in silver- 

 plating, that in soldering the parts of plated metal, if care be not 

 taken not to heat them unduly, the silver will disappear. This 

 eifect was supposed to be produced by evaporation, or the burning 

 off, as it was called, of the plating. It is not improbable that a 

 slow diffusion of one metal into the other takes place in the case of 

 an alloy. Silver coins slightly alloyed with copper, after having 

 lain long in the earth, are found covered with a salt of copper. 

 This may be explained by supposing that the alloy of copper at 

 the surface of the coin enters into combination with the carbonic 



