INVESTIGATION WITH A ROCK-SALT PRISM. 5 1 



Carbon Dioxide. CO2. 

 (Cell, 5.7 cm. ; barom., 74.0 cm. ; temp., 23° ; fig. 22.) 



Made from KXO3+H0SO, and dried in P0O5. 



This gas has been examined by nearly every person who has investi- 

 gated absorption spectra. It is noted for its variation in the location 

 of its maxima, in emission and absorption spectra. This is most evident 

 at 4.4 fi and 14 fi. 



The emission^ band of a Bunsen burner occurs at 4.4 ju,, while the 

 atmospheric absorption band occurs at 4.26 fj.. Rubens and Aschkinass,* 

 using a sylvite prism, studied CO, to 20 /x, and found a large band at 

 14.7 /A for absorption and at 14.1 fi for emission. The column of CO2 

 in the absorption work was 65 cm. in length. It will thus be noticed 

 that at 4.4 fjL the emission band shifts to the longer wave-lengths with 

 rise of temperature, while at 14 fi the shift is toward the shorter wave- 

 lengths. 



Angstrom found the CO2 bands at 2.6 [x and 4.32 fx., while in the pres- 

 ent examination they occur at 2.75 /x and 4.29 fx. The fact that the 

 atmospheric band of CO2 occurs at 4.26 /x leads me to think that the 

 present value is more nearly correct. A slight trace of CO would tend 

 to shift it toward that band, at 4.59 /a. The actual position of the atmos- 

 pheric band of CO,, as located in a radiation curve, depends upon the 

 temperature of the radiator. This is well illustrated in the curves of 

 the heaters (fig. 126), where for the more intense radiation the absorp- 

 tion band is shifted from 4.25 /x to 4.28 fx. The 14.66 fi band is in excel- 

 lent agreement with the value, 14.7 fx, found by Rubens and Aschkinass, 

 when we consider that rock salt is already quite opaque in this region. 

 As a whole, the work agrees w^ell with that of other observers, consid- 

 ering how precarious a problem it is to map out infra-red absorption 

 spectra. 



Carbon dioxide is the only gas studied which has no absorption bands 

 at 4.5 IX and 14 fx. 



From the fact that gases dissociate at high temperatures are we to 

 conclude that as the temperature rises the COg dissociates into CO, thus 

 shifting the emission maximum of COg from 4.26 ju, toward the absorp- 

 tion band of CO at 4.59 fx, and forming the well-known band at 4.4 /a? 



But how are we to account for the great Y band in Langley's holo- 

 graphs, which show a broad band extending from 4.3 fx to 4.65 fx? We 

 can not say that it is due to the combined action of CO and COg, since 

 the CO is not found in the atmosphere. Oxygen has slight bands at 



^Paschen, loc. cit., vol. 53. 



^Rubens & Aschkinass : Astrophys. Journal, 8, p. 191, iS 



