Vibration of Heated Metals. 141 



views which have presented themselves, though with diffidence 

 and brevity. 



1st, As the atmospherical air is thrown into currents when 

 unequally warmed, it may be supposed that the hot bar induces 

 such statical movements, as prove sufficient to occasion its vibra- 

 tions. 



I need not attempt, however, to explain in what manner the 

 currents of air could produce the tremors, seeing that, by Expe- 

 riment 12, &c. it is proved that the air has no concern in the vi- 

 bratory operation. 



2d, As the thermo-electric experiments and discoveries of 

 late years have shown, that when different metals in contact with 

 each other, differ considerably in their temperature, they acquire 

 and exhibit opposite states of electrization, it may be supposed 

 that as soon as the heated bar is laid upon the cold block, the 

 electric equilibrium is disturbed, and that the metals assume the 

 positive and negative conditions. A series and succession of 

 electric attractions and repulsions may be conceived then to com- 

 mence, which by frequent and rapid repetition cause the com- 

 mencement and continuance of the tremors. 



3d, A third hypothesis, and which appears to me the most 

 probable, ascribes the vibratory movements to the usual mecha- 

 nical changes which caloric occasions in passing from one sub- 

 stance into another: I mean the expansion and contraction 

 which accompany alternations of temperature. It is shown by 

 several of the experiments above recorded, that one of the two 

 metals must be a soft one, as lead or tin. Both of them possess 

 in a smaller degree the power of conducting caloric than the 

 harder or more elastic metals which were used for the bars. 



It also appeared, that some degree of roughness of surface of 

 the lead block is essential to the success of the operation. This 

 slight asperity arises of course from numberless points or ridges 

 projecting from the solid mass of metal. 



VOL. XII. PART I. T 



