upon different Metals. 5 



cury formed definite solid compounds with the several metals, 

 which were capable of being held in solution by an excess of 

 the fluid ; but were also capable, in favourable circumstances, 

 of separating from it, and crystallizing in peculiar forms. 

 Whether, at the same time, any other compound may have 

 been formed of an essentially liquid nature, I have not ex- 

 amined; but I may here remark, that the manufacturers of 

 looking-glasses have made the observation, that the mercury 

 which is pressed out of the tin amalgam, which they apply to 

 the backs of their plates, is in as pure a state as that which 

 they originally make use of. 



EXPERIMENT VII. 



A square bar of tin, about five inches long, and whose sides 

 were a quarter of an inch wide, was laid horizontally in a card- 

 tray, and just covered with mercury. To render the action as 

 equal as possible, it was frequently turned upon its different 

 sides, and examined. At the expiration of twenty-four hours, 

 minute fissures began to appear along all its lateral and termi- 

 nal edges. The process was continued, and the cracks widened, 

 until, on the third day, they opened to such a degree as to 

 shew that the bar was resolved into four equal trihedral, rect- 

 angular prisms, with two equal angles. They were readily 

 separated from each other by the point of a penknife, and two 

 similar pyramids, whose angles at their bases were 45, were at 

 the same time detached. This groupe is accurately represented 

 in their relative positions, a little separated, at Fig. 1, Plate I. 

 a, a, a, a are the small triangular prisms, which, when in contact, 

 made up the original square bar ; and b represents one of the 

 terminal pyramids. All the angles were as sharp and perfect, 

 and the faces as neat, as if they had been carved with tools ; 

 and when brought into contact with one another, they adhered 

 together with some force, from the cohesive attraction of a little 

 mercury which hung about them. This experiment I imme- 

 diately repeated, and obtained the same very remarkable results. 

 I was at first induced to consider this singular phenomenon 

 as dependent upon the original structure of the bar, from the 

 consideration of the following facts, which are well known to 



