IN the dstrophysical Journal for May, I92o, data were published on the gold-point palladium-point brightness ratio. At that time no analysis had been made of the palladium used. Inasmuch as the value for the melting point was somewhat higher than that obtained by Day and Sosman in their work with the nitrogen thermometer, a question was raised concerning the purity of palladium, it has since been possible to test this point, using two separate methods. Through the courtesy of' Doctor Sosman, of the Geophysical Laboratory, a small sample of the palladium used by Day and So.sman was obtained, making possible a comparison of its melting point with that used in this Laboratory.
Both samples were mounted between platinum wires, in the manner described in the original paper, and alternate melts were made in a platinum-wound black-body furnace. At the time of melt the e.m.f, of the thermoeouple was read and recorded. The furnace was then held at this temperature and readings made with the optical pyrometer. At least six readings were made with the optical pyrometer for each melt. The average results of twenty-six melts, twelve on one samp.!e and fourteen on the other, showed a difference between, the melting-point temt)eratures of the two samples o.f very close to one. degree, that furnished by the Geophysical Labo.ratory being higher. The thermocouple and pyrometer readings agreed to within less than this difference.
Small amounts of both samples were then sent t,o the Bureau of Standards for a spectroscopic analysis. Both samples were found to be nearly spectroscopically pure and showed only minute traces of silicon and lead. The sample from Nela Research Laboratory contained more lead but less silicon than that from the Geophysical Laboratory. They further reported that they * Communicated by the Director.