142 



THE APPARENT POSITION OF THE ZODIACAL LIGHT. 



fine the comparisons to the few adjacent groups which differ only a little in 

 longitude, since the unsystematic variations are evidently large. The remainders 

 in absorption and latitude, given in Table II., were therefore adopted as the 

 basis of the corrections. 



The first step taken was to change the signs of corresponding remainders when 

 necessary, so that those derived from the differences of absorption might all be 

 positive. Among the nineteen remainders in latitude formed from the evening 

 observations only four were then negative, and the corresponding remainders in 

 absorption were small. Four negative quantities also occurred among the fifteen 

 remainders in latitude formed from the morning observations. One of them cor- 

 responds to a large remainder in absorption; this is due to the influence of an 

 abnormal group of observations, which will again be mentioned. The general 

 ao-reement of the signs illustrates the dependence of the latitude upon the effects 

 of absorption ; but the variations in the relative magnitude of the remainders are 

 too irregular to allow a satisfactory system of corrections to be immediately 

 apparent from a graphical arrangement of the data. In order to avoid the pos- 

 sible effects of prejudice, without adopting a process involving more labor than 

 would be warranted by the degree of accuracy to be expected in the result, the 

 remainders derived from the differences of absorption were next arranged, accord- 

 ino- to their magnitude, in groups of five each, so far as practicable, and the 

 corresponding remainders in latitude were also collected. In this manner four 

 groups were formed from the evening observations, and three from the morning 

 observations. In each part of Table III. the successive columns give the number 

 of remainders in each group, their sums, their mean values, and their average 

 deviations from these meaus. 



TABLE III. 



