40 A. W. Wright—-Spectrum of the Zodiacal Light. 
trum to be measured, it was necessary to employ some method 
of fixing scale-positions without recourse to the use of a light, 
and, in fact, it was found indispensable to exclude all artificial 
lights from the raiet bonvore of the instrument during the 
observations, and for some time previous, on account of “their 
effect in blunting the pcnsthilits of the retin 
e means employed in securing these ER peeves 
the principal modification introduced, and were as follow 
The short joint connecting the eye-piece wiged the sliding take 
of the telescope was removed, and another substituted it, 
which was pierced on each side with a narrow openin In 
this, and perpendicular to the axis of the. tube, was ae 
ed rectangular frame, in which are two slides, or dia- 
phragms, moved by fine screws passing through the ends of 
to traverse the entire field. The inner ends of the slides termi- 
nate in sharp, saps edges perpendicular to their line of 
in place, they are so adjusted as to be accu- 
rately in the ee of the eye-piece. They are thus seen pro- 
jected upon the scale, and sharply defined, when the latter is 
illuminated. By turning the tube the terminal edges are made 
accurately parallel with the lines of the spec ectrum. As these 
are somewhat curved in the passage of the rays through the 
prism, the scale was placed in the middle.of the field, and the 
measurements made from the middle point of the slides, or, in 
some instances, the ends of the slit were covered with its of 
paper, reducing the spectrum to such a breadth that the curva- 
ture was small enough to be neglected. 
Before making an observation, the instrument was carefully 
adjusted, and the scale so placed that the division-mark 5 
coincided with the more refrangible edge of the sodium line, 
inasmuch as only one of the sides of the slit is pier ea and 
- opening it we mers widens toward the red end only. For 
e same reaso n fixing the limits of the spectrum, the 
eesti of the ‘slit must be added to the scale-number on the 
less refrangible side, in order to reduce the dimensions of the 
spectrum to what they would be with a linear slit, The situa- 
tion of any point in the spectrum, however faint, could be 
cod with a good degree of accuracy by moving the 
diaphragms up to it, and then, on illuminating the scale, read- 
ing its posee n by the scale-numbers. In the case of well- 
* like "that of the auroral spectrum, the error is 
not greater than half a division of the scale, and, with care- 
fully repeated determinations, may be made considerably less 
In making the observations upon the zodiacal light, the same 
method was pursiied as in the investigation upon its polariza- 
