338 SCIENCE IN SHORT CHAPTERS. 



stated simple laws, must happen in an ordinary English 

 apartment that is fitted, as usual, with one or more win- 

 dows more or less leak} 7 , and one or more doors in like 

 condition, and a hole in the wall in which coal is burning 

 in an iron cage immediately beneath a shaft that rises to 

 the top of the house, the fire-hole itself having an extreme 

 height of only 24 to 30 inches above the floor, all the 

 chimney above this height being entirely closed. (I find 

 by measurement that 24 inches is the usual height of the 

 upper edge of the chimney opening of an ordinary " reg- 

 ister" stove. Old farm-house fireplaces are open to the 

 mantlepiece.) 



Now, what happens when a heap of coal is burning in 

 this hole? Some of the heat from 10 to 20 per cent, ac- 

 cording to the construction of the grate is radiated into 

 the room, the rest is conveyed by an ascending current of 

 air up the chimney. As this ascending current is rendered 

 visible by the smoke entangled with it, no further demon- 

 stration of its existence is needed. 



But how is it pushed up the chimney? Evidently by 

 cooler air, that flows into the room from somewhere, and 

 which cooler air must get under it in order to lift it. In 

 ordinary rooms this supply of air is entirely dependent 

 upon their defective construction bad joinery; it enters 

 only by the crevices surrounding the ill-fitting ^windows 

 and doors, no specially designed opening being made for it. 

 Usually the chief inlet is the space under the door, through 

 which pours a rivulet of cold air, that spreads out as a lake 

 upon the floor. This may easily be proved by holding a 

 lighted taper in front of the bottom door-chink when the 

 window and other door if any are closed, and the fire is 

 burning briskly. At the same time more or less of cold 

 air is poured in at the top and the side spaces of the door 

 and through the window-chinks. The proportion of air 

 entering by these depends upon the capacity of the bottom 

 door-chink. If this is large enough it will do nearly all 

 the work, otherwise every other possible leakage, including 

 the key-hole, contributes. 



But what is the path of the air which enters by these 

 higher level openings? The answer to this is supplied at 



