of the Bessemer -flame. 89 



and of platinum 2000° C, it follows from these observations that 

 the temperature of the flame at the commencement of the blow 

 is below 1300°, but gradually rises, never, however, reaching 

 2000° C. 



This result is confirmed by the fact, that of the sodium-spec- 

 trum only the double line D is present. In the Philosophical 

 Magazine for August 1870, I showed that the sodium-lines 



1 5687 ma y ^ e em pl°y e( ^ as an index of temperature, since they 

 are present in the spectrum of any flame containing sodium the 

 temperature of which is high enough to melt platinum, but do 

 not appear at lower temperatures. The Bessemer-flame obvi- 

 ously contains abundance of sodium ; but this sodium double line 

 is absent. The lithium orange line also, which comes out at a 

 somewhat less temperature, is absent. 



The Bessemer-spectrum is seen more or less distinctly in several 

 other flames. The jet of flame which issues with the Spiegel 

 from the Spiegel cupola shows it brilliantly. This flame also 

 melts gold but not platinum. The Bessemer-spectrum is seen 

 also in the flame from the melting-cupola for the pig iron, in the 

 flame from the bottom of the blast-furnaces at work at Barrow, 

 and in the flame of the coke used in warming the cupola after 

 re-lining (which spectrum exhibits also the lines of sodium, 

 lithium, and potassium brilliantly), and in several other flames. 

 The flame obtained on adding the Spiegel, which gives the Bes- 

 semer-spectrum most brilliantly, is also incapable of melting 

 platinum, but melts gold. 



There are certain of the lines which seem to linger after the 

 rest, and which, when the blow has been carried rather far, are 

 occasionally seen alone (upon a continuous spectrum) after the 

 regular Bessemer-spectrum has disappeared. Some, but not 

 all of these, are iron lines. The following are those which have 

 been noticed : — 



5107, 5167, 5269 Fe, 5327 Fe, 5359, 5370Fe, 5395, 

 5433, 5443, 5453, 5462. 



The manganese present in the pig iron used at Barrow never 

 exceeds 0*6 per cent. T am indebted to Mr. Richards, chemist 

 to the Company, for the following numbers, representing the 

 mean composition of the pig iron and spiegel iron used : — 



