158 PROCEEDINGS OF THE AMERICAN ACADEMY 



these stars, by considering the Milky Way to be two magnitudes 

 brighter than the mean, instead of the minimum, brightness of the 

 sky. According to this estimate, the brightest parts of the zodiacal 

 light may commonly be three or four magnitudes brighter than the 

 surrounding sky. But at 60° from tlie Sun, it is not likely that the 

 zodiacal light exceeds the Milky Way in brightness, and for that 

 elongation we may suppose that its middle portions will not be more 

 than two magnitudes brighter than the sky around it. At its assumed 

 edo-es, the dilFerence of magnitude will depend upon the choice of the 

 observer with regard to the nature of the boundary to be employed 

 in his observations. Suppose that he endeavors to include all the 

 light which he can see, and that he distinguishes differences of a single 

 tenth of a magnitude, so that, at the boundary which he adopts, the 

 zodiacal light is brighter than the sky beyond it by one tenth of a 

 magnitude. This may perhaps be the case with Heis and Weber 

 (see p. 156). Taking their result of about 10° for half the extent of 

 the light in latitude (see Table I.), we should have a variation of 

 about one fifth of a magnitude for every degree, if we suppose the 

 light to vary uniformly, on the scale of stellar magnitude, so as to 

 reduce the excess of brightness from two magnitudes in the middle to 

 one tenth of a magnitude near the boundary. If the observer, in fix- 

 ing the upper and lower boundaries of the light, aims at equality of 

 brightness in the opposite points through which he draws them, an 

 excess of absorption at the lower edge amounting to one fifth of a mag- 

 nitude will consequently set back the lower boundary about a degree. 

 This conclusion is of no value as a representation of the true process 

 of observation, but it indicates that the differences of atmospheric ab- 

 sorption required to effect a sensible displacement of the boundaries 

 are probably not large. In this way, it appears to me to strengthen 

 the probability that absorption is an important, and perhaps the only, 

 cause of the variations of the zodiacal light in latitude. It may rea- 

 sonably be anticipated that future observation will prove the diminu- 

 tion of the light from the middle towards the edges to be not uniform, 

 but much more gradual near the edge than at some other places. In 

 this case, the effect of absorption would ^it first be considerably greater 

 than has just been supposed. But it would be useless, in the absence 

 of direct evidence upon the subject, to adapt an imaginary gradation 

 of light to the results obtained by observation for the displacement of 

 the zodiacal cone. At present, we cannot even assert with confidence 

 that the observers have aimed at equality of brightness between oppo- 

 site points iu the boundaries which they have drawn. 



