No abstract

No abstract

I T has frequently been stated that, on distilling amalgams, metals distill over with the mercury, but the experimental evidence is not conclusive. C. Michaelis l found that electrical conductivity offers a very delicate means of detecting a trace of a metal in mercury. He found no impurity in the distillates from amalgams distilled in a mercury pump vacuum, but when the amalgams were distilled in a partial vacuum of (26 mm.) impurities were detected in the distillates, however in the latter case it is mentioned that there was much bumping. Victor Meyer 2 tested this question by distilling mercury, containing Bi, Pb, Sn, Cu, etc., from porcelain retorts at atmospheric pressure. The distillates were tested analytically and found to contain impurities. Only after the twelfth distillation did the distillate give a negative test for other metals than mercury, but in this work also there is no assurance that some of the amalgam did not spurt over during the distillation, as there is much bumping unless mercury is distilled in a Torricellian vacuum.The well known method of preventing bumping by allowing a gas or vapor to bubble up through the liquid in the still will also effectively prevent the bumping of mercury. Mercury may therefore be distilled under diminished pressure and with very simple apparatus as proposed by one of us. 3 At that time mercury and amalgams were distilled in an aspirator vacuum of 20 mm. pressure, with air bubbling up through the mercury from a fine capillary, as described in the second part of this paper. With this method there is no suggestion of bumping, and although the distillation proceeds very rapidly there is very little ebullition. It is certain that nothing but vapors are carried over to the condensing tube.

391thle capacity should vary little with the pressure. This conclusion is confirmed by the result with chloropicrin reported in this investigation.The capacity per g. of charcoal was 0,2800 g. a t a pressure of approximately 5 X IO^ bars and 0.3100 g. a t 5 X xo4 bars, an increase of only 10% in acity with a Io-fold increase in pressure. The present data are, therefore, qualitatively in accordance with the one layer theory as rega.rds the author wishes to express his appreciation to Professor Prederick eyes of the Massachusetts Institute of Technology for criticisms and suggestions in connection with this work.