EFFECTS OF VARIATIONS IN MANGANESE. 241 charges of pig iron with different percentages of manganese were made in the side containing the manganese. The weight of the charges was from 45 to 60 pounds of iron with a' range of one to three pounds of manganese mixed with the charges. Heats Nos. i, 2, 4, 5, and 6 were made of Foundry pig iron. Heats Nos. 3, 7, and 9 of Charcoal pig iron, and heat No. 8 of Bessemer pig iron. The analysis of the ferro-manganese was silicon 1.65, manganese 80.34, phosphorus .354, total carbon 5.85, and no sulphur. In adding manganese to the metal in the ladle (this metal was always that coming from the side of the cupola free from the ferro-manganesc mixture), it was broken to the size of a pea and thrown gently on top of the molten metal, and then stirred well with a half-inch rod until all melted and in mixture with the iron, FIG.r making comparative tests, which involve delicate observations and affords a remarkably uniform conditions of fuel, blast, heat, etc., necessary to discover, the true effects of changes in the elements composing cast iron. The fact that the tests in Table 35 were obtained with the use of the cupola, Fig. 56, gives the writer greater confidence in the results shown than he could place in any others obtained in the ordinary way of making separate comparative tests, that is, having one heat taken off one day and another some other day, with the differences in fuels, blast, and heat conditions that usually exist in • making heats in ordinary cupolas.