[8TARKEY & HARNEs] DEFICIENT HUMIDITY OF THE ATMOSPHERE 209 



An effort was made to increase evaporation from the wet bulb in 

 order to increase the observed difference. The wick was lengthened and 

 more air circulated about the bulb. The readings were then observed to 

 bo 68°F. and 53°r., showing a difference .of 16°F. or 8.88°C.; no gi-eater 

 dilïerence could be obtained. 



This gives a relative humidity of 26 per cent., which was a little 

 nearer the truth. The observed difference between the two thermometers 

 should have been weTl over 20 °F. to have given the correct result. 



It is possible that a revolving hygrometer would have given better 

 results, and one was pnocured, but not in time for a test on this date. 

 Later in the season when the humidity was from 10 to 15 per cent one 

 of tJicse instruments was ti'ied but with little success. Possibly the form 

 we procured was not sufficiently accurate. 



It is evident from our comparisons that the wet and dry bulb in- 

 strument is not suitable for the measurement of relative humidity beyond 

 a fairly large percentage, probably on acoount of the conduction of heat 

 down the stem of the thermometer offsetting the cooling due to evapora- 

 tion. 



How far the instrument is suitable for outdoor readings, when the 

 air is very cold, we are not yet in a position to state, but we are inclined 

 to think it does not give very satisfactory results. For small differences 

 showing high relative humidity it is satisfactory, and. it is from such 

 readings that Glaisher's tables were doubtless compiled. 



We include a test here to show that satisfactory results may be 

 obtained under these conditions. The date of test was May 18th, 1906 in 

 the Physics Laboratory. The Eegnault hygrometer gave a dew point of 

 18.85°C. with the air temperature 21.7°C., which gives a relative hu- 

 midity of 82 per cent. Simultaneous readings with the wet and dry bulb 

 gave : 



for the dry bulb, 21.7 °C. 



" wet " 19.6°C. 



Difference, 2.1 °C. 



This gives a relative humidity of 82 per cent, and a dew point of 

 18.8°C., which is identical with the Eegnault instrument. 



We think then that for high relative humidities the wet and dry 

 bulb hygrometer yields correct values, with moderately high tempera- 

 tures, but that for low humidities the instrument is entirely misleading. 



In view of these discrepancies further tests were carried out in the 

 Hygiene Department, which tests consisted in comparisons between the 



Sec. III., 1900. 14 



