252 Dr. W. M. Watts on the Spectra of Carbon. 



determine the conditions on which the presence of one or the 

 other of these groups depends. A touch of the contact-breaker 

 will sometimes cause/ to disappear and be immediately replaced 

 by the other two groups. The change of temperature (if it be 

 so) thus caused is not attended, then, simply by the addition of 

 new lines, but causes the disappearance of one group and its re- 

 placement by two other quite different groups of lines. When 

 a Leyden jar is included in the secondary circuit, no trace of the 

 cavbon-lines is obtained if the jar be large enough, but instead a 

 brilliant spectrum, which is described afterwards as the fourth 

 carbon-spectrum and is represented in fig. 4«. I have employed, 

 instead of a Leyden jar, a graduating condenser consisting of two 

 opposed disks of metal, the distance between which could be 

 varied at pleasure. When the plates are separated, the conden- 

 sation is so feeble that the spark in carbonic oxide shows the 

 carbon-spectrum only; but as the plates are gradually approxi- 

 mated, the fourth carbon-spectrum appears gradually replacing 

 the old spectrum and finally completely extinguishing it. The 

 blue band /is the first to disappear, and is replaced by the group 

 123-133 of fig. 4, and the conspicuous line 76 of fig. 4 appears 

 nearly bisecting the interval between the first and second lines 

 of the group 8. 



When the density of the carbonic oxide is increased while the 

 spark (without condenser) passes through it, the gas is more 

 rapidly decomposed, the spark becomes more luminous, and the 

 spectrum more nearly continuous. At two atmospheres' pres- 

 sure the spectrum obtained is the carbon-spectrum, consisting 

 of 7, 8, 6, j", and 6 (the red end probably contains «), e, f, and d 

 being very brilliant. Increase of pressure up to about ten atmo- 

 spheres only produces the effect of filling up the intermediate 

 spaces with white light. 



The spectrum, including the groups f and 6, is also obtainable 

 from compounds of carbon with hydrogen. A Greissler's tube 

 enclosing naphthalin gives a splendid carbon-spectrum, in which 

 the groups f and 6 are especially brilliant. They are therefore 

 abundantly proved to be produced by carbon itself. 



By passing the spark through the vapour of carbonic disul- 

 phide, there can be obtained at will either Pliicker's sulphur- 

 spectrum of the second order or the carbon-spectrum on a 

 background of continuous light due to the separation of sul- 

 phur. 



The spark in the vapour of carbonic tetrachloride gives either 

 the carbon -spectrum or the chlorine-spectrum, according to cir- 

 cumstances. 



A Geissler's tube enclosing amylic alcohol gives the carbon- 

 spectrum, consisting of «, 7, 8, e, and/. 



