400 



ANNALS NEW YORK ACADEMY OF SCIENCES 



position relative to the eyes of the observer. This was to obviate any possi- 

 ble physiological influence which might cause the observer to measure the 

 polar and equatorial diameters differently. An investigation, however, 

 showed that the prism had a sensible effect upon the measures of all diame- 

 ters, equatorial as well as polar. The diameters measured with the prism 

 were all somewhat smaller than those measured without it. As the prism 

 was used continuously after October, 1891, the series of observations are di- 

 vided into two periods by this date. The mean results from the observations 

 in each period are given below, where the various values are expressed in 

 scale-divisions, one division of the scale being approximately equal to 40". 



Table IV. 



As a result of special measures made by Schur and Ambronn, both with 

 and without the prism, Ambronn concludes that all observations made 

 without the prism must be diminished by 0".4 or 0.01 scale-division, in 

 order to make them comparable with those made with the prism. 



Correcting all the observations made without the prism by this amount, 

 taking the general means, and reducing scale to arc, Ambronn finally obtains 

 for the definitive values of the sun's diameter at distance unity. 



Schur 1920".14 ±0".040 



Ambronn 1919''.80 ±0".036 



From these means the yearly residuals were found, and as given by 

 Ambronn are tabulated below. 



Table V. 



