ARC SPECTRUM OF ELECTROLYTIC IRON. 
989 
only when the development is nearly completed that dim light should be admitted 
through ruby glass in order to look at a plate. 
It was originally proposed to use a Howland grating for the production of the 
spectra, especially for the less refrangible portions, where the dispersion is so small. 
An attempt was made to carry this into effect, but it was soon found that the limited 
amount of electrolytic iron at my disposal was quite insufficient to permit the long 
exposure involved, so that prisms were used throughout the length of spectrum 
photographed. Even when this was done, the want of electrolytic iron prevented 
the proper exposure being given to the region from D to C, so the photograph 
of this portion of the spectrum does not contain so many lines as it would have 
done had more material been available. 
Heproductions of the photographs employed in this inquiry are given in Plates 12 
and 13. 
Reference to the Observations of Messrs. Kayser and Runge. 
Since my paper was commenced two important memoirs on the spectrum of iron 
have been published. Professors Kayser and Purge, of the Hanover Technical High 
School, have investigated the arc spectrum of iron between the wave-lengths 2230'01 
and 6750'36 (on Rowland’s scale), and compared their positions with those given by 
Thalen, Cornu, and Vogel.* The wave-lengths of the lines in their photographs 
were determined by micrometric measures, a number of standard lines being used to 
construct the interpolation curves. The electric arc was ^^I’oduced between poles of 
wrought-iron, 1 centim. in diameter, and the spectra were obtained by means of a 
Rowland grating having 14438 lines to the inch. No attempt was made to eliminate 
lines due to impurities in the iron although metal of the ordinary commercial quality 
was used. From this it will be seen that Professors Kayser and Runge have gone 
over very nearly the same ground as I have. But there are one or two important 
differences in our method of work. They determined wave-lengths by micrometer 
measures, my positions have been obtained by direct comparison with the solar 
spectrum. Their object was to investigate the spectrum of iron of ordinary purity, 
so lines due to impurities are not distinguished from those of iron. My idea has 
been to obtain the spectrum of the purest iron, and I have indicated in the tables, 
the lines ‘possibly having their origin in foreign substances. I have compared the 
lines given by Professors Kayser and Runge with those shown in my photographs 
from X 3900 to X 6500. The results are contained in the tables. 
Reference to McClean’s Work on the Spark Spectrum, 
Another paper on the iron spectrum was recently communicated to the Royal 
Astronomical Society by Mr. F. McClean.! 
* ‘ Abh. d. Akad. d. Wiss. zu Berlin,’ 1888. 
t ‘MontUy Notices, R.A.S.,’ vol, 52, November, 1891. 
