I20 INFRA-RED ABSORPTION SPECTRA, 



ered with a film of F2O3. The film was obtained by dipping the heater 

 in a FeS04 solution, which was then oxidized by heating.^ For the 

 curves a and h the spectrometer slit was the same width. Curve c rep- 

 resents the distribution of curve a for a wider slit. It was obtained 

 while finding the absorption of a certain compound, and is of interest 

 since, by its regularity, it shows the constancy of the radiation. 



The radiation from this heater was found when covered with borax, 

 also when a strip of mica 0.3 mm in thickness was wound on it, but in 

 neither case was the energy curve different from the original. The 

 mica did not show emission minima at 8.4 /a and 9.4 /u. as computed by 

 RosenthaP (fig. 126, curve h). 



That an emission band should exist at 5.2 /u, seemed doubtful. To 

 test this some finely ground quartz (French flint), pure feldspar, and 

 a number of kaolin clays were obtained from Prof. H. Ries, of the 

 department of geology. The original clay was removed from the plati- 

 num wire, and then the heater was covered with a thin paste of one of 

 these clays. After drying, the heater was used as before. In the course 

 of the investigation of absorption spectra as many curves (the "direct 

 deflections ") could be obtained as was desired. The quartz powder 

 would not adhere well to the heater, even when mixed with starch and 

 applied as a paste. The result was that the surface never became so 

 hot as it did when the finer-grained clays were employed. This, of 

 course, decreased the intensity of the radiation, as will be noticed in 

 curve d of fig. 125, where it will also be noticed that the sharp maximum 

 has disappeared, showing that it is not a selective emission band of 

 quartz, but that it is simply due to the adjacent absorption bands in the 

 atmosphere. 



There are really two emission maxima, viz, the partially suppressed 

 one at 2.5 ^i, due to the white-hot platinum wire^ and a second maximum 

 in the region of 3.8 ;u, which may be due to the cherry-red (color) 

 clay covering. The latter maximum is partly reversed by the strong 

 atmospheric absorption band of COo at 4.25 /u,. The platinum con- 

 ductor under the clay is very much hotter than it would be if exposed 

 to the air. It does not seem possible, however, that it is sufficiently 

 volatilized to give emission bands. It would seem as though the clay 

 would be a sufficient covering to suppress them. In fig. 126 is shown 

 the radiation curve c of a different heater, which shows the variation 

 in the distribution of the emission of diflferent heaters, this one being 

 made of a finer-grained clay than the preceding. This heater contained 



^Paschen : Ann. der Physik, 56, p. 762, 1895. 

 ^Rosenthal : Ann. der Physik, 68, p. 792, 1899. 



