118 



STATIC ELECTRICITY 



insulation. Boltzmann had found that the strain in a twisted win- 

 decays at this rate. 



But though we cannot as vet give a full quantitative explana- 

 tion of residual phenomena, there can be no doubt that they arc. 

 at any rate, largely due to conductivity in parts, and that 

 Max well's theory contains a large element of truth. On the one 

 hand exceedingly good insulators, such as air and other how 



no trace of residual charge, while on the other hand poor insulators, 

 with structure probably heterogeneous, show residual phenomena 

 in a marked degree. 



Rowland and Nichols* showed that a plate of Iceland >j>ar 

 exhibited no residual effect whatever, as might be expected it lit -tt n>- 

 geneity of structure is a necessary condition for its existence. 



Mechanical model illustrating the theory.f Tin- 

 model consists of a trough of semicircular < T<>^--r< -t KMI ( I'i^. iSl), 

 say 24 in. long, 6 in. diameter, and divided into eight equal compart- 



A 



A 



ir 



Fio. 81. 



ments by a middle partition along the axis and three crow- 

 partitions. 



It is supported at the two ends, so that it can rock about its 

 axis, and a pointer attached to one end moves in front of a scale. 

 Four pipes, with taps, connect the opposite compartments when 

 the taps are turned on. The trough is balanced In weights on an 

 upright, so that when empty it is in neutral equilibrium. On turn- 

 ing the taps off, and on pouring in water to the same depth in all 

 the compartments, the equilibrium at once becomes stable, and the 

 trough, if displaced round the axis, stores up energy. It may be 

 considered as analogous to a tube of strain connecting charges q 

 on the surfaces of two opposite conductors, the angle of displace- 

 ment representing the charge at either end, or the strain along the 



* Rowland's Physical Papers, p. 204. 



f Proc. Birmingham Phil. Soc.. vol. vi. p. 314. 



