between Metallic Masses having different Temperatures. 183 



will take place if the handle, which is the other point of sup- 

 port, terminate with a round end : should it, however, termi- 

 nate with an ornamental device, which affords two points upon 

 which it can rock, the necessary impulse will be gained. We 

 presume, therefore, that the time of contact of two points of 

 the metals must be longer than thai of the intermediate portions. 

 This condition is readily fulfilled by a vast variety of forms of 

 apparatus, and the rudest masses of metal, such as a poker, 

 when duly heated and placed upon lead, will produce active 

 vibrations. The variations of tone produced upon the appara- 

 tus by mechanical interference is easily explained : if a slender 

 rod, with metallic balls at its extremities, be placed across a 

 vibrating bar at right angles to its axis, the time and the arcs 

 of oscillation will be extended, the matter being thrown more 

 to the sides; hence the note will become much lower, and vi- 

 brations previously quite insensible will become visible. Again, 

 if while a bar is in a state of active vibration, it be gently 

 pressed from above, the extent of its vibrations will be dimi- 

 nished, and the time will be reduced ; hence the note will 

 rise. 



38. As it appeared essential to the experiment that the vi- 

 bration should take place between two points which were 

 longer in contact with the block than the other portions, it 

 seemed important to determine whether the connexion of 

 these points was essential. With a view to determine this, 

 I constructed a bar of lead of the same figure as those which 

 I usually employed, such as A B, fig. 5. I let into it a stud 

 of copper a, of which the surface corresponded with that of 



Fig. 5. 



the rest of the bar, and similarly two small ones be, forming 

 the two solid angles upon which the bar was to vibrate, but 

 totally distinct from one another. Whether upon the com- 

 plete bridge a, or upon the divided bridge be, the bar, when 

 heated, and placed upon a block of cold lead, vibrated pre- 

 cisely as if the entire bar had been made of copper. In an 

 early part of this paper 1 described an experiment, in which 

 the points of tlie block upon the bar impinged, were com- 

 pletely distinct pieces of metal : see fig. S, We therelbre 

 conclude that the impulse is received by a distinct and .s<p>a- 

 ratc process at each contact of the bar with the block, and that 



