490 
PiiOFESSOR W. N. HARTLF.Y AXT) MR. H. RAMAGE OX THE SPECTRA 
Variations in tlie Intensity of Iron Lines in Spectra from various Sources. 
WaA’e-lengths. 
Intensity. 
Bessemer flame spectra. 
Oxyhydrogen 
Identified 
in solar 
spectrum. 
Solar. 
Spark. 
Arc. 
Oxy¬ 
hydrogen 
flame. 
CO + 0 
Crewe. 
Middlesbrough. 
flame. 
flame. 
Temperature. 
I. 
II. 
III. 
Low. 
High. 
4063-75 
4063-759 
20 
10 
10 
3 
1 
1 
o 
3 
1 
o 
4071-92 
4071-908 
15 
10 
10 
2 
1 
1 
o 
21 
T 
o 
4132-24 
4132-235 
10 
8 
10 
1 
1 
2 
1 
4144-06 
4144-038 
15 
7 
10 
1 
1 
o 
11 
4202-20 
4202-198 
8 
9 
10 
1 
1 
*4216-39 
4216-351 
3 
2 
6 
1 
1 
2 
4 
4250-93 
4250-945 
8 
8 
10 
1 
1 
4271-93 
4271-934 
15 
10 
10 
4 
2 
1 
4 
1 
•/ 
1 
4308-06 
4308-OSl 
6 
10 
10 
4 
0 
1 
4 
1 
X 
2 
4325-94 
4325-939 
6 
10 
10 
4 
1 
1 
4 
1 
. . . 
1 
2 
*4376-03 
4376-107 
6 
4 
8 
0 
o 
3 
5 
4 
2 
o 
O 
4383-71 
4383-720 
15 
10 
10 
8 
4 
3 
6 
4 
1 
3 
4404-94 
4404-927 
10 
10 
10 
31 
2 
0 
4 
2 
1 
1 
4415-29 
4415-293 
8 
8 
10 
1 
... 
... 
1 
1 
*4427-46 
4427-482 
5 
3 
8 
2 
2 
3 
5 
5 
2 
31 
*4461-78 
4461-818 
4 
3 
6 
1 
1 
21 
4 
2 
1 
2 
*4482-34 
4482-338 
5 
4 
8 
1 
11 
3 
1 
1 
1 
*4489-86 
4489-911 
4 
1 
4 
1 
2 
... 
1 
Wlien the intensity is not stated in these columns the lines are either absent or exceedingly feeble, so 
that no value could be attached to them. 
T. Plate 17, spectrum (8). 11. Plate 14, spectra (5) and (6), and Plate 23 (2.) III. Plate 26, spectrum (2). 
A line at 4045 was hidden by overlying potassium lines. Two lines less refrangible 
than the above appear in seven of the Crewe spectra on Plate 8, and their intensity 
is 1. 
The lines vrith wavm-lengths 4215’0, 4375T9, 4426'7, 4461'5, and 44817 ma}’ all 
be classed as strong lines in the Bessemer flame spectra. In the spectrum of the flame 
during the earlier period of the blow they are practically the strongest lines of the 
above series. As the blow proceeds—that is to say, as the temperature increases — 
other lines, notably those with wave-lengths 4270'6, 43067, 4324'3, 4383‘3 (this 
especially), and 4405'0, become stronger, while those just indicated are not 
strengthened in the same proportion. 
The series of photographs taken at the Crewe works of the London and North- 
By interpolation. The other lines are coincident with strong lines in the spark spectrum. 
