POOR, THE FIGURE OF THE SUN 393 



observations showed no long-period variation, nor any long existing change 

 in the sun's diameter, of a greater amplitude than ± 0".2. 



As to the second class, or annual variations, Auwers investigated nine- 

 teen independent series of observations, made at seven different observa- 

 tories and with twelve different instruments. In all, these series contained 

 more than 21,000 observations of the horizontal diameter and about half 

 as many more of the vertical diameter. From sixteen of these series, con- 

 taining more than 26,000 observations, Auwers found, after eliminating the 

 effects of personal errors, that the diameter was either constant for the en- 

 tire year, or showed variations of such form and size as to be clearly the 

 result of temperature upon the instruments. In these series he found 

 the measured diameter to be the smallest when the image was sharpest, 

 and at those temperatures when the threads of the reticle were exactly in 

 the focal plane of the objective. 



The three remaining series showed annual variations which could not 

 be accounted for by the effects of temperature. In the Madras series, 

 Auwers traced the apparent variation to personal errors of the various 

 observers. The other two series which showed variations were those very 

 early observations of Maskelyne and Lindenau. To these series Auwers 

 devoted a separate paper,^ and in it he showed Lindenau's deductions to 

 be entirely unfounded. In the early Greenwich observations many cases 

 occurred in which the second limb of the sun was observed on different 

 threads from those used for the first. Such observations can only be util- 

 ized when the thread-intervals are known. Now Lindenau confined his 

 discussion to observations made on the same threads, while Auwers took 

 into account all the observations. To do this Auwers had first to find, 

 from other observations, the thread-intervals; but, while this work was 

 extremely laborious, it enabled Auwers ultimately to utilize three times as 

 many observations as did Lindenau. Still further, Auwers found the Green- 

 wich observations to be a heterogeneous mass made by different observers, 

 by Maskelyne and his assistants. He discussed the personal equations of 

 these various observers, eliminated their effects and finally obtained results 

 radicallv different from those of Lindenau, — results which showed no trace 

 of variability in the sun's diameter. 



The possible departure of the sun from a spherical shape was also dis- 

 cussed by Auwers in these three papers. At Greenwich, Washington and 

 Radcliffe, the vertical as well as the horizontal diameter was measured. 

 Auwers discussed these observations separately, and found the results as 

 given below: 



> Neue Untersuchungen uber den Durchraesser der Sonne, III (Sitzungsberichte of the 

 Berlin Academj% October 31, 1889). 



