the Occurrence of Enhanced Lines in the Arc. 861 



Plate XX. a gives the spark, water arc and ordinary arc 

 spectra of zinc in the ultra-violet region. Note the relative 

 strength in the water arc of the two spark lines A, 2502 and 

 A 2558, and the strong development in the two arc spectra 

 of the arc line triplet at X 2801. Spectra b and c show the 

 relative behaviour in ordinary arc and water arc at 80 volts 

 of the arc line triplet at X4811 and of the well-known spark 

 lines X4912 and X4924. Note absence of latter in the 

 ordinary direct arc (b). 



Photograph a on Plate XXL is a portion of the ultra-violet 

 region of the spark, water arc and ordinary arc spectra of 

 cadmium. Noteworthy are the relative behaviour in the 

 three sources of the spark lines XX 2195, 2313, and 2321 

 and the curious reversals of X 2288. Spectrum b shows the 

 abrupt drop in intensity of the enhanced line A, 4416 at 

 about the middle of the second phase of a water arc, whilst 

 its neighbour, the arc line A 4413, continues to be visible and 

 dies out only gradually. The behaviour of this enhanced line 

 is similar to that of the spark line of magnesium X 4481 

 illustrated on Plate XIX. b. 



§ 10. Discussion of results. 



The experiments described in this paper, together with 

 those related in our first communication, go to show that the 

 emission of light radiations in an electric arc is governed by 

 a number of factors which generally act simultaneously 

 and thereby complicate the task of tracing the origin of any 

 particular spectrum lines. By means of new experimental 

 methods it has been possible to bring out distinctly the effects 

 of that factor which is more directly responsible for the 

 emission of spark lines. Our preliminary work with lead 

 and tin had led us to the conclusion that the excitation of 

 these lines is connected with the existence of strong electric 

 fields, and it will therefore be necessary in the first place to 

 examine with the help of all the data now to hand whether 

 and in how far this conclusion was justified. As we have 

 shown, the spark lines attain their maximum of development 

 as regards both intensity and breadth during the first phase — 

 namely, within about 0*0001 second from the moment of 

 separating the electrodes. They continue to be emitted 

 after arcing has set in, and then fall off sharply in intensity 

 or stop quite abruptly before the end of the second phase. 

 Further, the spark line X 4481 is continuously emitted by 

 a stable arc burning in nitrogen. Thus a satisfactory 



