Iron • 247 
Second pattern measured in length - 11*100 inches. 
When cast in soft iron - - 10*975 

Equal to l-88rooth part the length of the pattern. 
The breadth of No. I. was 11 inches, that of No. II. 
81 inches ; the thickness in both was *475 inch. 
I shall now finish this paper with some experiments 
made upon the casting of cannon shot. This operation 
is performed by pouring the liquid iron into a mould 
which is divided into two semi-spheres. The mould is 
possessed of a joint, which preserves the sphericity of the 
shot. It is formed by careful turning to gauges made 
j with great care and exactness. This operation exhibits 
very distinctly the laws of shrinkage, contraction, and ex- 
pansion ; and from it I mean to prove the truth of what I 
only before assumed : 1st, that cast iron, when fluid, is 
then more dense than in any other state : 2d, That inime - 
I diately upon its passing from the fluid to the solid state 
) it acquires its greatest volume ; and 3d, That when cold, 
and always in proportion to the absence of heat, so will be 
| the diminished diameter of the shot. 
To prove that cast iron is denser in the fluid state, se- 
veral pieces of iron may be put into a ladle, and hot fluid 
iron poured upon them ; they will immediately rise to the 
surface, and expose a considerable portion of tjieir bulk 
above the surface of the liquid iron. This buoyancy di- 
minishes ; and as the pieces of metal approach more and 
more to the state of fusion that exists in the ladle, they 
gradually sink till they disappear entirely under the sur- 
face ; they then rapidly dissolve, and form a part of the 
fluid iron. 
Melted cast iron supports also lead and tin in the same 
manner ; but these soon become dissipated in the great 
heat of the fluid 
