238 Mr. G. A. Hemsalech on the Line Spectrum of 



assumed that, even with the electric current passing through 

 the graphite rod, the temperature of the latter never reached 

 that of the flame. This was borne out by direct measure- 

 ments with a pyrometer, which indicated 1100° C. without 

 and 1300° C. with the electric current passing through the 

 rod. Hence it would appear that the role of the graphite rod 

 in these circumstances is simply to shield the lower portion of 

 the flame from undue loss of heat by convection. The gain 

 in temperature, if any, would therefore be due to preservation 

 of the heat developed by the flame. 



Cooling of the flame was accomplished by passing cold 

 water through a brass tube 7 millimetres in diameter, which 

 was held in the mantle in place of the graphite rod. The 

 effect of cooling on the general appearance of the flame 

 charged with iron vapour is very apparent. The flame shrinks 

 appreciably and becomes very pale, although the cones seem 

 still fairly bright. The spectrum suffers a general and con- 

 siderable decrease in intensity — not only the continuous 

 ground in the visible part, but also the carbon bands and all 

 three classes of iron lines being similarly affected. There is, 

 however, no evidence of a selective action of cooling on any 

 of the iron lines. 



§ 13. Influence of Velocity of Gas Mixture issuing 

 from Burner. 



Dr. King has suggested that the absence of lines in the 

 cones of the high temperature flames, such as the oxy- 

 hydrogen or oxy-acetylene flames, might possibly be due 

 to the higher gas velocity in these flames as compared with 

 the Bunsen flame *. As a result of this high velocity, the 

 thermal forces at work in these flames are unable to exert 

 their influence for a sufficiently long time. This suggestion 

 is, of course, based on the assumption that the characteristic 

 cone lines observed in the Bunsen flame are due to thermal 

 and not to chemical actions. Although the chemical origin 

 of these lines had already been placed beyond doubt by the 

 results of earlier investigations, a direct experimental test 

 appeared to be called for. This was accomplished by passing 

 the gas mixture through the burner under such pressure as 

 to increase the lengths of the cones to at least ten times their 

 original size. The spectrograms obtained in this way ex- 

 hibit a general diminution in the intensities not only of the 



* S. A. King, Astrophys. Journ. vol. xxxvii. p. 276 (1913). 



