128 Prof. S. P. Thompson on the 



factorily observed. The ratio of the minor axis to the major 

 axis was not greater than 1 : 2. 



It is greatly to be wished that some more exact and more 

 reliable measurements were made of the conductivity of tour- 

 maline in the two principal directions along the axis and 

 across it. Although these more reliable measurements are 

 still wanting, I am disposed to think that my own experiments, 

 backed by the analogy of the thermal conductivity (about 

 which there is no uncertainty whatever), are sufficient to 

 establish qualitatively that the electric conductivity of tour- 

 maline is greater in directions at right angles to the axis 

 than in the direction of the axis. If this be established, the 

 theory of opacity advanced above is proved. 



26. There is, indeed, one other possible way of accounting 

 for opacity in a homogeneous medium besides the supposition 

 made above, that the medium should possess electric conduc- 

 tivity of one kind or another. This alternative way of explain- 

 ing opacity supposes that a luminous oscillation may pass into 

 a calorific one of the same period. The vibrations of light 

 being regarded as due to rapid alternations in vector-potential, 

 the particles of a medium transmitting the luminous waves 

 must be subjected to alternately directed electromotive forces. 

 Now let there be in the medium a particle of matter whose 

 natural period of oscillation coincides with that of the wave. 

 This particle will be subjected to alternate inductions which 

 will displace electricity in it from side to side, the (electro-) 

 kinetic energy of the movement changing to (electro-) 

 potential energy and back again to the kinetic form at each 

 half swing, as in the alternations of the swinging pendulum. 

 There is no difficulty in understanding that, under this action, 

 the mass of the molecule will be set swinging. It is easy 

 to set a pendulum, consisting of a metal ball hanging by a 

 silk thread, into vibration by subjecting it to periodic electric 

 inductions whose period agrees with that of the pendulum. 

 Thus the (electro-)kinetic energy of the wave passes over 

 into the (pondero-)kinetic energy of the oscillating particle ; 

 which is merely another way of saying that the light is ab- 

 sorbed and changed into heat. This action, however, can only 

 account for absorption on the assumption that the medium is 

 homogeneous, and its individual particles themselves conduc- 

 tors. But if the medium be homogeneous, all its particles will 

 have similar periods of oscillation. This hypothesis may 

 therefore account for selective absorption in media, though it 

 is not adequate to explain general absorption. If the medium 

 be not homogeneous, a third explanation of opacity is possible. 

 In a non-homogeneous medium there is no need to suppose 

 conductivity to exist; for heterogeneity of structure affords in 



