OF ARTS AND SCIENCES. 159 



Among the facts recorded with regard to the zodiacal light is one 

 which independently suggests the probability that the observation of 

 the boundaries is influenced by atmospheric absorption, althoufh it 

 gives no information respecting the extent to which it may displace 

 them. This fact is the customary distinctness of the lower edge as 

 compared with the upper. It is mentioned by Schmidt in his general 

 description of the zodiacal cone, and also later (S. 15, 27), and is 

 repeatedly noticed by Heis and Weber (H. 4, 5, 7, 11, 16, 20, 33, 37, 

 39, 44, 46, 53, 57) ; also by Brorsen, Schmidt, and Groneman (H. 15, 

 27, 58). Neumayei''s remark, that the southern boundary is deter- 

 mined with difliculty (H. 32), derives special interest from his south- 

 ern station, where this boundary is the upper one. Jones has a few 

 similar observations in the southern hemisphere (Jj. 568, 572, 586, 

 618), relating chiefly to the " Stronger" Light. He occasionally found 

 the "Diffuse" Light ill defined on its lower side (.Jj. 290, 330, 352). 

 Heis, on some occasions (H. 22, 23), found that the usual compara- 

 tive sharpness of the lower boundary was wanting. Lewis states as a 

 general rule that the southern side is more sharply defined and more 

 nearly parallel with the ecliptic (Ls. 438). He adds the interesting 

 remark that the axis of greatest brightness lies south of the axis of 

 symmetry. It is obvious that the increase of atmospheric absorption 

 with zenith distance will accelerate the diminution of light near the 

 lower side, and retard it on the other, which must have some effect in 

 makintr it easier to define the lower boundary, although we could not 

 be confident, without the support of direct observation, that the effect 

 would be considerable enough to attract attention. 



A few additional results from the tables of the Appendix acquire 

 some significance on the hypothesis that atmospheric absorption seri- 

 ously affects the observed position of the zodiacal light, but are hardly 

 to be regarded as strengthening that hypothesis, which must rest, for 

 the present, on the facts already set forth. The change in the " lati- 

 tude of the axis " consequent upon an increase of 30° in elongation is 

 here called "displacement," for the sake of brevity, and is considered 

 positive when the north "latitude of the axis" increases with the 

 elongation. This displacement has been computed only for those ob- 

 servations in which the position of the light was determined at each 

 of two elonsrations. The number of these observations is often too 

 small for any conclusive result, and possibly they are not worth study. 

 But those of Jones, Heis, and Weber seem to me to be numerous 

 enough to deserve attention. The result I find from them is that the 

 axis of the cone, as a rule, leans towards the ecliptic, so that its lati- 



