HOGG. — FRICTION AND FORCE DUE TO TRANSPIRATION. 143 



Decrement at 11.30 a.m. was 0.00234 

 " " 12.30 P.M. " 0.00234 



" " 4.45 P.M. " 0.00228 



In this time there had been about four hours of continuous pumping. 

 According to the gauge the pressure at the final stage was 0.00019 mm. 

 It is possible, however, that the pressure was not completely equahzed 

 through the apparatus, although the time allowed to elapse between 

 successive strokes of the pump was sufficient to insure that the equal- 

 ization at this time must have been nearly complete. At 9.30 p.m. 

 the pressure had become 0.00024 mm. and the decrement at the same 

 time was 0.00229. No pumping had been done since the last meas- 

 urements were made at 4.45 p.m. At 12.10 p.m. on the next day the 

 pressure was 0.00027 mm., while the decrement was 0.00236. The 

 limiting value of the decrement seems to be not much less than 

 0.00230. 



It is probable that this limiting value of / is due not only to the 

 friction in the fibre, but also to the friction of mercury vapor on the 

 disk. It was noticed that in the spectrum, after the current had been 

 allowed to run for some time, the mercury lines gradually appeared, as 

 if there were some slight deposit of mercury on the walls of the tube 

 which slowly evaporated under the heat due to the current. This 

 would indicate the possibility of a like deposit on the inner surface of 

 the other parts of the apparatus, for it will be remembered that the 

 spectrum tube is placed between the tube of silver and the main part 

 of the apparatus. 



The fact that the limiting value of / obtained is larger than had 

 been anticipated would be explained by the presence of mercury vapor. 

 Its presence would add a third factor to the resistance experienced by 

 the disk, and one which would diminish certainly while pumping was in 

 progress, but which would increase again as evaporation of the mercury, 

 or the diffusion of its vapor to the viscosity apparatus from the other 

 parts of the apparatus, proceeded. 



Proceeding for the moment on the assumption that the decrement 

 due to the air is proportional to the air pressure in the apparatus where 

 the pressure is small, we can deduce the value of that part of the decre- 

 ment due to friction in the fibre and the friction of the mercury vapor 

 combined. For if m represents this constant part of the decrement we 

 have, 



Pi ■.(li-m)::p^: (l^ - m) 



Also, for corresponding values for p and /, we find from Table I, 



