78 



MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



has beeu talven as 1, aud the intensities at otlier times divided by tlie intensities given by each 

 method at the chosen time. lu the first two methods the relative values are less than 1, because 

 the intensities at 12.00 on May 1 are the highest observed (with one or two exceptions) ; but in the 

 last method there are values greater than 1, because this is not the case. The relatively small 

 value at 12.00 on May 1 seems to be due to accidental conditions. 



The following table contains the relative intensities by the three methods : 



Comxi<(rison of relative intensities by the three methods. 



APRIL 28. 



APRIL 30. 



MAY 1. 



