KENT. — SPECTRA OF THE ZINC ARC AND SPARK. 109 



was a carbon arc of somewhat greater current than here used, but the 

 amount of vapor was never great, only small bits of metal being in- 

 serted in the arc, and the exposure always being made when it was 

 burning quietly. These two sets of standards were probably much 

 the same. Still, assuming them different, if the potential gradient 

 determine the enhancement of the red satellites and we accept Nutt- 

 ings classification of gradient, from low to high the order being, (!) 

 heavy current arc, (2) low current arc and inductance spark, (3) high 

 capacity and non-inductance spark, then the assymmetry of satellites 

 (and resultant shift) obtained in this investigation with low current 

 arcs as standards would be even less than that found with the some- 

 what higher current arcs previously used. However, as stated above, 

 in the arc there seems to be no regular, controllable nor enduring 

 enhancement of either red or violet satellites. 



Janicki's suggested explanation of the shifts obtained — namely, as 

 " unsymmetrical reversals like those of chromium and calcium, 

 reversals which their grating would not resolve and which appeared to 

 them as line-shifts" must then be replaced by this enhanced satellite 

 theory. 



The distances between the satellites in Plate 2, 48 (c) are approxi- 

 mately 0.05 Angstroms. We may then say that the removal of two 

 layers of inductance in coil (a) has shifted the center of gravity of the 

 line at least 0.02 Angstroms. In the extreme case then, with no 

 inductance in the circuit, the shift might easily be in the neighborhood 

 of 0.032 Angstroms, as formerly obtained. 



The writer wishes to record his appreciation of the kindness of 

 Professor Goodwin of the Massachusetts Institute of Technology in 

 loaning his Petitdidier echelon. To the Rumford and Bache Com- 

 mittees, and a personal friend, Mr. J. DeL. VerPlanck, the writer is 

 indebted for funds which made this investigation possible. In the 

 actual work of obtaining the results he wishes to acknowledge the 

 faithful assistance rendered by various students, especially Messrs. 

 Walter F. Burt, Russell T. Hatch, Charles H. Smith and Carl K. 

 Springfield. 



Physics Laboratory, Boston University, 

 June, 1912. 



