Heat Radiated by Rough and by Bright Surfaces. 87 



the hinder side as well as the front becoming heated, it was every 

 now and then changed for another one. 



To begin with, the prism was removed and the rays of light 

 from the glowing strip which passed through the diaphragms 

 were received on a white screen. It could thus be proved that 

 the strip in the flame and two diaphragms covered each other; 

 and by using both strips one after another, it was ascertained 

 that one of them took exactly the same position as the other. 

 The prism was then put in its place and the heating etfect ar. tne 

 various parts of the spectrum determined by means of the thermo- 

 pile, which, by means of a contrivance passing through the side 

 of the box, could be moved into different positions. In order 

 to ascertain that the effect of each strip was the same when 

 their surfaces were alike, they were both tried at first with a 

 bright surface. After having found that the thermo-piie gave 

 always the same deflection at the same part of the spectrum, 

 whichever strip was used as the radiant source, one of the strips 

 was covered with platinum-sponge, and the heating effects which 

 it produced at the different parts of the spectrum were deter- 

 mined, the thermo-pile being kept in the same position while one 

 strip acted upon it after the other. 



For the sake of more convenient comparison, the results ob- 

 tained are combined graphically, as in fig. 1, Plate I. The 

 graphic representation of the results is got by raising perpendi- 

 culars from the horizontal line A B, which represent the first 

 swings read off in millimetres on the scale of the galvanometer. 

 The distance which separates these perpendiculars is equal to 

 twice the amount by which the thermo-pile was displaced between 

 each two observations. The curve formed by these ordinates 

 and abscissse consequently represents the distribution of heat in 

 the spectrum. The longer ordinates belong to the platinized 

 plate. In the drawings, the observed values are connected 

 together by straight lines only. 



Before beginning the experiment, the pile was placed at the 

 limit of the visible red. The application of the name of this 

 colour to the ordinate whereby the deflection of the galvano- 

 meter is represented for this part, is consequently justified by 

 observation. The same is true of the yellow ; the other colours 

 of the spectrum could not be distinctly observed. 



Afterwards, in order to get a better separation of the several 

 colours, lenses of rock-salt were employed. 



Prismatic Analysis with Lenses. 

 The rays given out by the glowing strip fell first upon a slit 

 1 millim. in width, which was applied to the side of the box 

 above mentioned. Behind this there was placed, inside the box, 



