170 Abnormal Changes in the Spectrum of Lithium. [Nov. 19, 



Measurements were also made of two other lines in the arc spec- 

 trum. The wave-length of a weak and very diffuse line on one plate 

 was 4132*82, and that of a very weak sharp line, 4273*32. The 

 former line on another plate was stronger and broader; its middle 

 was at 4132*35, whilst the other line, much broadened towards the 

 red, had its strongest part at 4273*62. 



The current used in working both the open and enclosed arcs was 

 about 9 amperes. Kayser and Eunge employed a current of 25 — 35 

 amperes from one machine and, to bring out the weak lines in some 

 spectra, a current of 40 — 50 amperes from another machine. It seems 

 probable that thej^ worked with the intense arc which was only obtained 

 by the author when the carbons were very near together, and that they 

 observed only the spectrum of broadened lines which the author found 

 was emitted by the intense arc and near the negative pole by weaker 

 arcs. Their remarks on the appearances of the lines in the first and 

 second subordinate series confirm this view. Eder and Valenta 

 obtained a similar spectrum to Kayser and Eunge's arc spectrum in 

 the spark spectrum of metallic lithium when they used a condenser in 

 the secondary circuit. 



The author has pleasure in thanking Professor Liveing for the 

 interest he has taken in this work and for some useful suggestions. 



Conclusions. 



The lines in the principal series of lithium appear to broaden and 

 reverse normally. 



The lines in the second subordinate series do not reverse even in the 

 arc, but in strong arcs they broaden towards the less refrangible end 

 of the spectrum and become diffuse on that side. 



The first line in the first subordinate series, wave-length 6103*84, 

 is almost normal ; it broadens 'slightly more on the more refrangible 

 side than on the other. The other lines in this series also broaden on 

 both sides and become diffuse, but they broaden more rapidly on the 

 more refrangible side than on the other. The centres of the broadened 

 lines are more refrangible than the corresponding lines in the narrow 

 state. The inner core of intense arcs, and the parts near the nega- 

 tive poles of weak arcs and sparks, give a broad reversed line with its 

 centre about wave-length 4602*4; whilst the part near the positive 

 pole in weak arcs, and the flame of the arc, give a sharp bright line, 

 wave-length 4603*07, coincident with the lines in the spectra of the 

 oxyhyclrogen flame and uncondensed spark. The similar changes in 

 the other lines diminish with their refrangibility. The wave-lengths 

 hitherto recorded for these diffuse lithium lines would appear to be 

 those of abnormal lines. The true lines are the sharp bright ones 

 which occur, without complication, in the spectrum of lithium in the 

 oxyhydrogen flame. 



