10 On the Fraunhofer -lines visible in the Solar Spectrum. 



which, on the crystals being heated, radiated from a surface cut 

 parallel to the axis of the crystal *, showed no signs of polariza- 

 tion, whereas the heat from a disk of gypsum, parallel and easily 

 permeable, when heated to a temperature at which it lost its 

 water of crystallization, was evidently polarized. 



The intensity of the heat that radiates from a surface must 

 depend, 1st, on the velocity of the molecular vibrations ; 2ndly, 

 on the depth under the radiating surface to which the particles of 

 the body can be considered as participating in the radiation : and 

 as that depth must, at least in the case of tourmaline, be different 

 for the vibrations parallel to and perpendicular to the axis of the 

 crystal, it necessarily follows that the amplitude of a vibration 

 must be greatest in that direction in which the depth of the 

 radiating particles is least. 



This result has, with respect to tourmaline, been confirmed by 

 Mr. B. Stewart's interesting experiments " On the Nature of the 

 Light emitted by heated Tourmaline "f. 



§ 8. In the absorption of light and radiant heat, i. e. in the 

 transition from an modulatory motion of the particles of the sether 

 to a pendulous motion of those of the medium, the amplitudes 

 of the oscillations must depend (A) on the molecule itself, its 

 mass and volume, and (B) on the medium's elasticity, or the oscil- 

 lator!/ movements which are possible in that medium. 



Again, in the transition from the molecular motion of the 

 medium to that of the surrounding gether, i.e. in radiation, the 

 connexion stated under the rubric (A), as existing between the 

 amplitude and the mass, must continue to hold; so that one is 

 thus far justified in saying that for a given medium the absorbing- 

 and radiating -powers are equal, a consequence which in fact 

 seems to follow from Provostaye and Desprez's observations. 

 Since, however, the molecules of a body can assume only such 

 oscillatory motions as the medium admits of, and we are therefore 

 obliged to ascribe to the medium different absorbing-powers for 

 different colours, the converse of this theorem cannot be inferred ; 

 for it must be admitted that aether is with equal facility suscept- 

 ible of all possible undulatory motions. The circumstance that 

 a body most readily radiates just the kinds of light and heat that 

 it absorbs, is not therefore to be ascribed to any specific differ- 

 ence of radiating -power, but simply to the fact that the aether 

 cannot assume any other undulatory motion than that which 

 already exists in the radiating medium. 



In the preceding pages we have assumed that light and radiant 

 heat on absorption become therm ometric : this is not, however, 



* Namely, for rock-crystal and tourmaline ; for felspar the radiating sur- 

 face was at right angles to the symmetrical axis, 

 f Phil. Mag. May 1861. 



