PROFESSOE W. N. HARTLEY AND MR. H. RAMAGE ON THE SPECTRA 
49fi 
It was found from the inquiries made of the officials that whilst all three vessels 
were receiving similar charges of metal and lime, the lining of vessel No. 4 was 
almost new, and a new bottom had been fitted to it shortly before the time of 
observation. About forty charges had been blown in vessel No. 3, and about 
seventy in No. 2. The intensity of the line diminishes therefore with the age of the 
lining. 
The tar in the lining is not wholly decomposed when the lining is worn out, hut 
the rate of the decomposition and the quantity of the volatile products must be less 
than in a new lining, and it appears to us that it is the volatile products of the 
decomposing tar which cause these lines to appear in the sjDectra of the flames. In 
the “ acid ” process the flame drops as soon as the carbon is oxidised ; the hands and 
lines in the spectrum Immediately become diffuse, and the spectrum presents more 
the appearance of a continuous band of rays. According to Sir I. Lowthiax Bell, 
carbon monoxide is much more stable in presence of iron at high temperatures than 
carbon dioxide. Snellts (‘ Chem. News,’ vol. 24, p. 159) has shown that the proportion 
of the former oxide exceeds that of the latter in the gases issuing from the converter 
in the “ acid ” process during the latter two-thirds of the time, and towards the end 
there is no carbon dioxide present; so that all the carbon in the volatile products 
from the tar will exist in the flame during the “ over-blow ” as carbon monoxide. 
The hydrogen in these products and in the water vapour in the blast will pass out of 
the converter in the elementary state. These reducing gases acting on the com¬ 
pounds of manganese in the slag reduce some of the metal, which then j^asses off as 
vapour. The combustion of the hot gases, carbon monoxide, and hydrogen outside 
the vessel will give a hotter flame, and the increased temperature, together with the 
reducing gases in the flame, will account for the increased intensity of the manganese 
spectrum. 
There are two points regarding the flame of the “ over-blow ” which require expla¬ 
nation. First, the great brilliancy of the flame during the latter part of the blow ; 
secondl}^ the absence of the spectra of iron and calcium, and the feeble character of 
the spectra of potassium, sodium, and manganese. 
Immediately the carbon is burnt out of the iron the flame drops to about one-sixth 
of its height; its luminosity also diminishes, and the quantity of fume increases. 
There can be no considerable variation in the temperature of the fluid charge during 
the few moments in which this great change takes j^lace ; the gases will therefore 
leave the converter at about tbe same temperature tbrougbout this time. 
It is probable that the impurities in the iron are oxidised chiefly by secondary 
reactions, but a portion of them is oxidised directly by the oxygen of the air. The 
volume of the iron in the converter is veiy much greater than the volume of the 
maiiganese, and yet almost the whole of the latter is oxidised and passes into the 
slag in the first few minutes of the “ blow.” The fact that the flame maintains its 
size until the sudden drop confirms this hypothesis. If this hypothesis is true the 
