THE APPARENT POSITION OF THE ZODIACAL LIGHT. 145 



Other groupings were also tried. In all cases the evening observations seem 

 to be sufficiently accordant, but the morning observations so much more irregular 

 than when they were grouped according to variations of absorption that they 

 cannot be advantageously combined with the evening observations. An attempt 

 to do this gave corrections for the evening observations, which themselves mimis- 

 takably required systematic correction for the inclination of the ecliptic. For the 

 evening observations separately a uniform system of corrections, represented graphi- 

 cally by a straight line from the origin through the point corresponding to the 

 fourth group in Table IV., appeared to be as satisfoctory as any. The corrected 

 latitudes given for the evening observations in the last colinnn of Table II. were 

 thus obtained. For the morning observations, the quantities in the same column 

 were obtained from a system of corrections represented by a straight line from 

 the origin through the point corresponding to the mean of the final groups of 

 evening and morning observations in Table IV. The choice of this point is arbi- 

 trary, and the result, accordingly, uncertain. 



The three columns of corrected latitudes in Table II. exhibit sj-steraatic vari- 

 ations dependent upon the longitude, and both the evening and the morning ob- 

 servations agree in placing the axis of the light in north latitude for about 

 fifty degrees on each side of the autumnal equinox. Near the vernal equinox 

 the results are more irregular, with a tendency to south latitude. Our present 

 knowledge does not warrant the assumption that the axis should lie on any great 

 circle ; on the other hand, we should not be entitled to regard the results of the 

 evening observations as showing that the axis crosses the ecliptic more than twice. 

 But after making any reasonable allowance for error in the observations and in 

 the method of reduction, it can hardly be doubtful that the axis of the " Stronger 

 Light," as seen by Jones at the elongation 60°, was decidedly north of the 

 ecliptic near the autumnal equinox, and considerably farther south, if not actually 

 in south latitude, near the vernal equinox. It will appear on examination that 

 only an obviously excessive correction for absorption or for the position of the 

 ecliptic will remove the evidence of this variation in the evening observations, 

 while it happens in the case of the morning observations that those made near 

 the autumnal equinox cannot be subject to large corrections on any system. 



It will now be interesting to compare this result with one obtained from obser- 

 vations of the faint light, called " Gegenschein," and occasionally seen in approxi- 

 mate opposition to the Sun. Its centre is usually somewhat to the north of the 



ecliptic, which may indicate an effect of atmospheric absorjition ; but this question 

 VOL. XI. 11) 



