344 KECORD OF SCIENCE FOR 1886 



a velocity greater than that of the gaseous inoleeules themselves at that 

 temperature, about 485 meters per second. But in the experiments 

 now made, even with a notable counter-pressure, velocities of 0,000 me- 

 ters a second were observed. (Ann. Chim. Phys., March, 188G, VI, vii, 

 289.) 



Schneebeli has determined the absolute value of the friction -coef- 

 ficient for air by measuring the volumes which passed through a capil- 

 lary tube of known length and diameter, between the two extremities 

 of which a constant difference of pressure was maintained. The values 

 obtained range from 0.0001090 ,to 0.0001734, the mean being 0.0001707, 

 closely according with that of Obermayer, which was 0.0001700, when 

 the difference of pressure was variable and 0.0001701: (corrected) when 

 this pressure was maintained constant. (x\.rch. de Geneve, 1885, xiv, 

 339: J. Phys., June, 1880, II, v, 290.) 



Toadinson has studied the viscosity of air by means of the torsional 

 vibrations of a pair of cylinders or a pair of spheres, suspended verti- 

 cally from and at equal distances from the center of a horizontal, cylin- 

 drical bar, the whole oscillating in a sufficiently uucontined space. The 

 bar was supported by a rather fine wire of copper or of silver. The co- 

 efficient of viscosity was determined from observations of the logarith- 

 mic decrement of amplitude of vibration, produced by the resistance of 

 the air to the oscillating spheres or cylinders, the distance of the cylin- 

 ders or spheres from the wire being such that the main part of the loss 

 of energy resulting from the friction of the air was due to the pushing 

 of this air. Five sets of experiments were made with this apparatus, 

 giving for the extreme values of the coefficient at 0°, 0.00017201 and 

 0.00017401, the mean being 0.00017 .'90, with a probable error of only 0.14 

 per cent. (Nature, February, 1880, xxxiii, 403.) 



A subsequent experiment was made by Tomlinson at the suggestion 

 of Stokes, in which a hollow paper cylinder about 2 feet in length and 

 inches in diameter was suspended through its axis to a light, hollow, 

 horizontal bar about seven inches long, to the middle of which the wire 

 was soldered. The mean value obtained for the coefficient was 

 0.00017740 at 12.050; that previously obtained having been 0.00017711 

 at 11.79°. In the latter case the loss of energy is due to the dragging 

 of the air. (Nature, December, 1880, xxxv, 105.) 



Holman has studied the efiect of temperature upon the viscosity of 

 air and of carbon dioxide, using for this purj)Ose cai)illary tubes about 

 30 centimeters in length. The results of the measurements seem to 

 show conclusively, as the author thinks, that the variation of the viscos- 

 ity with the temperature of the gas, at least in the case of dry car- 

 bonic acid and of dry air freed from carbonic acid, which maybe taken 

 as typical gases, is not proportionate either to the square root or to any 

 numerical power of the absolute temperature reckoned from — 274°. 

 They point therefore to the inference that all hypotheses yet advanced 

 to account for the variation of the viscosity of gases and hence also for 



