Flame Spectra observed at Swedish Bessemer Works. 93 



appears very intense and constant with a very Lot charge, and then 

 the proportion of silicon is about 1 per cent, at its appearance. The 

 last tenths per cent, of the silicon are not oxidised until all the 

 carbon is burnt, the final product receiving a very high temperature. 

 Here the charge is thought to be too hot. With a normal charge 

 the proportion of carbon begins to decrease upon the appearance of 

 the blue flame, but as long as this is visible (J to 1^ minutes) the 

 silicon decreases more rapidly and when the flame has become 

 white, one may be quite certain that there is no large percentage of 

 silicon left in the bath. The "blow" lasts about 15 minutes. With 

 extra cold charges the blue flame appears almost immediately after 

 the erection of the converter and it becomes white very soon. 



I did not, however, succeed in observing a spectrum of the blue 

 flame under these conditions. During my stay at Ulfshyttan the 

 charge was never too hot but was always normal, and the blue flame 

 appeared to the naked-eye more milky than at Langshyttan, but it did 

 not give any distinguishing spectrum. A Bessemer spectrum was, 

 however, observed both during the "boil" and during the decarbon- 

 isation, and this differed essentially from the Bessemer spectra 

 already described. This is a matter of course, as the silicon here 

 takes the place of the manganese. 



Only the most brilliant manganese-oxide bands are to be found 

 in this spectrum. The wave-lengths of the bands and lines appear 

 in the table (p. 94). 



This spectrum has quite a different appearance from that of the 

 ordinary Bessemer spectra. In the red field there are two strong 

 bands which are always very distinctly visible. Then there are two 

 lines, one of which (X, = 6088) seems, however, to coincide with the 

 •stronger edge of a faint band shading away towards the left (red) 

 and on which the other line (X — 6108) is to be found. The 

 yellowish-green field is chiefly remarkable for two bright bands, 

 which seem to be separated by an absorption band. Between those 

 bands and the sodium line some bright lines and a band are to be 

 found and, in addition, other bands and lines which, however, rarely 

 appear. 



To the left of the bands above mentioned there also appear on the 

 back-ground of continuous rays a narrow band and several distinct 

 lines or possible narrow bands. 



The order in which the lines appeared was somewhat different in 

 different " blows." The Na a - and Lia-lines invariably appeared first, 

 and then the bands 5611 and 5580. Then on one occasion 5415 and 

 5377 were observed before the other lines, while at another time 

 5642 appeared after 5611 and 5580, after which the bands and lines 

 presented themselves in the following order: 5814, 5784, 5867; 

 5524, 5377, 5258, 5355, 5312, and lastly 5415, 5388, 5159, 5099. 



