HOGG. — FRICTION AND FORCE DUE TO TRANSPIRATION. 139 



depend upon how long the Hght is allowed to shine on the vane, be- 

 cause the vane, after it is heated, becomes a radiator, and the surround- 

 ing vessel is heated by radiation from the vane. This increase in the 

 force goes on for some hours, and in that time there is a chance for very 

 inconvenient changes in the temperature of the vessel and vane due 

 to extraneous sources of heat, e. g. the different distribution of the 

 heat in the room when the lamp is lighted, and when the experimenter 

 is in the room. The source of light was enclosed in a double- walled 

 asbestos chimney, which communicated at the tc^ with a flue, but the 

 sides of this chimney became heated in time. It was found best to 

 set the illuminated vane as accurately as possible (a process taking 

 from a half to three quarters of an hour), and then close the shutter 

 and set again. The difference is taken as the torsion necessary. The 

 results given below were obtained in this way. It was found that 

 the zero position of the vane, i. e., the position of equilibrium after 

 the shutter had been closed, did not change as much as 5° in an hour. 

 At this point, however, an error is introduced which tends to make the 

 result too large. The maximum angle of torsion obtained was 250°, 

 so that the error here considered is not very serious for such an angle 

 of torsion ; but as the pressure is diminished the angle diminishes, 

 and at pressures where the angle of torsion has fallen to, say, 25°, 

 this error becomes serious. 



McLeod Gauge.<i. — In reading the gauges, special attention was 

 paid to these points : 



First, the mercury, when lowered, was never allowed to sink much 

 below the point of connection between the volume tube and the pres- 

 sure tube. The reason for taking this precaution is that the narrow 

 bore of the latter offers considerable resistance to the passage of gas 

 into it, and thus when the mercury is raised after having been allowed 

 to sink well down, more than the right quantity of gas is thrown up 

 into the bulb, and the pressure will read too high.21 Still further to 

 ensure accuracy on this point, the mercury is made to rise very slowly 

 indeed until it has passed the entrance to the pressure tube. 



Second, during the ascent of the mercury the bulb and tubes are 

 tapped repeatedly ; and in setting on any mark, care is taken to tap 

 both tubes until one is certain that the mercury has really settled 

 down. 



Third, in the measurement of each pressure the mercury is raised 

 twice, and, for each time, generally one measurement is taken at each 



" Kamsay and Baly mention the necessity of taking tliis precaution. PliiL 

 Mag., [v], 38, 1804. 



