STEEL 



907 



The following series of analyses, by the same chemist, f metal taken at different 

 stages of the blow, show very distinctly the gradual removal of the carbon along witli 

 the silicon : 



The difference in the amount of copper, which is much larger in the Styrian steel 

 than in that from Dowlais, is to be attributed to the fact that the pig-iron used in 

 the former is entirely smelted from spathic ore, while in the latter only the spiegel- 

 eisen is due to that source. Copper pyrites, in small quantity, is almost invariably 

 present in spathic carbonates, and however carefully they may be washed and weathered, 

 some copper, as a general rule, is reduced and passes into the iron in the blast- 

 furnace. 



The progress of the conversion of the charge can be controlled to some extent by 

 observing the spectrum given by the flame with the spectroscope ; and more particu- 

 larly the moment of complete decarbonisation may be determined with considerable 

 accuracy, especially if the flame be bright and free from smoke. The spectrum pro- 

 duced when the combustion is most active is characterised by groups of numerous lines 

 in the yellow and green portions, that of sodium being the most prominent and the 

 first to appear among the former. There is also a well-defined group of lines in the 

 blue field, and under the most favorable conditions the violet and red lines of potas- 

 sium and lithium, together with an extra violet line accompanying the former are seen. 

 For this, however, an instrument of great defining power and an extremely bright 

 flame are essential. When the metal is completely decarbonised, the yellow and green 

 lines disappear, bxit the sodium is persistent, sometimes even after the tipping of the 

 converter. On the addition of the spiegeleisen, the whole of the lines reappear with 

 great brilliancy. When there is much manganese in the cast iron employed, as is the 

 case in Styria, the use of the spectroscope is difficult, owing to the brown smoky 

 character of the flame. 



At Seraing, it has been found that the disappearance of the dark absorption-bands, 

 which alternate with the bright lines, can be more readily determined than the latter, 

 which often reappear after their apparent extinction, and is therefore to be preferred 

 as admitting of much closer and easier observation. 



The exact chemical character of the spectrum of the Bessemer flame has not as yet 

 been made out, although it has been the cause of considerable controversy, there being 

 two different opinions as to its origin. One of these supposes the lines to be due to 

 carbonic oxide, and their cessation to the complete combustion of the carbon ; while 

 the other considers that they are mainly produced by manganese, and that their sudden 



