J68 fVipt. Al'iiuy and Col. Festing. [Juno 19, 



The radiation from different carbon filaments gave almost iden- 

 tically the same results, whether the surface was bright or whether 

 the filament was dead black, a result for which we were not quite 

 prepared. For owing to the kindness of Messrs. Woodhoose and 

 Rawson, who specially prepared a lamp, we were able to compare 

 the two legs of a filament, one of which was bright and the other 

 dead black, though the resistance per unit of length was the same, 

 and the legs were in fact identical except in this respect. The bright 

 leg of the loop radiated visible radiation much more freely than the 

 dead black leg, and we therefore expected to find a larger proportion 

 of dark rays in the spectrum. As before stated such was not the 

 case, and we found that the bright filaments were able to stand a 

 higher temperature than the dull filaments before giving out. 

 Table XI shows the result with a filament having a metallic lustre, 

 whilst that of which Table IX is a record was jet black. Tables XII 

 and XIII give respectively the wave-lengths and the constants k 

 and I of this lamp. 



Table XIII. Showing the constants k and I for the different turns of 

 the screw in Table XI. 



In formula w 3 =A;R ! ZR : 



Turns of screw. 



For 17 /=-186 A; =103 



It seemed of interest to compare a platinum incandescence lamp 

 with a carbon filament, and the results we show in Table XIV, and 

 fig. 5. 



