visible in the Solar Spectrum. 9 



§ 6. When the bodies are such that a change from one form of 

 aggregation to another occurs without any sensible change in the 

 velocity of the molecular vibrations or any internal operation 

 taking place, the latent heat may be explained by a simple 

 rearrangement and varied grouping of the molecules. If we call 

 m' and m the mass of a molecule in the solid and fluid state of 

 a body whose mass is M, so that 



and if we furthermore put t 7 and r, p 1 and p for the molecule's 

 time of oscillation and radius vector, we shall have 



2 



G-f5-*-=-j>«}--* 



or 



2(m'. A' 2 , s' 2 - m . A 2 , s 2 ) = M . L, 



where L expresses the latent heat, and A', A the amplitudes of 

 the molecule's undulations. 



If a change in the form of aggregation take place without a 

 change in the velocities of the molecules' vibrations, then 



A' 2 .s' 2 = A 2 .s 2 , 

 and accordingly 



2m'. A' 2 .*' 2 . (l-^ = M.L. 



If we call the body's specific heat in the solid and fluid forms 

 respectively d and c, then experience shows us that in general 



m.c—m'.c 1 ', 



and as the zero of the Centigrade scale is arbitrary, we may 

 assume 



A' 2 .s' 2 = & + *, 

 whence we get 



(b + t°).{c-c ! ) = L, dlcr=L, 



if the specific heat of water be taken as the unit. 



This formula, which was already known in the time of Craw- 

 ford and Gadolin, and occurs in their researches on specific heat, 

 has in later times been given by Persoon, who has also verified 

 it for a variety of different bodies. 



§ 7. In a memoir " On the Meaning of the Plane of Polariza- 

 tion*," I adduced some experiments to show that, for crystals with 

 different axes of elasticity, the velocity of the molecular vibrations 

 is different in different directions. In experiments with rock- 

 crystal, tourmaline, and felspar (the two last opake), the heat 



* Poggendorff's Annalen, vol. xc. p. 582. 



