He was the first, in 1867, to examine the spectrum of the aurora borealis, and detected and measured the characteristic bright line in its yellow green region; but he was mistaken in supposing that this same line, which is often called by his name, is also to be seen in the zodiacal light.
The only source of doubt as to the validity of the conclusion that this is really the zodiacal light arises from the possibility that, after the close of the ordinarily recognized twilight, there may be a faint illumination arising from the reflection of light by the very rare upper atmosphere, shown by the phenomena of meteors to extend some hundred miles or more above the earth's surface.
The problem of separating a possible effect produced in this way from the zodiacal light proper may seem to offer some difficulty.
But the few observations made show that, after ordinary twilight has ended in the evening, the northern base of the zodiacal light extends more and more toward the north as the hours pass until, towards midnight, it merges into the light of the sky described by the two observers mentioned.
He also describes the moon as adding to the zodiacal light during her first and last quarters, a result so difficult to explain that it needs confirmation.