MOUNT WILSON OBSERVATORY. 



247 



The mean of noon measures is X6291.2147, and of the morning and 

 evening measures X6291.2142. To confirm Perot's observations, the 

 noon measures should have exceeded the low-sun measures by 0.043 a. 

 This result was checked by a direct comparison between the iron-arc 

 and the oxygen lines on May 27, 1920, in the Pasadena Laboratory, 

 and on June 16, 1920, on Mount Wilson. 



On the Water-Vapor Lines at X5900. 



In 1921, 12 lines of water-vapor were measured with high and low 

 sun. Because of their weakness at times of high sun on Mount 

 Wilson the measures were difficult. The solar lines were used as 

 standards. The deviations from the mean for high and low sun are: 



For recession, as found by P^rot for the oxygen line, the morning 

 and evening measures shoiild have been 0.038 and 0.041a less than 

 for the high sua. 



In these observations oxygen lines of the a band have been referred 

 directly to iron-arc lines by both grating and interferometer, and indi- 

 rectly through solar standards obtained by direct comparison. From 

 the agreement the following conclusions seem to be justified: 



1. The wave-lengths of the atmospheric lines are practically con- 

 stant. 



2. In view of possible convection currents, the wave-lengths of the 

 solar lines at the center of the sun are also remarkably constant. 



3. The international wave-lengths of the solar standards used are 

 determined to a high precision. 



WAVE-LENGTH OF SKYLIGHT AND OF SUNLIGHT REFLECTED FROM VENUS. 



From the discussion of 56 spectrograms of Venus and 41 of the sky, 

 Messrs. St. John and Nicholson find that after all known corrections 

 for velocity have been applied, the observed differences between the 

 wave-lengths in skylight and in sunHght reflected from Venus at 

 various phases can be well represented by the empirical formula 



cot h 

 aX103 = 1.3-8.5— ^ 



where D is the semidiameter of the planet in seconds of arc and h its 

 altitude at the time of observation. The formula is derived from the 

 following considerations : A correlation with altitude is shown by spec- 

 trograms taken at both high and low altitudes on the same night, 

 the plates at the lower altitude giving the shorter wave-lengths. A 



