MECHANICAL PARADOXES. 



shut there may have been a distinct cold 

 draught from it. If the room was fairly warm 

 and the outside air cool, there certainly would 

 be. For the cold outside air cools the glass. 

 The cold glass cools the air inside which comes 

 into contact with it. The cooling makes it 

 contract, become dense and heavy, and con- 

 sequently sink downwards and move towards 

 the centre of the room while other masses of 

 air move in the upper part of the room towards 

 the window to be cooled in their turn. Thus are 

 established the convection currents, of which 

 one part is the cold draught from the window, 

 which may well have given the old gentleman 

 his cold, in spite of the fact that the window 

 was shut. 



The same convection currents on a large 

 scale are set up in an assembly hall which has 

 a blank wall on one side and a row of windows 

 on the other. On a cold night the cool air 

 outside cools the glass windows, which cool 

 the adjacent air inside. This inside air, con- 

 tracting with the cold and so becoming denser 

 and heavier, sinks and crosses over the lower 

 part of the room, where it is warmed by the 

 heat of the assembled people and perhaps by 

 artificial heating. On the side opposite the 

 windows it rises upwards, and higher up crosses 

 over to the windows to be cooled and made 

 to descend again. If the roof also is largely 

 of glass or other thin material the cooling 



22 



