[ 261 ]

the quantity os metal, the finer the vegetations are; thata mixture of two parts of lead and one of silver gave finervegetations than pure silver ; that pure lead had its surfaceperforated too hastily, and that its jets hardened withoutrising high ; that copper is not easily made to vegetate atall, its surface growing so hard as to afford more resist-ance to the fluid underneath than the bottom of the cu-pel does, which last accordingly bursts; and that gold,instead of jets which continue fixed at the lower end tothe surface of the mass, throws off small round grains,sometimes to the distance of ten inches.

M. de Reaumur, from some phenomena in the castingos iron, suspected that this metal expands in fixing ; andaccordingly made several experiments, which are relatedin a paper in the memoirs of the French Academy for theyear 1726, for determining whether iron really expandsat this period of its cooling, and whether it is the only me-tal possessed of that remarkable property.

He observes, that lead, tin, copper and silver, cast intoingots, are always concave or depressed on the upper sur-face, which seems a mark of their having shrunk in fix-ing. He melted each of these metals, separately, in smallcylindrical crucibles, which being quite filled with thefluid metals, a plate of iron was passed over the surfaceto take off such part as might have risen above the edges :when grown solid, they were all found to have sunk con-siderably in the vessels ; and on melting them again theywere found to fill up the space which they had forsakenin cooling. Having melted pieces of each of these me-tals in separate crucibles, he dropt into them pieces of therespective metals unmelted ; they all funk beneath thesurface, and some fell with a thump on the bottom ofthe crucible : from whence it is plain, that silver, copper,lead and tin, are heavier, or more dense, in their solid

1 than