Arc Spectra of some Metals. 499 



quartz, which projected the arc upon the slit of a 10-foot 

 concave Rowland spectrograph. 



The hydrogen was furnished by three electrolytic cells, 

 made with sheet-lead electrodes in dilute sulphuric acid. 

 Each cell absorbed 12 amperes of current, so that the 

 hydrogen produced by 36 amperes was continually flowing 

 through the hood. This hydrogen, which was introduced 

 through a stopcock on one side, was allowed to escape through 

 a stopcock on the opposite side of the hood, where it burned : 

 continuously as a " pilot " flame about 2 centimetres high. 



In the current of hydrogen, a trap, instead of a drying- 

 tube, was used ; for it was found by experiment that there' 

 was no difference in effect between hydrogen which had been, 

 dried by phosphorus pentoxide and sulphuric acid, and 

 hydrogen which had not been dried at all. 



The most marked effect of hydrogen on the arc spectrum 

 is a general diminution of intensity. Consequently, in order 

 to photograph the spectrum of the arc in hydrogen so as to 

 give, on the negative, an average intensity equal to that of 

 the arc in air, one must make the exposure in hydrogen from 

 5 to 100 times as long as in air. 



The most interesting effect of the hydrogen atmosphere, 

 however,, is not the change in the average intensity of the 

 entire spectrum, but the change of relative intensity in the 

 lines of any one substance. 



This change was studied by photographing on each 

 negative three spectra, side by side. The first was that of 

 the arc in air, the second that of the arc in hydrogen, the third 

 again that of the arc in air. But the exposure was always 

 so timed that the spectrum in hydrogen had an intensity 

 which was intermediate between the two intensities in air. 

 The advantage of this is that, in comparing intensities, if a 

 line is apparently weakened one has a still weaker spectrum 

 with which to compare it, and, hence, can decide whether 

 the weakening is due to under-exposure or due to some effect 

 on this particular line and not on the whole spectrum. In 

 like manner, if a line is apparently intensified, one has a still 

 stronger spectrum with which to compare it, and to convince 

 himself that the intensification is not due to over-exposure. 



The effects of hydrogen on the spectra of magnesium and 

 zinc are partially summarized in Tables I. and II., which 

 follow. 



