.MiM-.iuVI.] TIIK DIl'TUACTION Sl'ECTltUM. , 



on this topic may be seen in the Philosophical Magazine, 

 1-Vbruary, is i 7 (and in Memoir V. of the present work). 



The centre of the yellow space is therefore the point 

 of a change in photographic action. From the commence- 

 ment of the red to that point the action is negative; from 

 that point to the extreme violet it is positive. The phase 

 of action changes in the centre of the yellow. 



M. Eisenlohr has made an allusion to the heating pow- . 

 er of the spectrum, and to that point my attention has 

 also been directed. This is the result at which I event- 

 ually arrived that the centre of the yellow is the hottest 

 space, and that the heat declines equally to the two ends 

 of the spectrum. I attempted to form a diffraction spec- 

 trum without the use of any dioptric media, endeavoring 

 to get rid of all the disturbances which arise through 

 the absorptive action of glasses by using as the grating 

 a polished surface of steel, on which parallel lines had 

 been drawn with a diamond, and employing a concave mir- 

 ror instead of an achromatic lens; and though my results 

 were imperfect and incomplete, I saw enough to convince 

 me that this is the right course to be taken in investi- 

 gating the problem. It is absolutely necessary to employ 

 a spectrum which has been formed by reflection alone. 



I will also add that, in the experiments here referred 

 to, a method was employed for determining the tempera- 

 tures of narrow spaces, which may be recommended to 

 those who are disposed to resume such inquiries. It is 

 to use a blackened platinum wire, f of an inch long, and 

 fa of an inch in diameter, one end of which is fastened 

 to the end of a bar of bismuth, and the other end to the 

 end of a bar of antimony, each of these bars being of 

 an inch square on the end and 4 inches long, their dis- 

 tant extremities communicating with a galvanometer. 

 By carrying the platinum wire transversely along the 

 spectrum, I expected not only to determine the distribu- 



