276 METHODS OF VENTILATION. 



perature of the air entering the house and the atmosphere of 

 the house itself, the greater will be the movement produced 

 among the particles. The motion is in exact proportion to the 

 difference of temperature ; and hence the necessity of admitting 

 the external air, in small quantities, when the external ther- 

 mometer is low. The slightest cause that disturbs the equilib- 

 rium of the air produces a motion. It is more sensible than 

 the most delicate balance. It is put in motion by the slightest 

 inequalities of pressure, and by the smallest change of tempera- 

 ture. It is speedily rarefied by heat, and thereby rendered 

 specifically lighter than the neighboring portions, so that it 

 descends, while colder, and consequently denser, flows in, to re- 

 store the equilibrium. It will be easily seen, from the very 

 nature of this law, that an equilibrium cannot be maintained in 

 the artificial atmosphere of a hot-house, since the source of 

 radiation must necessarily be confined to too small a surface to 

 equalize the ascending heat; and, on the other hand, the con- 

 densation by cold is too irregular throughout the heated vol- 

 ume. This irregularity, produced by its unequal action on 

 different parts of the house, must ever render it impossible to 

 obtain an equality of temperature throughout an atmosphere 

 heated by artificial means ; and the larger the house, the greater 

 will be the difficulty of maintaining an equilibrium in its various 

 parts. So much so is this the case, that, as has been already 

 stated, the difference has been found to amount to 100. 



"Gaseous bodies expand equally for an equal increase of 

 temperature, as measured by the thermometer. Gay Lussac 

 showed that 100 measures of atmospheric air, heated from the 

 freezing to the boiling point, became 137.5 measures ; conse- 

 quently, the increase for 180 Fahrenheit is \^ of its bulk. 

 Dividing this quantity by 180, we find that a given quantity of 

 dry air expands T ^ of the volume it occupied at 32, for every 

 degree of Fahrenheit. New experiments have been made by 

 Kudberg, within a few years, giving ^ T as the ratio of expan- 

 sion for one degree of Fahrenheit ; and these results are con- 

 firmed by Regnault. This last number may be adopted as the 

 true increment. 



"If we wish to ascertain the volume which 100 cubic inches 



