8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 68 



reduced total pressure. It may give a fair estimate of the error 

 committed in using these laboratory values for the transmission of 

 the atmospheric water-vapor in the path of the sun's or earth's radia- 

 tion, to use the observations of Miss von PJahr just tabulated. Using' 

 the distribution of aqueous vapor at different altitudes as given by 

 Humphreys/ a vertical column of air, which would give a transmis- 

 sion of 88 per cent with a total pressure uniform throughout at 76 cm., 

 would give, with a distribution of pressures such as actually exists in 

 the atmosphere. 90 per cent in summer, 89 in winter according to the 

 measures of Miss von Bahr. With the distribution of vapor above 

 Mount Wilson, the transmission comes out 90 per cent for both 

 summer and winter. That is, it would take a slightly greater amount 

 of vapor to produce an absorption noted in the spectrum of a celestial 

 body than would be expected from the laboratory data. If the 

 observations are made at the surface of the earth, the difference 

 would be I or 2 per cent and about 3 per cent if made at Mount 

 AVilson. 



Of the effect of temperature on the amount of absorption of gases 

 comparatively little is known. Miss von Bahr ' considers that its 

 eft"ect is certainly different from that of pressure. She interprets 

 her results to mean that an increase of temperature causes a consider- 

 able damping by the absorbing particles whereas an increase in pres- 

 sure up to a certain limit merely influences things by increasing the 

 number of absorbing particles. A comparison of the results obtained 

 with steam with those obtained under atmospheric conditions as 

 regard to temperature indicates that less absorption would be ex- 

 pected at the lower temperature. 



APPARATUS AND METHOD OF OBSERVING 



The arrangement of the optical apparatus is shown schematically 

 in figure 2. Radiation from the source N, composed of Nernst 

 lamps, passed 42.5 meters through the tube T, containing the water 

 vapor, to the mirror M^, 51 cm. in diameter, thence, collimated, 16 

 meters to the flat mirror Mo, 76 crn. in diameter, thence back to M^, 

 and then to focus on the slit of the spectroscope at S. Before enter- 

 ing the spectroscope the beam could be returned over the path 

 through the water vapor by means of two flat mirrors, F^ and Fo, 

 close beside the slit. The first arrangement gave a path through 

 the water vapor of about 117 meters, to which must be added the 



^Bulletin of the Mount Weather Observatory, 4, p. 121, 1911. 

 ■ Annalen der Physik, 38, p. 206, 1912. 



