394 



ANNALS NEW YORK ACADEMY OF SCIENCES 



Table H. 



In each of three series, therefore, the vertical diameter appears as larger 

 than the horizontal. 



The inclination of the sun's equator to the horizontal line changes with 

 the season of the year. If, therefore, the sun had a marked ellipticity, 

 the measured horizontal and vertical diameters would each show periodic 

 variations, the length of the cycle being six months. If the equatorial 

 diameter exceeded the polar by 1", then Auwers showed that, at the date 

 of the observations, the measured mean for any month should exceed the 

 yearly mean by the value 



± (0".01 sin 2t + 0".09 cos 2t) 



where t is the fractional part of the year; and the upper sign applies to 

 horizontal measures, while the lower one applies to vertical diameters. 

 Giving the Greenwich and Neuchatel observations weights of 3, the Wash- 

 ington observations 2, and those of Radcliffe 1, Auwers, from the mean of 

 all the observations, found the expressions 



Horizontal diameter — 0".01 sin 2t + 0".06 cos 2t 



Vertical diameter — 0".01 sin 2t— 0".04 cos 2t 



And these expressions show absolutely no relation with the real form of 

 the expression as above given. 



Auwers concluded, however, that a polar compression of 



4 0' 



or an 



excess of 0".5 of the equatorial over the polar diameter, would best repre- 

 sent all the observations. 



HELIOMETER MEASURES. 



In the first of the before-mentioned series of papers,* Auv^ers examined 

 two short series of heliometer measures made by Schliiter and ^Yichmann 



1 Monatsberichte of the Royal Academy of Sciences at Berlin, May, 1S73. 



