APPENDIX A. 219 



lision of two bodies of differing hardness lead and 

 iron for instance. The first of these metals be- 

 comes very hot, while the second does not vary sen- 

 sibly in temperature. But the motive power is 

 almost wholly exhausted in changing the form of 

 the first of these metals. We may also cite, as a 

 fact of the same nature, the heat produced by the 

 extension of a metallic rod just before it breaks. 

 Experiment has proved that, other things being 

 equal, the greater the elongation before rupture, 

 the more considerable is the elevation of tempera- 

 ture. 



(2) [The remainder is blank.] 



When a hypothesis no longer suffices to explain 

 phenomena, it should be abandoned. 



This is the case with the hypothesis which re- 

 gards caloric as matter, as a subtile fluid. 



The experimental facts tending to destroy this 

 theory are as follows : 



(1) The development of heat by percussion or 

 the friction of bodies (experiments of Rumford, 

 friction of wheels on their spindles, on the axles, 

 experiments to be made). Here the elevation of 

 temperature takes place at the same time in the 

 body rubbing and the body rubbed. Moreover, 

 they do not change perceptibly in form or nature 



