460 Prof. FORBES on the Vibrations which take place 



case when heat is uniformly diffused through a body, and which 

 is manifested by that universal tendency to diffusion. Hence the 

 element of heat is in a state of equilibrium, and the only force 

 which could be excited successfully to produce a separation, would 

 be between the heat residing in the last molecule of one body, 

 and the first of a separate one in contact with it, but not bound 

 to it by cohesive attraction. Suppose, however, this second or 

 free body cooler than the other, a current of heat will be imme- 

 diately created, which, as it is more or less easily received by the 

 cold body than parted with by the hot, will create a stagnation, 

 or a rarefaction of the elements of heat, respectively ; in the for- 

 mer case producing a repulsive action, or recoil through the whole 

 string of elements set in motion ; in the latter we are led to an- 

 ticipate that the action would be attractive. If this view be cor- 

 rect, (and being theoretical, I do not attach great importance 

 to it,) it is easy to see why repulsion takes place only when the 

 cool body has less conducting power than the hot, and why the 

 repulsive energy depends on the difference of these conducting 

 powers. In the case of very bad conductors, such as antimony 

 and bismuth, I conceive that the current has not had time to esta- 

 blish itself. 



63. In the case of electricity, a remarkable similarity of ef- 

 fect is observable, depending on the conducting power of the 

 material through which it passes. All those remarkable repul- 

 sive actions which produce destructive effects in the case of light- 

 ning, take place during the accumulation of impulses in bad con- 

 ductors *. 



* I might point out another analogy in the sudden and forcible action of the 

 hydraulic ram, where the accumulated effect of small impulses produces sudden 

 and intense results, but I am afraid of extending unwarrantably such speculative 

 analogies. The two preceding paragraphs of this paper have been somewhat mo- 

 dified since it was first read. 



