1889—90.] Prof. C. Michie Smith on the Zodiacal Light. 143 
adopted the following, which is probably not new, though I have 
never seen it described. The photographed scale is removed from 
the end of its tube, and is replaced by a small slit with adjustable 
jaws. The tube is then mounted on a vertical axis concentric with 
a fixed horizontal, graduated circle. The tube is rigidly attached to 
the arm carrying the verniers of this circle. When the position of 
any line has to be determined, the slit is adjusted so as to admit 
just sufficient light to render its reflected image visible, and the tube 
is rotated till this image coincides with the line. It is of course 
easy to adapt Prof. C. P. Smyth’s illumination to this arrange- 
ment. 
The attempts so far made to photograph the zodiacal light 
spectrum have been total failures, but I hope that the use of plates 
with a greater range of colour sensitiveness may yet prove successful. 
The spectrum as usually seen is extremely faint, so that the 
observer ought to prepare himself for his work by remaining in a 
dark room for at least ten minutes before he attempts to make 
observations, but if this precaution be taken the spectrum can be 
seen on any clear night when there is no moonlight and neither 
Jupiter nor Venus is very near the place of observation. As 
ordinarily seen, very little colour can be distinguished in the spec- 
trum, but under favourable conditions a distinct tinge of red can be 
observed. Except on a few occasions the spectrum, as I have seen 
it, was continuous and quite free from bright lines, but on several 
nights in 1883 I saw what appeared to be a bright line. The 
following extracts from my note-book give all the cases of this 
appearance. 
1883. March 7th . — Spectrum in direct vision spectroscope distinct 
up to at least 45° altitude: apparently continuous, but with part 
more brilliant than the rest. 
March 8th . — Spectrum clearly visible. The most notable feature 
is the rapid variations in brightness. There seems to be either a 
bright band or a part of maximum brightness somewhat like that 
seen on the more refrangible side of the D line towards sunset. 
(Observation made with long collimator and glass prism.) 
March 28 th, 7 h 20 m to 7 h 50 m p.m. — Short collimator. Light 
faint and very diffuse : no clouds : eight stars easily visible in Pleiades. 
Spectrum flickering and faint, but shows almost certainly a bright 
