J. M. Silliman— Examination of the Bessemer Flame. 301 
covered a black band which he considered an absorption-band, 
ather. 
nitrogen is most abundant, and it is for this reason that the first 
investigator, Roscoe, expressed himself as confident that the nu- 
merous lines of the spectrum were caused by this gas, although 
he could obtain no coincidence. 
Brunner* states that “no part of the Bessemer spectrum is 
ever visible in the flame when the converter is heated for the 
first time after being re-lined, but that when the lining is not 
new, Lielegg’s group of green lines (CO7) appears in the spec- 
trum, which then contains also the lines of potassium, sodium, 
and lithium.” From which he concludes that this spectrum is 
not to be identified with carbonic oxyd, but must be produced 
by other constituents of pig-iron. Others state that the Bessemer 
spectrum is sometimes visible while the converter is being 
heated after a blow. I made an observation of the flame from 
the converter while it was being heated the first time after being 
re-lined, and obtained with great distinctness the potassium, 
lithium, and sodium lines, but have not under any circum- 
stances detected any other lines while the converter was being 
reheated. 
never observed it, but Dr. Wedding, who has summed up the 
observations of others, states that he has repeatedly seen it. 
Its position is at 185$°. 
he instrument used in my investigations was constructed 
by Alvan Clark of Cambridge, and consists of an equiangular 
flint-glass prism, in a metallic box, into the sides o which at 
the requisite angles are screwed an inverting telescope with a 
* Van Nostrand’s Hclectic Eng. Mag., vol. i, page 508. 
