Experiments on the Flame Spectra of Carbon Monoxide. 217 



si Experiments on the Flame Spectrum of Carbon Monoxide." 

 By W. N. Hartley, F.R.S., Royal College of Science, 

 Dublin. Received March 6 —Read March 18, 1897. 



In the investigation of flame spectra at high temperatures, ' Phil. 

 Trans.,' A, vol. 185, p. 676, Part I, 1895, it was found that carbon 

 monoxide when burnt with oxygen gave faint lines or bands hi its 

 spectrum coincident with edges of some of the carbon bands which 

 have been observed in the flame of coal-gas. The experiments were 

 made simply with the view of ascertaining what might be usually 

 seen in a carbon monoxide flame, and looked for in a flame from a 

 Bessemer converter. The spectrum was photographed with a fairly 

 wide slit, and the exposure was of one hour's duration. It was com- 

 posed almost entirely of a continuous band of rays, but there were 

 very feeble indications of lines or the edges of bands which, after 

 careful inspection, being barely visible, were just capable of being 

 measured. Their positions lay between wave-lengths 5945 and 4249. 

 Other faint bands were seen at 6337, 6172, 4224, and 4183. These 

 all lie within the region of coloured rays and apparently belong to 

 carbon, but they were by no means considered an important feature 

 in the spectrum. 



The exact nature of the carbon monoxide spectrum bears upon the 

 researches of Professor Smithells (' Trans. Chem. Soc.,' 1894, p. 603), 

 so that it seemed desirable to decide whether the lines observed were 

 really due to carbon monoxide gas, or to some impurity which gained 

 admission to the material burned, or to the flame produced by its 

 combustion. A little consideration showed that possibly carbon- 

 aceous dust, in other terms, organic matter floating in the air, might 

 be concerned in the production of the feeble carbon bands, as the 

 exposure was a long one. Again, as the gas was conveyed from the 

 gas-holders and also from the oxygen cylinders through tubes of 

 india-rubber, there was some amount of uncertainty as to whether 

 the tubes had not contributed to the gases some minute quantity of a 

 volatile hydrocarbon. To obviate any such complication, a further 

 series of photographs was taken with a spectroscope slit of narrower 

 dimensions to obtain better definition. 



The gases were conveyed through patent flexible steel tubing 

 which had not been used for any other purpose. The flame was pro 

 duced by burning the carbon monoxide by means of a Deville blow- 

 pipe fitted with a polished platinum nozzle. Oxygen was conveyed 

 into its interior by a polished platinum jet. The image of the flame 

 was projected on to the slit, and the photographic plates were 

 exposed from thirty-five to forty minutes. One part of the spectrum 



VOL. LXI. R 



