PHYSICS. 591 



thor formulates the general proposition that the atmosphere acts as a 

 vacuum, and offers no resistance to the discharge of air of all pressures 

 above two absolute atmospheres. {Phil. Mug., December, 1885, V, xx, 

 531.) 



Hoffmann has determined the law of the flow of air through a capil- 

 lary tube too short for it to obey Poiseuille's law. He concludes (1) 

 that the principal cause of the failure of this law under these circum- 

 stances is to be sought for in the perturbations produced at the two 

 orifices, particularly at the entering one ; and (2) that if the times of 

 flow of any mass of air whatever be taken as ordinates, and the lengths 

 of the tubes (of constant diameter) through which the air flows as ab- 

 scissas, very flat curves are obtained which ascend rapidly and are in- 

 termediate between a straight line and a hyperbola. ( Wied. Ann., xxi, 

 470; J. Phtjs., November, 1885, II, iv, 512.) 



Hausemann has experimented to determine the law of the diffusion 

 of gases through a porous partition, the gases used being oxygen and 

 hydrogen. The diffusiometer was constructed so as to allow the diffu- 

 sion to take place under considerably higher pressures, for a single gas, 

 than had been hitherto used. From the data obtained the calculation 

 of the difl'usion-coelflcient for the two gases was made without diffi- 

 culty. Experiment furnished perfectly concordant numbers, but the 

 calculation of the coefficient of reciprocal diffusion by means of these 

 coefficients did not succeed. Values were found for this coefiicient of 

 iuterdiffusion which decreased with the time and the upper limit of 

 which is considerably smaller than the coefficient measured directly by 

 Loschmidt. Hence Stefan's theory is not confirmed ; it imj^lies the 

 hypothesis that diffusion is effected in the vacuities of porous bodies 

 according to the same law as in open space. But it is evident that the 

 porous partition exerts a special action which has not been taken into 

 consideration by Stefan. {Wied. Ann., xxi, 545; J. Phys., November, 

 1885, II, IV, 518.) 



Amagat has repeated his experiments on the compressibility of air 

 and of nitrogen with reference to the use of these gases in manometers, 

 extending the range of pressure, which in the earlier series was never 

 less than 20 to 30 atmospheres, down to that of 1 atmosphere. These 

 experiments were made in one the towers of the church at Fouvi^res, 

 where a clear height of 63 meters was available. The method was es- 

 sentially the same as that used in the Verpilleux shaft, though with 

 some improvements. The results agreed perfectly with previous deter- 

 minations, the values for nitrogen not varying in general by one 

 thousandth of the total pressure. But the discrepancies with Cailletet's 

 figures still exist, amounting between 40 and GO meters to more than 2 

 meters of mercury. A table is given of the product pv for nitrogen and 

 air at the temperature of 1C° and at pressures varying from 0-76 to 65 

 meters of mercury. Within these limits either of these gases may be 

 used for filling manometers. Between 85 and 430 atmospheres nitrogen 



