OF FLAMES RESULTING FROM BESSEMER PROCESSES. 
493 
The line 4642 was visible in all the above spectra. It was stronger in (2) than in 
(3), but strongest of aU in (4). 
The density of (5) was much greater than in any of the other photograjDhs, but the 
intensity of the line was not proportionately increased, for there appeared to be very 
little difference between its intensity in Spectra (4) and (5). 
Provided the temperature is sufficiently high to cause the emission of a ray with 
wave-length 4642, its brilliancy or intensity of chemical action is increased by 
diminishing the quantity of vapour or vapour-pressure in the flame, and thereby 
permitting the molecules so great a freedom of internal motion that it becomes 
possible for them to vibrate in the particular manner which causes the emission of 
this ray. 
Taking a survey of all the plates upon which the line 4642 has been observed, and 
of those in which it is absent, and also having regard to the conditions under which 
the sj^ectra in each case were produced, it is not so much reduction of temperature as 
reduction of quantity of vapour in tlie flame which increases its intensity. 
The lines next in order are two of potassium with wave-lengths 5112 and 5098, 
which occur also in the oxyhydrogen flame, the arc, and spark spectra of potassium. 
In the case of a line with wave-length 5112 it is probably intensified in some of the 
Bessemer flame spectra by a closely adjacent iron line, wave-length 5109, observed in 
the oxyhydrogen spectrum of iron. 
About wave-length 5269 there is an iron line and at the more refrangible edge a 
manganese band. The iron line wave-length 5268'9 is the strongest in this region in 
the oxyhydrogen spectrum of iron. 
At wave-length 5328 there is a line due in part each to iron and potassium ; at 
\ 5343 a potassium line is seen, and at X 5361 a line of potassium and the edge of a 
band of manganese. 
In the earlier part of the blow a diffuse band occurs about 5540, and towards the 
end of the first part of the blow it appears to be replaced by two sharp lines. The 
change is best seen in the spectra on Plate No. 14. The band is due to calcium, but 
the lines are not calcium lines, for in tlie sixth spectrum on this plate the violet 
calcium line X 4227 is much weaker than in the fifth spectrum, while the two lines are 
stronger. 
The sixth spectrum is remarkable for aii increase in the strength of the lines of iron 
throughout the whole spectrum, while the lines and bands of potassium and calcium 
are weaker. It is probable, therefore, that they are really iron lines. They are 
present in some Bessemer flame spectra photographed at Crewe in January, 1895, and 
occur in these only when the lines of iron are strongest. In the oxyhydrogen spectra 
of iron and its compounds the continuous spectrum is very strong in this region, but 
on very careful examination of a number of spectra we find a few in which traces of 
these two lines are present on the strong continuous spectrum. The relative intensity 
of the lines as compared with the other lines is, however, not equal to what obtains 
