368 SCHAEFFER. 



get a record of the shorter wave-lengths, that no satisfactory measure- 

 ments of opacities could be made on them, except for a very limited 

 range of wave-lengths. The opacities in this case for X3404 were con- 

 siderably greater than for the same wave-length recorded through 

 2300 meters of air while for X2748 and X2573 the opacity was much less 

 than for that wave-length recorded through the shorter distance. 

 However, the plates for the 2300 meters distance give quite a definite 

 idea of the effect. These exposures were taken with the spectograph 

 at Jefferson Physical Laboratory and the source of light at the Chapel 

 at Tufts College. 



The decrease in intensity for the shorter waves may be seen by 

 comparing the opacities of the plates for the different spectral images 

 obtained with a long air path, with the corresponding opacities pro- 

 duced when the air path between the source and the plate is negligible. 

 If /i is the intensity of the light which reaches the photographic plate 

 for a gi\'en wave-length Xi, and I-i, the same for light of shorter wave- 

 length Xo, then the ratio iJh for the case with a long air path is less 

 than the same ratio for a short air path. The ratio is different be- 

 cause the apparent absorption is greater for the shorter wave-lengths. 

 This can easily be seen by direct comparison of exposures (1) and (4). 

 If i? = h/Ii is the ratio of the intensities of these two lines as produced 

 in the laboratory when the air path is negligible, and IV = I'ljl'i is the 

 ratio for the long air path, then /?'//? gives the relative intensity of the 

 light of shorter wave-length after passing through the long column of 

 air, compared with the intensity of Xi. This ^alue of R' IR was de- 

 termined from the opacities produced on the photographic plates for 

 the principal lines in the ultra-violet spectrum of Cd, X3404 being used 

 as the standard. The absorption for light of wave-length greater 

 than X3404 was found to be small and this line was most suitable since 

 the opacity of the silver deposits for this wave-length was not too great 

 to be easily measured. 



Photometric mctJiod. — The notation generally accepted in photo- 

 graphic photometry '' is used in this paper. Z) = log 0, D = K log 

 E and E = It^ where D is the density of silver deposit, the opacity, 

 K a constant, E the exposure, I the intensity of the light which acts 

 for / seconds and j^ is a constant. 



The opacities were measured with Dr. Stetson's ^ apparatus for 

 determining stellar magnitudes. In this apparatus the photographic 



7 Sheppard and Mees, Investigations on the Theory of the Photographic 

 Process, (Longmans), p. 45. 



8 Stetson, Astrophysical Journal, 43, p. 253, 1916. 



