26 INFRA-RED ABSORPTION SPECTRA. 



This heater is not adapted to the region from 0.8 to 2 ;«, on account of 

 the weakness of the radiation. Here the " glower," or an acetylene 

 flame, is preferable. 



The constancy of the radiation and the cheapness of the " heater " 

 make it a most useful source of infra-red radiation. Moreover, it has 

 no products of combustion, such as water vapor, hence the room can 

 be kept free from moisture, thus protecting the prism. 



METHOD 01^ OBSERVATION. 



The method of observation consisted in projecting successive por- 

 tions of the spectrum upon the radiometer vane, noting its deflection 

 when the absorption cell was before the slit Fi (fig, 5), and also the 

 deflection when a clear piece of rock salt was substituted. The ratio 

 of the deflection obtained for the radiation, from the " heater," which 

 passed through the cell to the deflection for the radiation which passed 

 through the clear plate of rock salt gave the percentage of transmission 

 through the liquid directly and more accurately than by finding the 

 absorption of the empty cell and deducting it. This also meant the 

 reduction of the work by almost one-half. By plotting these ratios as 

 ordinates, and the wave-lengths corresponding to the circle readings as 

 abscissae, the " transmission curves " of the different substances are 

 obtained. 



The general method of observation consisted in observing the zero 

 reading through the telescope, then, with the cell before the slit, raising 

 the iron shutter, by drawing a cord which extended to the observer in 

 the outer room. The vane would then be deflected and come to rest in 

 a new position in about 30 seconds. The carrier of the absorption cell 

 was then raised, by drawing a second cord, until the clear piece of rock 

 salt came before the slit. This height was regulated by a suitable stop 

 on the ways, which permitted the cell to rise to a fixed height. The 

 vane would then suffer a still greater deflection, and, when it came to 

 rest, the shutter was dropped — when the deflection decreased to zero. 

 If there was a shifting from the former zero reading, it was noted and 

 the deflections were corrected. The spectrometer was then set for a 

 new position in the spectrum and the operation repeated. If the zero 

 shift was of any significance the reading was repeated, especially when 

 going through an absorption band. 



This method of observation meant a still further saving of time, so 

 that the time to make a single measurement was reduced to about 1.5 

 minutes, while the exploration of the entire spectrum required from 

 3.5 to 4 hours. In the mean time the observations were plotted upon 



