1904.] Cotton by Water and ly Water Vapour. 235 



of the belt a slight jolt was thus given to the thermometers which 

 served the purpose of an automatic tapper in preventing the mercury 

 from sticking. 



The actual conduct of an experiment is as follows : The covered 

 thermometer is dried at about 110 for an hour as already described, 

 the jacket tube being placed separately in the oven at the same time. 

 The counterpoise thermometer and tube, if used, are treated similarly 

 in this and in all subsequent operations. The instrument is then 

 removed and quickly inserted into its tube, both being well above 

 100, and is left to cool to the temperature of the balance. Its weight 

 is taken after a momentary loosening of the cork to equalise pressure, 

 and it should not vary in a series of tests more than about O'OOl gramme 

 from the mean. In the twenty experiments of Series V, where the 

 duplicate instrument was used as a counterpoise, the variation was only 

 half as much. The weight of the dry wool itself may be obtained by 

 deducting the tare of the instrument, taken before the bulb was 

 covered. 



The thermometer, still protected by its tube, is then suspended in 

 the immersion vessel (water or saturated air as the case may be) and 

 the cover put on, and readings are taken by the telescope from time 

 to time till their constancy indicates that the covered bulb has reached 

 the fixed temperature of the thermostat. About an hour suffices. The 

 final reading is noted as the initial temperature of the experiment. 

 The thermometer is then as rapidly as possible withdrawn, removed 

 from its tube, and re-hung in the proper position, and a stop-watch is 

 started at the moment of immersion. The cover is replaced, the 

 telescope is adjusted vertically so that its scale occupies exactly the 

 same position as before with respect to a fixed mark on the thermometer, 

 and readings begin. These are taken at short intervals, and always, 

 except at the maximum temperature and at the finish, at the moment 

 when the mercury is crossing a scale line, the time being noted to the 

 nearest second so long as seconds are of any importance. 



Generally the experiment is stopped at an exact pre-appointed time 

 by removing the thermometer as rapidly as possible to its jacket 

 tube, and it is then allowed to take the temperature of the balance and 

 weighed. The difference between this and the original dry weight 

 gives the amount of moisture absorbed during the immersion. As the 

 final temperature cannot be read at the actual moment of stoppage, a 

 very slight extrapolation is generally necessary to complete the curve ; 

 for the exact final temperature may be wanted for a purpose that will 

 appear later. Of course, in the case of water immersion the weighing 

 is omitted, as absorbed moisture cannot be distinguished from that 

 which mechanically adheres to the cotton. 



At the completion of an air immersion experiment, after the instrument 

 has been weighed, it may be at once used for a second experiment in which 



T 2 



