﻿98 Miss Ellen O'Connor on the Spectrum of 



A comparison o£ the photographs d, e,f, g (PL V. fig. 2) 

 in the order given shows a gradual change in the spectrum. 

 The continuous background decreases, the lines become 

 sharper and more numerous. Since in the cases d, e, and / 

 the inductance in the oscillatory circuit is constant, and in 

 the cases e and g the capacity is the same, the change cannot 

 be regarded as due to variation of inductance or capacity 

 alone. The order of these spectra is, however, the order of 

 increasing frequency of the oscillations, and since the current 

 through the arc varies inversely as the frequency (approx.), 

 one or both o£ these are to be regarded as associated with 

 the change in spectrum. 



In those cases where there is a great difference in frequency, 

 such as in d and g, where the frequencies are in the ratio 

 1 : 3J, the differences in the spectra are considerable, while 

 in g and i, where the frequencies are in the ratio 1 : 1^, the 

 spectra are the same. 



In the low-frequency arc, when the oscillatory circuit has 

 an inductance of 163,200 cm. and a capacity of '01245 mfd., 

 the spectrum consists principally of the b triplet, the lino 

 4481, the oxide fluting which is somewhat lost in the con- 

 tinuous background, and four lines 5528°75, 4703*33, 4352*18, 

 4167*81 of the Rydberg* series. 



Any alteration in the oscillatory circuit within the limits 

 used in these experiments appears to have little effect upon 

 either the b triplet or the line 4481. They always remain 

 the strongest lines in the spectrum, their intensities being 

 approximately equal. The lines of the series become sharper 

 and stronger with increase of frequency. This can be seen 

 in d, e, f, i (PL Y. fig. 2), where the frequencies were 

 1-1 x 10 5 , 2-7 x 10 5 , 3-9 x 10 5 , and 6*5 X 10 5 . 



In the case of the lowest frequency all the lines, with the 

 exception of 5528, are diffused towards the red end of the 

 spectrum, as in the case of the spark, and still more in the 

 direct-current arc at atmospheric pressure (PL V. fig. 2). 



With rise of frequency this diffusion disappears. If the 

 frequency is kept constant and the capacity and inductance 

 varied, there appears to be no alteration in the intensity of 

 the lines. Photographs / and g show the spectra of the arc 

 when the capacity in the oscillatory circuit was '00101 mfd. 

 and '002 mfd. respectively, the corresponding inductances 

 being 163,200 cm. and 82,410 cm., the frequency in each case 

 being 3*9 x 10 5 per sec. 



A distinctive feature of the high-frequency arc-spectra is 



* Rydberg, Wied. Ann. I p. 625 (1893), Hi. p. 119 (1894). 



