406 



ANNALS NEW YORK ACADEMY OF SCIENCES 



mean results for the different years differ radically. The plates in 1871 

 show the equatorial radius to exceed the polar by some O^.S; while the 

 plates of 1870 and 1872, on the other hand, show the polar radius to be 

 the greater by some 0''.2. Forming the mean by weights of the results 

 obtained from the plates in the different years, we see that the yearly means 

 are as follows: 



1870, Sept. 22 +0".50±0".10 



1871, July 19 — 0".32±0".16 



1872, July 2 +0''.22±0".09 



These measures thus seem to indicate a change in the relative sizes 

 of the polar and equatorial radii of the sun. During the interval 1871-72, 

 the polar radius was increasing relatively to the equatorial, and by 1872 

 was decidedly the greater. These changes in the shape of the sun are 

 apparently real changes, and can hardly be accounted for in any other way. 

 The plates were all taken with the same instrument and under the same 

 conditions, and in corresponding seasons of the year. They were nearly 

 all taken in the morning hours and at approximately the same distance 

 from the meridian. So far as can be determined from the data at hand, 

 there is no instrumental explanation for the difference between the results 

 in the different years. 



The conclusion from this investigation is, that during this period, 1870- 

 72, there was a real change in the shape of the sun ; the equatorial diameter 

 first increasing and then shrinking relatively to the polar diameter. 



Forming the mean by weights of the entire series, it is found that dur- 

 ing the years 1870-72 the polar radius exceeded the equatorial by 



p -E.= — 0".06. 



NORTHFIELD PLATES. 



Under the direction of Professor W. W. Payne, Dr. H. C. Wilson has 

 taken a long series of solar photographs at Northfield, Minn. Only a few 

 of these photographs are available for measurement. Dr. Wilson selected 

 and sent to Columbia University for measurement nine plates, which were 

 taken during the years 1893 and 1894, all of which were well oriented and 

 had on them the necessary data for measurement and reduction. 



These nine plates were measured in the same manner as were the Ruther- 

 furd plates, with the following results: 



