112 Mr. Piazzi Smyth on Carbon and Carbo- Hydrogen, 



pure tube-carbon-spectrum, just as carbo-hydrogen adds its 

 brilliant lines and many linelets when it has the opportunity 

 in one of its own tubes. 



Exactly so ! But in such case what does a carbo-nitrogen 

 tube do ? 



On examining carefully I found that it added a feature ex- 

 cessively faint, but perfectly peculiar to itself, and somewhat 

 reminding one, though at extreme distance, of the bands of 

 nitrogen ; for there were many rather regular haze-bands, be- 

 ginning sharply, sometimes with an actual line, on the side of 

 least refrangibility. One such line I would particularly call 

 attention to, of intensity .1*0, and at W.N. place 49,543; for 

 while it appeared so very clearly in a cyanogen-tube, I con- 

 vinced myself again and again that it did not exist (unless 

 homoeopathically as an impurity trace) in either carbo-hydrogen 

 or carbo-oxygen tubes. 



Here, then, we have arrived at a most notable stage in our 

 general inquiry ; for by pushing the examination to further 

 exactitude than has been usual, we have found, in direct oppo- 

 sition to general belief hitherto : — 



1st. That each of the three varieties of compound carbon 

 gases, viz. carbo-hydrogen, carbo-oxygen, and carbo-nitro- 

 gen, gives the later, or more violet-ward, details of the tube- 

 carbon's band-spectrum (under 1 inch induction-spark) diffe- 

 rently from the other. 



2nd. That the feature which is thus peculiar in a carbo- 

 hydrogen tube is undoubtedly the well-known and most bril- 

 liant coal-gas and air blowpipe-flame's carbo-hydrogen spec- 

 trum. 



3rd. Whence we conclude that the pale and weak residual 

 feature which is peculiar to a carbo-oxygen tube must there- 

 fore be the carbo-oxygen flame's spectrum in open air, and 

 the feature peculiar to the carbo-nitrogen tube the carbo- 

 nitrogen flame's spectrum similarly. 



4th. In which case, attending to what has been- already 

 remarked as to respective intensity and faintness of lines and 

 haze in the observations, we may see that whatever carbo- 

 hydrogen succeeds in introducing into one of its own tube- 

 spectra is just as remarkable for enormous overpowering force 

 as what either the carbo-oxygen or carbo-nitrogen introduces 

 is for ultra weakness. 



5th. Wherefore in flame-spectroscopy, when burning carbo- 

 hydrogen gas in a blowpipe, the carbo-oxygen and carbo- 

 nitrogen impurities, even though present to a large percentage, 

 cannot make their faint spectra appear in the presence, or to 

 the prejudice, of the carbo-hydrogen ; while, again, if we take 



