HI.] RAIN AND DEW. 51 



it will be found that the outer surface of the glass gradually 

 loses its brightness ; a dimness soon creeps over the surface 

 that was previously clean and bright, and, before long, 

 drops of water may be seen trickling down the sides of the 

 vessel. It is true that certain kinds of glass, such as some 

 of the old Venetian, constantly exude moisture, or "sweat," 

 so that after the surface has been dried it soon becomes 

 moist again — an effect probably due to an excess of soda 

 in the glass. But the moisture which appears on ordinary 

 glass, under the circumstances indicated above, is clearly due 

 to an entirely different cause, since it is produced with 

 equal readiness on glass of any chemical constitution, or 

 indeed upon a vessel of polished metal. It is evident then 

 that the moisture is not derived from the substance of the 

 vessel itself; neither is it obtained by percolation through 

 the walls of the vessel, for the metal has no sensible pores. 

 The only remaining source of moisture is the surrounding 

 medium, or the atmosphere. That medium always contains 

 more or less watery vapour ready to be deposited upon any 

 object when sufficiently chilled, and the necessary refrigera- 

 tion is brought about by the cold water in the glass or the 

 metal vessel. Moisture which is thus deposited upon an}' 

 cold surface, without production of mist, is termed detv. 



The proportion of watery vapour that can be held in the 

 atmosphere depends principally on the temperature of the 

 air ; the lower the temperature the less it retains. If charged 

 so highly with moisture that it can take up no more, the air 

 is then said to be saturated. When a body of moist air is 

 cooled, the point of saturation is gradually reached ; and, 

 when saturated, any further cooling causes a deposition of 

 dew : hence the temperature at which this occurs is called 

 the dew-point. This point may be determined in a variety 

 of ways, but it is interesting to note that some of the instru- 



E 2 



