494 PROFESSOR W. N. HARTLEY AND MR. H. RAMAC4E OX THE SPECTRA 
in the Bessemer flame spectra. Failing to find corresponding lines in the spectrum of 
any other element, we have no hesitation, on the above evidence, in attributing them 
to iron. A line A. 5792 Avas identified in a similar AAmy. 
The line X 5803 is a potassium line, Avhich is seen as a strong line in the oxy- 
hydrogen spectrum of potassium. The line X 5833 is also present in the oxyhydi'ogen 
flame spectrum of potassium, and is, in some of the Bessemer flame spectra, intensified 
Ijy an iron line, 5834'5, observed in oxyhydrogen flame spectra. 
In the region of the spectrum just considered there are many bands, and these are 
almost wholly due to manganese or compounds of manganese. They coincide Avith 
bands in the oxyhydrogen flame spectra of manganese and its compounds; they are, 
in both series of spectra, all degraded toAvards the red, and the more refrangible 
edges are sharp. The measurement of the more refrangible edges only are gAen in 
the tables, and these measurements vary on different photographs as the intensities 
of the bands vary. Sometimes Avhen there is a strong continuous spectrum present, 
the sharp edge of a band has the appearance of a broad or nebulous line. The three 
more refrangible bands terminate about Avave-lengths 4561, 4402, and 4245, and near 
these positions there are lines in the Bessemer spectra Avhich, according to their 
intensities, modify the appearance of the edges of the bands. The line near 4556 is 
the crnsium line 4555, and in the other tAvo cases, iron lines of AvaAm-lengths 4404 and 
4251, 
Lines and Bands Less Refrangible than the D Lines. 
We haAm been able to identify the lines and bands less refrangible than the D lines 
(Avhich were photograplied Avhen using Lumiere B plates) only by direct comparison 
Avith other spectra. We have no lines of reference on the Middlesbrough j^lates from 
AA’hich to draAv a curve. A series of spectra AAms photograjfiied at CreAve in January, 
1895, on a plate stained Avith cyamine, and this includes both solar and Bessemer 
flame spectra. A curAm of Avave-lengths Avas draAvn and the AvaA'e-lengths of the 
unknown lines and bands in the Bessemer flame spectrum Avere determined from it. 
By direct comparison of the Middlesbrough spectra AAnth this plate, and by measure¬ 
ments of the oxyhydrogen flame sjDectra of the metals or compounds of lithium, 
manganese, and iron, Ave have identified the lines and bands in the Middlesbrough 
spectra. They Avere recorded in the analytical table of the second spectrum of Plate 
No. 26, and call for no special remark beyond the fact that the bands and lines are 
due chiefly to manganese, but are modified by bauds and lines of iron, and possibly 
by the orange lithium line 6108. 
The red lithium line 6708, and the' red potassium lines 6938 and 6911, do not 
appear on the photographs, but Avere seen by the eye obserA’ations. 
