Arc Spectra of some Metals. 503 



The lines 2601*03, 2575-15, 2562-70, 2138-3, and the re- 

 maining six triplets of Kayser and Runge could not be obtained 

 in the hydrogen arc under any reasonable exposure. 



The Iron Spectrum. 



In the case of the iron arc, the number of lines is so enor- 

 mous that we can here merely illustrate the effects of a 

 hydrogen atmosphere by giving the principal changes which 

 occur in the region covered by a single negative. In the first 

 column of the following table is given the wave-length of the 

 line. These wave-lengths are reliable only to about 0'05 of an 



Angstrom unit. The second column gives the intensity of 

 the line in an atmosphere of hydrogen, on a scale which runs 

 from "j 1 " for lines just easily visible to " 10 " for the very 

 heavy lines. In the third column is given the effect pro- 

 duced by hydrogen. Here the word " new " is used to 

 indicate that the line does not appear in the ordinary iron arc. 

 " Enh." is a contraction for the word "enhanced"; and the 

 number which follows " enh." is intended to indicate, roughly 

 of course, how many times greater the intensity of the line is 

 in hydrogen than in air. The term " Dim." is a contraction 

 of " diminished ; " and the number following it indicates how 

 many times weaker the line is in hydrogen than in air. 



The illustrations given are sufficient to show that the 

 hydrogen atmosphere does more than slightly alter these 

 spectra : it profoundly modifies them. 



The explanation of these changes may be simple ; but it is 

 not patent. A number of hypotheses thrust themselves upon 

 one's attention ; but as a rule they are very difficult to test by 

 experiment, and they have therefore little value. For 

 instance, it does not appear impossible that the resistance of 

 the arc may vary considerably with the nature of the atmo- 

 sphere surrounding it : and, if so, then the temperature and 

 the spectrum may vary. But whether, in such case, hydrogen 

 would serve to increase or to diminish the temperature, it is 

 not easy to predict. 



In nearly all the spectra which I have photographed an 

 average direct current of about 2 amperes has been employed, 

 while the pressure between the electrodes has varied from 75 

 to 100 volts. The current was constantly varying in intensity, 

 and was often completely interrupted. Is it not possible that 

 an atmosphere of hydrogen increases the rapidity of these 

 interruptions, and hence alters the effects of self-mduction, 

 thereby changing the temperature of the arc? 



Still a third hypothesis is that the introduction of hydrogen 



