2 1 6 Historical Review of Experiments 



expansive effort of that portion which experienced the 

 greatest pressure, and consequently acquired the greatest 

 temperature, without being able to assume the dimen- 

 sions proper to this same temperature on account of 

 the less heated and less expanded parts ; consequently 

 there must have arisen, necessarily, a certain condensa- 

 tion of the metal in respect of its natural dimensions, 

 which condensation gradually decreased from the point 

 where the pressure was greatest to the surface. We may 

 suppose that this operation took place in a similar man- 

 ner in all parts of the cylinder. 



"As a consequence of this decrease of volume, an 

 amount of caloric was given out equal to that which 

 would have caused a similar increase of volume, on the 

 supposition, that is, that the specific heat of the metal 

 does not change through this range of the scale of the 

 thermometer, and that the expansions are equal ; and 

 this, considering the range of temperatures and the 

 consequent expansions, is probably not far from the 

 truth. The entire amount of heat disengaged would 

 have raised the cylinder to about 180 of Reaumur's 

 scale ; and if the expansion of brass by heat is equal 

 to that of iron, which has been found to be y^or for 

 each degree of the thermometer, the 1 80 degrees would 

 have caused an expansion of y|f^ in each direction, and 

 the decrease of volume must have brought about the 

 same degree of heat if we suppose that the pressure 

 stood in equal relation to this expansion. 



" Now there is a change, and sometimes a very consid- 

 erable one, wrought in the specific gravity of a metal, 

 by percussion, by the action of a fly-wheel, or by the 

 compression of a wire-drawing machine. It appears, 

 for example, that the specific gravity of platina and of 



