1903.] The Spectra of Metallic Arcs in an Exhausted Globe. 257 



General Conclusion. 



So far as they go, the experiments seem to suggest that the modifi- 

 cations of the arc . spectra in an exhausted globe may be due to the 

 presence of hydrogen liberated from the heated poles. It has already 

 been pointed out that one of the effects of a hydrogen atmosphere on a 

 metallic arc is to introduce enhanced lines into the spectrum, and the 

 presence of hydrogen under the new conditions of experiment is 

 indicated in the case of magnesium and zinc by the appearance of the 

 H0 line, and in magnesium, zinc, and cadmium by flirtings which are 

 known to appear in the presence of hydrogen. 



It is well known that hydrogen is occluded by many metals, and 

 experiments made to determine the relative amounts of the gas given 

 ff on heating in vacua show that the amounts are roughly proportional 

 to the relative strengths of the enhanced lines appearing when the arc 

 passes in the exhausted globe; that is, magnesium and zinc gave off 

 the greatest quantities of the gas, cadmium the least, and iron an 

 intermediate amount. The apparent absence of the F line in the 

 spectra of cadmium and iron in the exhausted globe may, therefore, be 

 due to its feeble intensity on account of the smaller quantity of gas 

 driven out, while the absence of hydride " bands in the case of iron 

 may perhaps be explained by supposing that combination of hydrogen 

 with iron does not readily take place. 



The exact nature of the action of hydrogen on the arc requires 

 further investigation. 



The experiments described in the foregoing paper have been made 

 in Sir Norman Lockyer's laboratory, to whom we desire to express our 

 thanks for the privileges afforded. We also wish to express our thanks 

 > Mr. F. W. Jordan for assistance in carrying out the experiments 



