The experiments described below were undertaken in the course of the Geophysical Laboratory's work on optical glass during the war, in order to determine the order of magnitude of the loss due to volatilization of lead oxide from lead silicate melts under certain conditions. An exact determination of such data, with perfect control of all determining factors, would involve a very comprehensive set of experiments, but approximate results may be obtained with simple means. 1 he material for the

more »

... periments consisted of homogeneous glasses containing, apart from incidental impurities, only lead oxide and silica in various proportions. The heatings were done in the same electric furnace in which the experimental glasses had been made, with the opening covered but not tightly closed. The broken glass was filled into a platinum crucible, which had been previously tested for approximate constancy of weight under the conditions of the subsequent experiment. The crucible with charge was heated until the glass formed a melt with level surface, then cooled and weighed. The furnace was then regulated to the temperature desired for the experiment and the crucible was introduced and held for a certain time, taken out , cooled, and weighed again. The difference between the two weighings gives the loss in the melt and as the only volatile constituent in every case was lead oxide, the total loss may be safely calculated as PbO. The area of the surface was then measured. A'few tests showed that the actual loss from the same melt at the same temperature was proportional to the area of the exposed surface and independent of the weight of the melt. The factor determined was therefore the rate of loss per unit of the exposed surface of the melt. The units used were I hour, I square ct!ntimeter, and I milligram. ,.