174 Mr Glazebrook, On the effect of moisture in [Oct. 27, 



The values of e observed on one occasion were : 



For Damp air 19-45 mm. 



For Normal state ..11-99 „ 



For Dry air..... 8'44 „ 



the saturating tension being 20'66 mm. ; thus the air was nearly 

 saturated by the draught of damp air. 



The anomalous behaviour of the glass under a draught of 

 moist air may, I think, be explained by some experiments of 

 Magnus, Fogg Annalen, vol. cxxi. Phil. Mag. Ser. iv., vol. 27. 



Magnus found that blowing damp air on to the surface raises 

 its temperature while blowing dry air cools it; the change of 

 temperature being due presumably to slight condensation of 

 moisture on or evaporation from the surface, respectively. I 

 repeated his experiments both with a thermometer and thermopile, 

 and observed the effects. With the thermometer as at first fitted 

 up the effect was very small. Thus there was as the mean of 

 several observations a cooling effect of 0°'03 C, due to dry air, a 

 heating effect of 0°05 C. due to damp air, while the galvanometer 

 needle was affected about equally in opposite directions owing to 

 the draught. I then cleaned one thermometer bulb B carefully 

 and slightly soiled that of another A. The two were enclosed in 

 the same glass bulb, and a draught of air blown on to them both. 

 „ The readings were as follows : 



[Dry 25° 24°-9 24°-8 24°-8. 



[Damp 25°-4 25°-3 25°-2 25°'2. 



B. ff ry 1 25° in all cases. 

 (Damp) 



Thus the draughts of air produce no effect on the clean bulb, 

 while damp air heats the soiled one, and dry air cools it. 



I then supposed that the draught of damp air might slightly 

 heat the surface on which it was blown, and investigated what 

 the effect of this would be. For this purpose a small spiral of 

 platinum wire was placed near the point of the surface on which 

 the light was incident, and an electric current passed through it. 

 The wire thus being heated raised the temperature of the glass 

 locally, thus straining it and it was found that the analyser 

 readings were lowered at once ; and in a very short time the light 

 became so strongly elliptically polarised that it could not be 

 quenched. Thus local heating produces a lowering in the reading ; 

 so that if we suppose that when damp air is blown on to a glass 

 surface which has not been recently cleaned that surface is slightly 

 heated the apparently anomalous optical phenomena observed 

 when the draughts of dry air are blown would be accounted for ; 

 the glass is heated and the reading lowered. 



