162 Prof. Magnus on the Polarization of Heat at 100° C, 



was still possible that it was only the luminous portion of the 

 heat which was polarizable. If, on the contrary, we could prove 

 that the heat which is radiated at an oblique angle from bodies 

 of any temperature, and therefore at a very low one, is also partly 

 polarized, it would be proved that in the case of dark bodies the 

 heat which they radiate comes in part from their interiors, and 

 is propagated in them by transverse undulations. I think, then, 

 we should be justified in asserting that the conduction of heat, 

 or its propagation in athermanous bodies, also depends upon 

 transverse oscillations. 



MM. La Provostaye and Desains state that they obtained polar- 

 ized heat from a platinum plate whose temperature was below incan- 

 descence. Their experiments were made with platinized plates at 

 a temperature of between 330° and 360° C. ; and with these they 

 say they have obtained very distinct polarization, though to a less 

 degree than with smooth plates. It was shown in the above- 

 mentioned experiments with incandescent plates that these do not 

 show any polarization if they are perfectly platinized. The plates 

 used by MM. Provostaye and Desains, therefore, could not have 

 been fully platinized. These experimenters do not enter into any 

 explanation of the polarization of the heat radiated from these 

 plates. It is also clear that by the employment of rough plates 

 they were not in a position to arrive at the conclusion that the 

 polarization depended upon the refraction of the heat issuing 

 from the interior. For the same reason, their results, although 

 very interesting, determine nothing in regard to the point at 

 issue. It was therefore necessary to undertake new experiments, 

 in order to decide whether the heat which is radiated from bodies 

 of a lower temperature (say, 100° C.) is polarized. 



The means hitherto employed for the investigation of the po- 

 larization of heat depend upon its passage through double-re- 

 fracting plates or through columns of mica plates. Neither of 

 these could be here employed, because they do not allow the 

 passage of the dark heat. There only remained, therefore, to 

 make use of reflection for this purpose. Since, however, as is 

 well known, only a small portion of the heat falling upon a mirror 

 is reflected, especial precautions and arrangements had to be made 

 in order to measure the reflected heat. 



The essential portion of the apparatus was a reflecting mirror, 

 65 millims. wide and 130 millims. long, of black polished glass. 

 This was placed at the end of a horizontal tube of 60 millims. 

 diameter and 120 millims. length. This tube was made of stout 

 pasteboard, in order to prevent heat reaching the mirror by con- 

 duction. Both ends of the tube were closed by plates having 

 circular openings of 35 millims. diameter. The axis of the tube, 

 or the line joining the centres of the openings, passed, when pro- 

 duced, through the centre of the mirror, which could receive any 



