438 EXPERIMENTAL INVESTIGATIONS. 



tirely forgotten ; and it was not until some time had elapsed 

 that I could recollect anything about the experiments con- 

 tained in it. I now remember that they were made at the 

 London Institution, and it must have been from ten to fifteen 

 years ago. I have no recollection of the reason why I did 

 not publish them, and can only guess that it was in accord- 

 ance with my general habit of not publishing negative results. 

 The results here, however, though negative, seem to me in- 

 teresting, as positive results would a priori be expected ; and 

 may be worth publishing in the ' Philosophical Magazine.' 



The difference in appearance to an observer of a polished 

 surface when at ordinary temperatures and when ignited is 

 sufficiently marked. The self-luminous character of the ignited 

 body apparently removes the impressions of surrounding 

 objects, and would lead to the belief that reflexion, at least 

 that of the character yielded by polished surfaces, was de- 

 stroyed. Such has been the a priori impression of those whose 

 opinions I have asked on the subject. My own belief was, that 

 if polished surfaces when ignited reflected light, they at all 

 events broke up or scattered the reflected rays, and would 

 cease to have the character of a polished surface ; and that if 

 they reflected light at all or notably, they would reflect it as 

 paper or snow does, dispersing the rays so as to produce a 

 general impression of luminosity, instead of throwing them 

 back in a parallel beam, or one in which they preserved their 

 original relative inclination. 



The subject appeared worth investigation ; and as I could 

 not find that it had been attempted, I determined to make 

 a few experiments upon it. The difficulty which immediately 

 presented itself was, that the surfaces which are mainly em- 

 ployed for polished reflexion being oxidable metals, their phy- 

 sical structure would be changed by the oxidation consequent 

 on incandescence. Gold or platinum, therefore, were the 

 only substances which promised any success ; and the latter, 

 from its reflecting white light and more ready capability of 

 retaining a high temperature, was selected. A strip of plati- 

 num-foil, 2 inches long by 0*2 broad, was firmly stretched on a 

 piece of plate glass, polished with putty-powder and tripoli 

 until it had reached as high a lustre as it could be made to 



