Iron Vapour in Air-Coal Gas Flame. 239 



iron lines, but also of the carbon bands. In fact, the latter 

 seem to be relatively more affected, because many of the iron 

 lines, which with the ordinary low velocity flame are more 

 or less masked by carbon bands, stand out quite plainly with 

 the high velocity flame. 



A few of the groups of Class I. iron lines are, however, 

 less affected by the higher flame velocity and they, therefore, 

 appear relatively enhanced as compared with the majority of 

 the lines in the spectrum. To these exceptions belong the 

 following groups : — quartets 7 at 4376 and ft at 3856 ; 

 the triplet at 38G0 and a group at 3920. The behaviour of 

 all these groups is in striking contrast to that of the triplet 

 at 3720, which, for reasons already given (§2, c), may be 

 considered to consist of true temperature lines. This triplet 

 is relatively very feeble with the higher velocity, and we may 

 conclude that the increase in the gas velocity is accompanied 

 by an appreciable diminution in the temperature of the flame. 

 That the formation of nitrides proceeds almost normally at 

 the higher velocity is shown by the presence of Class III. 

 group at X4957, which shows the usual development. Also 

 the relative behaviour of Class II. lines is not sensibly 

 affected. 



From these results we may infer that an increase in the 

 velocity of the gas mixture influences the thermal actions in 

 the flame to a relatively greater extent than the chemical 

 actions. The abnormal behaviour of some of Class I. groups 

 can be best explained by assuming them to be to some extent 

 also sensitive to chemical actions, whereas such characteristic 

 temperature lines as constitute the triplet at 3720 are com- 

 pletely unaffected by them. 



It is thus evident that the higher velocities of the oxy- 

 hydrogen and oxy-acetylene flames do not account for the 

 complete absence in their cones of the lines of the supple- 

 mentary spectrum of iron. 



§ 14. Summary of Results. 



1. There is no connexion between the chemical actions which 



underlie the emission of the carbon bands and those to 

 which is due the characteristic cone spectrum of iron. 



§ 2, b. 



2. The relative behaviour of iron lines emitted by the mantles 



of various flames is such as to suggest their being in a 

 great measure controlled by the temperatures of these 

 flames. § 2, c & d. 



S2 



