PHYSICS. 443 



Franlv Waldo Las suggested the use of Wriglit's apparatus for dis- 

 tilling mercury in the direct filling of barometer tubes. By attaching 

 the tube to be filled to the end of the distilling-tube, the mercury, as it 

 condenses, passes into it at once, and without coming in contact with 

 the air. Certain precautions are detailed in the paper. {Am. J. Sci.j 

 January, 1884, III, xxvir, 18.) 



Schumann has made a long series of experiments on the coefficient of 

 friction of gases and vapors and its dependence on temperature. He 

 concludes (1) that values of this coefficient calculated by the formula 

 of Maxwell show greater deviations from each other than would corre- 

 spond to errors of observation, and this particularly at high tempera- 

 tures; (2) that by introducing a correction this formula gives numbers 

 which at ordinary temperatures show close agreement with those ob- 

 tained by the method of transpiration ; (3) that, owing to absorption, the 

 method of transpiration gives for gases too high values and for vapors 

 too low values of the coefficient of friction; (4) that the dependence of 

 the coefficient on the temperature increases with the temperature ; (5) 

 that the coefficients of all the vapors examined have almost the same 

 function of temperature; and (6) that the relation found by Puluj to 

 exist between length of path and refractive index holds for the vapors 

 of homologous ethers at corresponding temperatures. ( Wied. Ann., 

 xxiii, 353, November, 1884 ; Phil. Mag., December, 1884, V, xviii, 544.) 



Winkelman has studied the phenomena of diflusion of gases and va- 

 pors, in order to compare the results obtained with those calculated from 

 the formulas of Stefan and Meyer. His experiments were made in a 

 tube drawn out at the lower end into a portion of uniform but of much 

 finer bore than the upper portion, and graduated. This lower jjortion 

 contained the liquid which was to be vaporized, the gas into which it 

 ,was to difiuse being passed into the upper portion by means of an 

 entrance and exit tube. For diffusion of steam into hydrogen, into 

 carbon dioxide, and into air, discrepancies between the actual results 

 and the theoretical ones deduced by Meyer's formula were observed. 

 For alcohol and for ether vapor diffusing into these three gases, a fair 

 agreement with Stefan's formula was observed. The mean length of 

 free path of the vajior molecule can be obtained from the latter for- 

 mula, at least approximately. {Wied. Ann., June, 1884, xxii, 1; Am. 

 J. Sci., July, 1884, III, xxviii, 70.) 



Guglielmo also has determined the coefficient of diffusion of the vapor 

 of water into air, carbon dioxide, and hydrogen. He gives for the first, 

 12.8G at 8o and 13.05 at 15°; for the second, 46.95 at 18°; and for the 

 third, 8.38 at 18°. The last two values are 3.52 and 0.628 times the co- 

 efficient of diffusion into air. By Meyer's formula the author has cal- 

 culated the mean length of free jjath of the molecule of vapor of water 

 in an atmosphere of this vapor. He finds the value 0.00000801, that 

 directly obtained by Kundt and Wai;burg being 0.00000640. The fun- 



