the Occurrence of Enhanced Lines hi the Arc. 839 



behaviour characterizes the spark line X 4481. During the 

 first phase it shows the usual symmetrical expansion, but in 

 the second phase it appears as a fairly sharp bright line which 

 at a certain definite moment stops quite abruptly whilst all 

 the other lines continue and are not even affected in their 

 intensities. This curious fact clearly indicates that some 

 factor absolutely essential to the emission of A 4481 either 

 disappears suddenly at this moment or reaches some critical 

 minimum value below which it is no longer capable of 

 stimulating this particular radiation. 



High voltage experiments have likewise been made with 

 concentrated solutions of magnesium sulphate and chloride 

 between graphite poles. In the case of the sulphate solution 

 the lines of the Rydberg series are the most affected ; they 

 are winged on the red side during the first phase and in the 

 early stages of the arc flash. The flame line \ 4571 is absent, 

 as it also is with the 80 volts arc; but the band at X5007 

 shows throughout the arc flash. The triplets and the spark 

 line \ 4481 are enhanced as compared with the low voltage 

 spectrum. With a chloride film, under otherwise similar 

 conditions, the magnesium spectrum appears on the whole 

 more feebly developed. Nearly all the lines in the first 

 phase are intensified or widened as compared with their 

 appearance during the subsequent arc flash. The flame line 

 A- 4571 and the band at A, 5007 are absent; the lines of the 

 Rydberg series show only as traces in the film spectrum. 

 The series triplets and the line X4481 rise to the same level 

 in the second phase, thus indicating that also in this case 

 \ 4481 is particularly affected by the more powerful arc flash. 

 A similar effect of the higher voltage upon spark lines has 

 been noted in the case of cadmium by using a film of a con- 

 centrated solution of the chloride of this element between 

 poles of graphite, namely the two spark lines A 5338 and 

 A. 5379 attain almost the same level as the arc lines. 



As a general result of our observations on high voltage arcs, 

 we are led to conclude that the application of a high potential 

 appreciably prolongs the life of the spark lines during the 

 second phase and, when the arc passes between poles of 

 magnesium metal, brings into prominence the lines of the 

 Rydberg series. 



§ 4. Thickness of liquid film at moment of rupture, and 

 relative durations of the first and second 'phases. 



As was explained in § 5 of our first communication, the 

 introduction of a liquid film between the poles of an arc 

 gives rise to a transient luminous phenomenon composed of 



