50 !>!>. W. CKOFFREY DUFFIELD ON THE 



differences do not conclusively prove that the spectra are not the same, as special 

 conditions may govern the appearance of these violet heads. Further evidence upon 

 this poinf. is required. 



The most valuable information regarding the origin of the banded spectrum is 

 afforded by the photographs reproduced in Plate 1 at 5 and 20 atmospheres, when an 

 incipient band spectrum seems to originate from the strong lines a and c, and, since 

 at higher pressures there is a gradual extension of the energy outwards from these 

 lines until the whole of the plate is covered with bands, it .seems definite that the 

 replacement of the line by a banded spectrum is most intimately associated with the 

 vibrating system producing the line of the 1st sub-series in this particular part of 

 the spectrum ; there is, however, a member of the 2nd sub-series just off the plate 

 which is, perhaps, responsible for some of the bands on the red of Plate 1. 



The discovery that a silver arc burning in an atmosphere of hydrogen under certain 

 conditions gives rise to the spectrum of that gas* necessitated an inquiry into the 

 possibility of the air spectrum being produced when the same arc is burned in air 

 under unusual conditions ; but a comparison of the band spectrum with the spectra of 

 oxygen, nitrogen, and air gives no indication that its origin is due to these gases. 



0. H. BASQUIN, ' Astrophys. Journ.,' vol. XIV., p. 1. 





