-1859] 



WRITINGS OF JOSEPH HENRY. 



219 



serves to increase the density, and consequently the repulsive 

 energy of the vapor previously existing. Hence, as we have 

 shown before, the expansive power of vapor or steam in- 

 creases in a geometrical ratio, while the temperature in- 

 creases in an arithmetrical ratio, that is, an addition of a few 

 degrees of heat produces more than a proportional degree 

 of elastic force. The case however is very different with 

 vapor separated from the water from which it is produced; 

 it then obeys the same law as atmospheric air and increases 

 in elasticity with equal additions of temperature. 



It has been stated in a previous article that the atmos- 

 phere increases its elastic force by one four hundred and 

 ninetieth part for every degree of Fahrenheit above the 

 freezing point; the vapor of water follows the same law. 

 These facts are readily proved by the apparatus exhibited 

 in Fig. 2. So long as any water e remains 

 above the mercury in the tube a, the latter 

 may be drawn up or pushed down into the 

 reservoir without altering the height of the 

 column of mercury c e. The higher the tube 

 is drawn up, the more water will spring into 

 vapor, while the tension or repulsive energy 

 remains the same, as shown by the invariable 

 height of the mercurial column. When the 

 barometer tube is pushed down into the basin 

 and the space above diminished, a portion 

 of the vapor is converted into water, and this 

 portion increases as the space is made to 

 diminish. If however we draw up the tube 

 so that all the water will pass into vapor, a 

 further elevation of the tube will produce 

 an elevation of the height of the mercurial 

 column ; the vapor will become rarified and 

 its elastic pressure will consequently be 

 diminished, and hence the increased length 

 of the column of mercury. If sufficient cold 

 Fig. 2. and pressure could be applied to atmos- 



pheric air, it is not improbable that a portion might be con- 



