220 Proceedings of tlic Royal Irish Academy. 



dently, whose accordant measures indicate its wave length to be ap- 

 proximately 5355. The more refrangible end of the spectrum was 

 extremely ill-defined, and the measures give it a wave length which 

 is probably much too great, being that of the poiat where the index 

 in the field of the eye-piece ceased to be visible, namely, 4640 seventh- 

 metres. 



Date.—'SmiQ 15<i, OJ''. 



Place of Observation.— Lat. 13° I^. Long. I7i° ^Y. 



Spectrum. — As on June 12"^, with the exception of the dark band 

 at 5355, which was not visible, with all attention. The other details 

 of the spectrum were visible with the usual distinctness, but the 

 measures of the bright line were unsatisfactory, indicating its wave 

 length to be 5720, instead of 5670, or 5680, as previously determined. 

 This untoward discordance may have been caused by an accidental 

 shift of the recording levers, due either to insufficient tightening of the 

 clamp which connects them with the telescopes, or, as is more pro- 

 bable, judging from the tested stability of the zero, to my having 

 mistaken the acute angle formed by the less refrangible edge of 

 the bright line with the visible edge of the triangular pointer for 

 the proper point of reference, namely, the apex of the triangle. 



On directing the instrument to any part of the sky, except that 

 occupied by the Milky Way, the same spectrum, micch diminished in 

 intensity, but otherwise recognisable, was seen as when viewing the 

 central regions of the Zodiacal Light. The same fact was observed on 

 many subsequent occasions in widely different localities, and with 

 another instrument. It is, perhaps, scarcely possible to attribute this 

 spectrum to any diffused reflection of the solar rays, for, at the hour 

 of observation, the sun's depression below the horizon was nearly 40°, 

 and, furthermore, the luminous appearance which yielded the spectrum 

 its greatest brilliancy retained its form and position among the stars 

 sensibly unchanged during the evening. 



The facts mentioned in the last words of the preceding sentence, 

 taken together with the apparently unique character of the spectrum, 

 appear to preclude the supposition that it was due to any terrestrial 

 aurora, the spectrum of which cliffers much from the Zodiacal Light 

 spectrum, as regards both its aspect and the positions of the lines 

 hitherto recorded, so far as they are known to me. 



A suggestion made many years ago, I believe, by Sir John Hers- 

 chel, to the effect that the observed extension of the Zodiacal Light 

 fi'om the sun indicated that the earth is probably immersed in the 

 outer portions of a luminous haze of lenticular form, symmetrically 

 disposed about the sun and having its principal section nearly co- 

 incident with the plane of the Ecliptic, seems to afford a possible 

 explanation of the diffusion of the light over the whole sky, indicated 

 by the spectroscope, and also of certain anomalies in the form of the 

 denser portion, hereafter to be recorded. 



