140 Mr TREVEL, VAN'S Experiments on the 



29. If a hard-metal bar be placed on a thick lead cylinder, 

 or on a lead block, with the end overhanging, and a spirit-lamp 

 under it, it soon becomes sufficiently heated to produce both 

 vibration and sound, which I have kept up for an hour and 

 a half. 



20. The shape of the bars and blocks is of little consequence, 

 as the hard metal of almost any form vibrates when heated and 

 placed on a lead block. 



31 . If thin bars of metal be employed, sound is often produced 

 though the tremors are not visible. 



From the above related experiments, it appears, 



1st, That, in order to produce the vibrations, two different 

 metals must be employed, the one soft and possessed of moderate 

 conducting power, viz. lead or tin, the other hard ; and they take 

 place whether soft metal be employed for the bar or block, pro- 

 vided the soft metal be cold, and the hard metal heated. 



2e?, That the difference of temperature between the bar and 

 the block must be considerable, though the exact amount has not 

 been rigidly determined ; it has succeeded with a difference of 

 170. 



3d, That the surface of the block shall have a certain degree 

 of unevenness, for when rendered quite smooth, the vibration does 

 not take place ; but the bar cannot be too smooth. 



4th, That the interposition of any matter prevents vibration, 

 with the exception of a burnish of gold-leaf, the thickness of 

 which cannot amount to the two hundred thousandth part of an 

 inch. 



5th, That the air has no share in the production of the vibra- 

 tory movements, however much its presence is essential to the 

 production of sound. 



The cause of the vibrations is by no means obvious. I shall 

 take the liberty of mentioning those speculations or hypothetical 



