500 Mr. R. Rossi on the P 



res sit re 



Passing from air to carbon dioxide, for instance, the dis- 

 placements ought to be very nearly in the ratio 2 to 3. 



Some work on spark spectra in different gases under 

 pressure has already been done by Hale and Kent* and by 

 Anderson f. The latter found larger displacements in carbon 

 dioxide than in air; but, as pointed out by Humphreys J, his 

 results were not conclusive as to the effect of speed tic inductive 

 capacity on the pressure displacement. 



In the following work the displacement of some lines of 

 the spectrum of an iron arc burning in air and carbon dioxide 

 under pressure are compared. Hydrogen was also tried ; 

 but the arc burns so poorly in that gas under pressure, that 

 too long exposures of the photographic plate would have 

 been needed, and the results would have been spoiled by 

 leakages of the apparatus and changes of temperature of the 

 room §. 



Photographs were taken at 15, 30, and 50 atmospheres 

 with the 2H feet concave grating spectrograph of this 

 laboratory. A small region of the spectrum containing some 

 sharp lines was chosen, thus enabling the measurements to 

 be made more accurately. The arc was found to burn in 

 carbon dioxide just as well as in air. 



The accompanying table gives the displacements at the 

 various pressures. Figures in brackets denote doubtful 

 readings, the letter R indicates that the line was found to be 

 reversed at that pressure. It can be seen that, with a few 

 exceptions, the displacements are the same in the two gases 

 within the limits of experimental error. The mean displace- 

 ment per atmosphere of all the 23 lines studied is found to 

 be '00411 and "00401 Angstrom unit for the arc burning 

 in air and carbon dioxide respectively. 



There also are no very noticeable differences between the 

 appearances of the spectra in the two gases. There are a few- 

 more reversals of lines when the arc burns in carbon dioxide 

 than when it burns in air ; and there seems to be a slight 

 tendency of the lines to be broader and more diffuse in carbon 

 dioxide than in air. The relative intensity remains the same. 



So far then as this evidence from only two gases is worth, 

 it points to the fact that the specific inductive capacity is of 

 but secondary importance in the cause of the displacement 



* Publications of Yerkes Observatory, vol. iii. Part II. (1907). 



t Astrophvsical Journal, vol. xxiv. p. 221. 



X Astrophvsical Journal, vol. xxvi. p. 18. 



§ At 15 atmospheres the exposure necessary when the arc was burning 

 in air or C0 2 was on the average 1 or 2 minutes, while it was estimated 

 that for the same arc burning in hydrogen an exposure of 3 hours would 

 have been needed. 



