704 Scientific Proceedings, Royal Dublin Society. 
The range of spectrum examined is extensive, and lies between 
wave-lengths 6000 and 3200, and lines capable of being photo- 
graphed were carefully observed. It will be noticed that in the 
tabulated statement on next page, after the symbol of the ele- 
ment, an index number from 1 to 9 shows the relative strength of 
the lines, the figure 1 indicating the weakest, and 9 the strongest 
appearance of the same lines in the several spectra. In the case 
of the principal constituents of the meteorites this is unnecessary, 
as lists of the ines measured are given, but where only two or 
three lines are visible, the substances being in minute proportions, 
the index figures serve a very useful purpose. Symbols in italics 
indicate traces only. The tabulated statement clearly shows 
the elements which are present, with variations in the composition 
of the different specimens. 
The lines were identified by measurements made on the photo- 
graphs, and wave-lengths were obtained from curves based on 
Rowland’s Standard, the particular wave-lengths quoted being 
those of Kayser and Runge. Lists of the lines of iron, nickel, 
cobalt, sodium, potassium, rubidium, gallium, copper, silver, and 
lead are given. A calcium line was recorded in all specimens, and 
the bands of magnesia in all the stony meteorites. 
We were at first inclined to doubt whether calcium, sodium, 
and potassium, were really constituents of the meteoric irons, but 
the lines of the alkali metals, which are very weak, were proved to 
belong to the metallic iron by burning it without having recourse 
to a support of any kind, and thus the spectrum observed was the 
same as that obtained by burning filings of the metal on supports. 
A portion of the file used upon the metal was also burnt, and 
this showed a composition differing from that of the meteoric 
irons, since it contained manganese, but no nickel, cobalt, or 
gallium. 
