Flame Spectra observed at Swedish Bessemer Works. 81 



however, from the Langshytte spectrum. The line 6252 in the 

 group a was not present; the two faint lines in the group /3 and the 

 group e appeared very indistinctly, and no new lines or bands were 

 perceived. 



The Nykroppa Spectrum. (Fig. 1, spectrum II.) 



The wave-lengths of the lines and bands are given in the following 

 table. They are arranged in the same manner as in that of the 

 Langshytte spectrum. 



This spectrum has a different aspect in the region of red rays from 

 that of the Langshytte spectrum, which, as I have mentioned already, 

 only consists of three red lines on a bright background. Here the 

 group a consists of a faint continuous spectrum which extends from 

 the Li line, and gradually becomes more conspicuous and splendid 

 towards the more refrangible rays. The band 6255 — 6211 is defined 

 on the less refrangible edge by this continuous spectrum, and on the 

 more refrangible by a dark background extending to 6028 on which 

 appear the line 6174 and the band 6106 — 6079. I cannot state 

 positively w r hether these bands are really bright bands or if they 

 only form part of the continuous spectrum, and are flanked by 

 absorption bands, especially as the brightness of these bands is not 

 greater than that of the continuous spectrum before the " boil." 



From 6028 the splendour of the background gradually becomes 

 more intense towards the ISTa line.- On that background there 

 appear sometimes the lines 5972 and 5023. They are, however, of 

 short duration. 



The group /3 shows this difference, that the continuous spectrum 

 extends behind /3 2 to 5568 before we notice the dark shade which is 

 its usual termination. In addition, some lines appear that were not 

 observed at Langshyttan. 



7 and 8. contain respectively each one line on the dark background 

 near the preceding group, which lines were not observed at Langs- 

 hyttan. The group e seems sometimes to be only a bright nebulo- 

 sity, but gradually there appear four very distinct bands. 



The order in which the different bands appeared could here be 

 better observed than at Langshyttan, where the spectrum of the blue 

 flame passed into the ordinary Bessemer spectrum. Just when the 

 "boil " begins, a faint yellow line appears very distinctly on the bright 

 continuous background. Na a (5895), and then the second (5889) of 

 the Na lines, and at the same time the Lialine, then almost simultane- 

 ously the red band of the group a and the green ones j3 h 7 b Si, which 

 remain during the whole " blow." Lastly appear the lines 5972 and 

 5923 and the bands of the group e, these are also the first to 

 disappear. The disappearance of the bands is generally contrary to 

 the order of their appearance. 



VOL. lix. G 



