MEMOIR XXVIII. ] DISTRIBUTION OF HEAT IN THE SPECTRUM. 399 



screen should be a little beyond D. The light on the 

 face of the pile will now be greenish-blue. 



The screen n is then placed so as to intercept the in- 

 troraitted beam. When the needles of the multiplier 

 come to rest, they give the working zero, which must be 

 noted. 



The intromitting screen n being now removed, the mul- 

 tiplier will indicate all the heat of the more refrangible 

 rays ; that is, from a little beyond D up to H 2 . The force 

 corrected for the working zero is to be noted. 



The screen i is then removed to the line A, so as to give 

 all the radiations between the lines A and H 2 . The light 

 on the face of the pile is white, and the multiplier gives 

 the whole heat of the visible spectrum. By subtracting 

 the foregoing measure from this, we have the heat of the 

 less refrangible region ; that is, from A to the centre of 

 the spectrum. 



As a matter of curiosity the experimenter may now, if 

 he please, remove the screens h, i ; the light on the face 

 of the pile will still be white, and the multiplier will 

 give the force of the entire radiations, except so far as 

 they are disturbed by the thermochrose of the media. 

 These measures, as they do not bear upon the prob- 

 lem under consideration, I do not give in the following 

 tables. 



Instead of advancing the screen i from the less tow- 

 ards the more refrangible regions, I have very frequently 

 moved li from the more refrangible to the less. When it 

 is brought down from H 2 to the centre of the spectrum, 

 the light on the face of the pile is of an intense orange- 

 red it might perhaps be called a bromine-red. I need 

 not give further details of this mode of experimentation, 

 as I did not find that its results differed in any impor- 

 tant degree from those obtained as just described. 



The variation in different experiments may generally 



