[ 330 ] 



XXXIX. On the Constitution of the Lines forming the Low- 

 Temperature Spectrum of Oxygen. By Piazzi Smyth, 

 Astronomer Royal for Scotland*. 



I HAD added (at p. 140 of my paper on " Gaseous 

 Spectra") to the "small dispersion" account therein 

 given of the above-mentioned spectrum of luminous oxygen gas, 

 that two, if not four, of its very few and scanty lines 

 appeared, when viewed with much higher dispersion, to be 

 double, but that I hoped to give a more exact account of 

 them, after completing some arrangements then in progress 

 for increasing both the dispersion and magnifying-power of 

 my spectroscope. 



These improvements, together with a great advance in 

 definition, were finished last November; and almost the first 

 result they yielded was, to much more than confirm what I 

 had only suspected before, and to do so, moreover, with such 

 vigour and certainty as to make me inquire right and left for 

 several weeks, to ascertain if what I then saw was really a 

 new discovery, or had perhaps been known long before to 

 older and better spectroscopists. 



So far as I have been able to gather, the thing is new, and 

 promises to be important to theorists in molecular vibrations, 

 on account of what it fulfils. To explain this, let me refer to 

 Dr. Arthur Schuster's valuable paper in the Transactions of 

 the Royal Society, London, for 1879, " On the Spectra of 

 the Metalloids : Spectrum of Oxygen." Beginning his ex- 

 position of the low-temperature spectrum (after having treated 

 very fully of the exceedingly different high-temperature spec- 

 trum) of oxygen with the late Prof. Pliicker's account of it, 

 and concluding with his own confirmatory observations, Dr. 

 Schuster shows that the low spectrum had always been found to 

 consist of only four, single, wide-apart lines — viz. one in the 

 orange, one in the citron, another in the green, and another 

 still in the indigo-violet; and to that spectrum, containing 

 only four such solitary, single, simple lines, the learned 

 Doctor gave the second or alias name of the " compound-line " 

 spectrum of oxygen. 



The reason for such an apparent misnomer, or lucus a non 

 lucendo, was, that at the very moderate temperature of electric 

 illumination at which this trifling-lined spectrum of oxygen 

 appears, most other gases do give forth very compound-line 

 spectra indeed — spectra with hundreds or even thousands of 



* From the Transactions of the Royal Society of Edinburgh, vol. xxx. 

 part 1. Communicated by the Author. 



