178 CARNEGIE INSTITUTION OF WASHINGTON. 



be of the same order of magnitude as that derived from the study of Green- 

 wich, Pulkova, and the Cape. Many cases may be cited where the refraction 

 that reconciles observations above and below pole will, if used for the equa- 

 torial stars, give very anomalous results. Odessa 1910, observed especially 

 for refraction determination, corrects all the standard catalogues by a second 

 or two at the equator, increasing southward. The anomalous nature of the 

 result is shown by the fact that the sun observations stand out from the 

 tables by like amounts. These anomalies are of the kind to be expected if 

 the observations are affected by differential refraction. 



This investigation, which is still proceeding, is not expected to give definite 

 corrections to the Preliminary General Catalogue. Such a determination 

 must naturally wait for the completion of the reductions of the San Luis 

 and Albany observations, which were especially designed for the solution 

 of this question, among others. It is to be regarded as reconnaissance work. 

 The revision of the Preliminary General Catalogue can be more intelligently 

 planned if such work is done in advance. 



Mr. Varnum has arrived at similar results in quite a different way. Em- 

 ploying the systematic corrections to forty catalogues as given in the Pre- 

 liminary General Catalogue, he applied a formula of the form C^- at -f ht"^, 

 both in right ascension and declination. The derived corrections in dechna- 

 tion agree very closely with those obtained by Mr. Raymond from other 

 considerations. The second-order term was introduced because of the 

 failure of the linear term to fit the observations. Its significance is not clear, 

 but, as has' been pointed out by Mr. Raymond, part of the effect may be due 

 to the different refractions used. 



When the derived corrections are applied to the solution for solar motion, 

 the apex is placed at A = 271? 2, D= -f-26?7, in harmony with the apex 

 determined from radial velocities and that obtained from the space motions 

 of the stars with velocities under 80 kilometers per second. The mean paral- 

 lactic motion is reduced from 3''85 to 3'' 53. The corrections to the positions 

 in the Preliminary General Catalogue are in the direction indicated by Mr. 

 Roy's discussion of the systematic errors of catalogues published since the 

 formation of the Preliminary General Catalogue. 



The correction to Newcomb's equinox for 1908 derived from Mr. Varnum's 

 discussion amounts to — 0^057, in very good agreement with the recent values 

 derived from observations: — 0!052 by Eichelberger for the 1908 Washing- 

 ton catalogue, and —0^068 by Hough for the 1908 Cape catalogue. Dr. 

 Lambert' finds a correction of -f0''0048 to the annual proper-motions of the 

 Preliminary General Catalogue in declination -f39°. Mr. Varnum derives 

 the value, +0r0043. 



DIFFERENTIAL REFRACTION IN ECLIPSES. 

 In order to illustrate the possible effect of differential refraction at the time 

 of an echpse, Mr. Varnum has computed the displacement which might be 

 expected at Albany under the same atmospheric conditions that existed 

 at Wallal during the echpse of September 21, 1922. The resulting value of 

 -MT82 justifies a more serious consideration of the part played by differential 

 refraction at the time of total solar eclipse. 



' Special Publications of the U. S. Coast and Geodetic Survey, No. 80. 



