346 RECORD OF SCIENCE FOR 1886. 



gases by filliDg air and bydrogeu iuto two equally long barometer 

 tubes so that the mercury is at the same heiglit in each. If, now, ether 

 in excess be added to each tube, the mercury sinks more rai)idl^Y in the 

 tube containing hydrogen, and after a few minutes the tubes show a 

 marked difference of pressure, proving the vapor to diffuse far more 

 rapidly in hydrogen than in air. Ultimately, after some hours, the dif- 

 ference of pressure in the two tubes diminishes and becomes zero. 

 (Wied. Ann., 188G, Xo. 3; Phil. Mag., May, 1S8C, V, xxi, 451.) 



Lommel has described an aerostatic balance, useful for demonstrat- 

 ing the specific gravity of gases in lecture experiments. Under one 

 scale-pan of a balance is hung, by means of a wire, a closed glass bal- 

 loon which is inclosed in a glass vessel having in its cover a small hole 

 for the wire. This vessel has a side tube near the bottom, provided 

 ■with a stop cock. The instrument is balanced, while the vessel is filled 

 with air. If, now, another gas is allowed to stream in aud displace the 

 air, the balloon rises or sinks according as the gas is heavier or lighter 

 than air. By adding weights in one scale pan or the other equilibrium 

 is restored, and it is then easy to find how much, more or less, a volume 

 of gas equal to that of the balloon weighs than the same volume of air at 

 the same temperature and pressure. (Wied. Ann., 1886, No. 1; Nature 

 February, 1886, xxxiii, 397.) 



Grunmach has reported to the Berlin Physical Society on his baromet- 

 ric investigations, and has described at length the arrangement of the 

 normal barometer, the vacuum of which was measured in an electrical 

 way. A combination of the barometer vacuum with a Geissler tube 

 permitted the exhaustion to be examined even beyond the limits of the 

 pressures measurable by the cathetometer. The occurrence of the phos- 

 phorescent light in the spectrum tube is a standard for the highest de- 

 gree of rarefaction, in which the vacuum is filled with mercury vapor 

 having a tension of only 0.01 to 0.02™™. A still better vacuum would 

 be obtained when the mercury was satisfactorily absorbed, a condition 

 which he had in vain tried to accomplish with selenium. A large num- 

 ber of normal barometers were compared with this, by a method already 

 described at length, using the developed reduction formulas. As a 

 result it api)eared that the impurity of the free mercury-cup increased 

 the height of the meniscus aud so the recorded height of the barometer. 

 In the subsequent discussion Goldstein proposed for the electrical 

 measurement of the vacuum, instead of a Geissler spectrum tube, the 

 employment of a wide tube which lets the phosphorescence become more 

 evident. But for the determination of the highest degrees of exhaustion 

 he maiutained that the thermometer was better adapted than the phos- 

 phorescent tube. If a thermometer be placed in a vacuum tube whose 

 positive pole was a jioint, and whose negative electrode was a steel 

 plate nearly filling the tube opposite the cathode, then the thermometer, 

 when the exhaustion readied the point that light ai>peared on the ca- 

 thode, would rise 80° to 00° above the teuiperature of the room. At 



