174 O. N. Rood on a Secondary Spectrum. 
with its aid the distances between the lines in the original, or 
in the secondary spectra, were measured. The collimating 
telescope was arranged so that its optical axis could be directed 
across any desired portion of the supporting brass plate. 
these parts were capable of being independently leveled. 
In the experiments described in this paper only two prisms 
~ were used, their angles being approximately 60° ; one was made 
of ordinary flint glass, the other was filled with bisulphide of 
bon. Mr. Rutherfurd kindly supplied me with two of his 
superb diffraction gratings ; both were ruled on glass, the finer, 
No. 1, having 6480 lines to the inch, while the other, No. 2, in 
the same space contained 4820. The equality of the intervals 
between the lines was equal, if not superior, to that in the grat- 
ings of Nobert which have come under my observation ; the 
spectra were brilliant. 
The lines of the solar spectrum are not adapted for the study 
of the arrangement of the colors of the secondary spectrum, 
owing to its peculiar structure, consisting as it does of a spec: 
trum bent back or folded on itself and in some portions very 
much contracted; the identification of solar lines becomes, hence, 
all but impossible in the majority of cases. Accordingly I se- 
lected a number of chemical lines of easy identification, which 
in point of fact proved to answer quite well. They were as 
follows: the potassium line at the extreme red end of the spec: 
trum (Ka), the red lithium line (Li), the red hydrogen line (Ha), 
the sodium line (Na), the thalium line (Th), the blue hydrogen 
line (Hf), and in some cases the violet hydrogen line (Hy 
For the most refrangible portion of the spectrum I employed 
two mercury lines, viz: the final line, which is somewhat less 
refrangible than the solar line H, and which, is marked by Hug: 
gins with the number 5158; also the line next but one to th 
4775, Huggins. This last line is a little less refrangible than 
the violet hydrogen line (Hy), and is far more brilliant. These: 
lines I have indicated in the spectra given below by the symbols 
Hgt and Hg’. The most fehetantte of the caesium and ru- 
bidium lines were also sometimes employed to divide up, for 
purposes of measurement, the rather extensive tracts in the biué 
and violet portions. The potassium line was best obtained | i 
a flame of common illuminating gas somewhat under-fed wit 
oxygen, so as to avoid danger of fusing the platinum wire; the 
salt selected was the nitrate. The hydrogen lines were from 
spectral tube containing the rarefied gas, the thalium line from 
a mixture of the sulphate of thalium with carbonate of sod®, 
which ensured the necessary presence of the line due to the lat- 
ter. The mercury lines were generated by the aid of platinum 
and mercury electrodes, in connexion with an induction coil and 
Leyden jar. All these lines were plainly visible in both the 
