990 
PROrESSOR J. N. LOCKTER ON THE PHOTOGRAPHIC 
In this case, however, the spark spectrum, and not the “ arc ” spectrum, was 
photographed in juxta-position wuth the solar spectrum. McClean has not tabulated 
the wave-lengths of the lines exhibited in his photographs, but, from a set of 
enlargements he has had the goodness to send me, I have been able to determine 
them with reference to lines in the solar spectrum, in precisely the same way as in 
my own photographs. The results of this comparison in the region k 3900-X 5740 are 
shown in a separate set of tables. 
So fai' as I am aware McClean has not published any detailed account of the 
apparatus which he employed. 
Exijlanation of the Tables. 
All the lines in the arc spectrum of iron shown on the photographs have been 
mapped in their exact positions with reference to the lines in Professor Powland’s 
photographic map of the solar spectrum (first series). In the following tables, how¬ 
ever, the wave-lengths have been transferred to the scale of the second series. The 
first and second columns give respectively the wave-lengths and intensities of the 
lines photographed at Kensington, and the third and fourth those estimated by 
Kayser and Punge. The lines tabulated in the first column have been taken 
from the enlarged photographs, of which reproductions are given in the Plates 
accompanying this paper (Plates 12, 13). At placed after the wave-length of 
certain lines denotes that corresponding lines were observed by Thalen. The first 
two figures of the wave-lengths are inserted only at the top of each column and 
where they undergo a change. In each case the scale of intensity used is such that 
1 represents the strongest and 6 the weakest lines. The last column is reserved for 
remarks on the probable origins of lines (generally faint) which appear in the spectrum 
of iron, but which are possibly due to impurities. An origin stated without further 
comment signifies that there is a long line at that wave-length in the spectrum of the 
substance named; but where a ? is added the coincident line of the substance is not 
one of the longest. Coincidences with lines of cerium have not been considered. 
III. Discussion of the Pesults. 
Im'purities in Electrolytic E'on. 
The impurities as indicated (by the method previously referred to) in the appended 
tables may be summarized as follows :— 
Impurities undoubtedly p)Tesent. —Mn, Ni, Cr, Co, Ba, Sr, Ca, Cu, Ti, Di. 
Impuritiesp)robably p)resent. —Zr, U, Pu, La, Er, Mo, Zn, V, W, Os, Al. 
The evidence for the elements in the first column depends upon the work of others 
