12 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 68 



SOURCE OF RADIATION 



The radiation used in the following experiments was finally fur- 

 nished by a lamp of 44 Nernst-lamp double-glowers or 88 filaments. 

 It would have been impracticable to use any inclosed radiator because 

 of the absorption in the inclosing material of the very radiations 

 required. 



Within a rectangular frame of soapstone (a), figure 4, was 

 mounted the series of filaments (b). To the rear of the filaments 

 the soapstone was cut away so that the radiation observed from 

 between the filaments was that from the walls of the room at nearly 

 the temperature of the rest of the spectroscope and not radiation from 

 the hot material of the frame of the lamp.' The glowers were 

 clamped between copper blocks on each side. In the under copper 

 blocks were ruled grooves into which the platinum ends of each 

 glower were carefully placed. Over these ends were laid strips of 

 platinum foil, then the upper blocks were clamped tightly down upon 

 the ends of the glowers. The foil, as well as several horizontal deep 

 saw cuts in the somewhat stifif upper blocks, was necessary to insure 

 sufficient contact on the filaments to hold them in place and to furnish 

 them with electrical current. 



The Nernst glowers do not conduct an electrical current at 

 ordinary temperatures. They Were heated by a blast lamp until 

 they would conduct. Then when once conducting they would become 

 hotter and hotter, conducting better and better as their temperature 

 rose, until they would have melted if they had not been provided with 

 so-called " ballast " series resistances. These iron-wire ballasts were 

 so chosen that their increase in resistance with the temperature nearly 

 compensated the decrease in resistance in the filament due to rising 

 temperature. The ballasts were located in a separate box. The 

 lamps were pushed to their utmost, output and they frequently burned 

 out. Although rated at nearly an ampere each the whole 44 in the 

 second lamp would stand but a little over 15 amperes on an open- 

 circuit voltage of no volts. The black-body curve best fitting their 

 energy curve indicated an effective temperature of about 2,250° K. 



^ This provision was important. In the earlier lamp the filaments were 

 inclosed in a hollow in the soapstone ; but the soapstone, although doubtless 

 helping to keep the lamp at a steady temperature, was not near enough to- 

 the temperature of the filaments to produce approximately equal radiation 

 intensity as a part of a black-body source and too near this temperature to 

 have its radiation negligible, whence the error earlier mentioned. The room 

 temperature was so near that of the bolometer that its radiation seen through 

 the interstices of the filaments was negligible. 



