496 METALLURGY OF CAST IRON. the iron complained of. As a founder, I know there are ways in which the original character of pig metal can be so altered in mixtures as to place upon the ftirnaceman the blame for bad results for which he is not justly responsible. In such cases, the remelting of a sample by him might often exonerate him.* The expense of a small sample cupola need not alarm any furnaceman; he can erect one for twenty dollars. In fact, the author erected one at the Spearman Fnr-nace, Sharpsville, Pa., January 17, 1896, which did not cost six dollars, and took but seven hours' labor of one man from the time ground was broken until the cupola was at work. A cast was made in ten minutes after the iron was charged. This cupola was made of an old shell, twelve inches in diameter and thirty inches long, which was lying around in our foundry yard. It had been used a few years previously in an industrial street parade, for casting horseshoes, which were thrown to the people as the wagon went along, the blast being furnished by means of an old pair of hand-bellows. If iron can be melted under such conditions, in such a baby cupola, no one need hesitate to believe that it can be conveniently done in a small cupola at a blast furnace, where all the blast required can be steadily supplied. The following Tables 105, 106, and 107, seen on next page, give chemical and physical tests of a furnace-cast, taken January 18, 1896, at the Spearman furnace, Sharpsville, Pa., and is chiefly given to present one good form for such records: ' * If founders knew that furnacemen tested their own iron by remelting it in a cupola and kept a regular record of all their tests, it would have a great tendency to make many investigate thoroughly to find whether the fault was not their own before entering complaints to the furnacemen.practice, inudi better with round bars than with square »»nis.burs..........................*fi$7 "