OF ARTS AND SCIENCES. 203 



boiling. This gave a sample of the oxide which was very nearly white 

 after intense heating in a double porcelain crucible. 



In Experiments 22 and 23, pure oxygen was introduced into the 

 flame of the blast lamp used for igniting the zincic oxide, the bottom 

 of the outer crucible being melted where the flame struck it. The 

 material analyzed in Experiment 22 was taken from the layer nearest 

 to the zincic silicate formed in the bottom of the crucible by fusion of 

 the glaze, and that used in Experiment 23 was taken from the top. 

 The zincic oxide of Experiment 24 was heated to bright yellow heat 

 in the same way. All the edges of the porcelain were rounded, 

 and the cover and the crucible itself were fused to the heavy iron sup- 

 ports on which they rested. After twelve minutes the iron itself 

 began to burn and the ignition was stopped. The oxide was found 

 to be " sintered together," and was of a grayish color, as if it had been 

 reduced in part. The inference was corroborated by the presence of 

 zincic oxide on the cover of the crucible and on the sides of the stack 

 of the furnace. The zincic oxide used in Experiment 25 was treated 

 in a similar manner, but proved to be still darker in color. It was 

 afterwards ignited for some time in an atmosphere of oxygen, in 

 order that no reduced metal might be left. 



The materials used in Experiments 26 and 27 were separate samples 

 heated in a double crucible, with the cover of the inner crucible fitting 

 inside the outer one, instead of covering both. This arrangement 

 rendered less likely the entrance of reducing gases, and allowed freer 

 play of air in the furnace. After the latter had been thoroughly 

 heated by the ordinary blast lamp, oxygen was slowly turned on 

 through a Y tube, to replace a large part of the air in the flame. This 

 gave an intense white heat, which melted and burned the tip of a file 

 in the fraction of a minute. In Experiment 27 the glaze of the inner 

 crucible was fused, and the platinum which separated the two crucibles 

 was cemented to the outer crucible. The crystalline residue of zincic 

 oxide after this treatment was white with the faintest trace of yellow, 

 but no sign of gray. As the crucible would not stand a higher heat, 

 and as any quantitative determination could scarcely be carried on at 

 a higher temperature, no attempt was made to push the heat further. 



The fifth series of determinations was made with the object of dis- 

 covering whether the trace of silica and alkali dissolved from the glass 

 and porcelain might be responsible for the occlusion of the gases. 

 First, a sample of the purest zinc that could be obtained was treated 

 with a small amount of pure nitric acid in a porcelain dish, and the 

 solution evaporated to small bulk in the presence of an excess of zinc. 



