278 CARNEGIE INSTITUTION OF WASHINGTON. 



COMPARISON WITH SPARK DISSYMMETRIES. 



The iinsymmetrical lines given by the disruptive spark were described 

 in last year's report. These results have been closely compared with 

 the tube-arc effects, additional photographs of the spark spectrum 

 being taken for this purpose. In addition to iron, the tube-arc and 

 spark effects for selected regions of the titanium spectrum have been 

 compared. The resemblance between the dissymmetries in the two 

 sources is very close throughout, indicating that the features of the 

 tube-arc which make its spectrum similar to that of the spark are 

 responsible also for the dissymmetries of the individual Unes. 



INVESTIGATIONS OF THE IRON ARC. 



The study of the pole-effect in the iron arc has been continued by 

 Mr. St. John and Mr. Babcock as in the preceding year. The Ust of 

 affected lines now contains 286 which shift toward the red at the pole 

 and 80 wliich shift toward the violet. 



Additional evidence of the reaUty of the displacements has been 

 secured: first, by confirming with the Koch microphotometer the 

 measures made with the Hartmann instrument; second, by interferom- 

 eter measures of the pole effect directly; and third, by superposing 

 upon the iron spectrum the absorption spectrum of iodine vapor at low 

 pressure. For this purpose the spectrograph slit was placed in coinci- 

 dence with the axis of the enlarged image of the arc. The iodine vapor 

 was contained in a spherical glass bulb, placed in the beam of fight. It 

 happens that absorption lines due to iodine fall upon some of the 

 emission lines due to iron, and as the former are among the narrowest 

 fines known, they serve as fiducial fines of the highest quality. With 

 reference to them, unmistakable displacements of the maxima of the 

 iron lines in passing from center to pole of the arc are shown upon 

 our photographs. 



A possible dependence of the pole-effect upon the density and tem- 

 perature of the iron vapor has been examined by electric-furnace tests 

 upon selected lines, with the result that no change of wave-length was 

 shown either for a tenfold change of vapor density or for a temperature 

 variation from 2100° to 2600° C. 



The arc in vacuo has been tested for pole-effect at pressures of 0.5 

 cm. and 10.0 cm. of mercury. In general, the pole-effect disappeared 

 under these conditions, although the length and current-strength of 

 the arc were the same as those used under normal pressure. This 

 would indicate that the fall of potential plays a minor role, if any, in 

 producing the pole-effect. Further study of this question is to be made. 



A comparison of the pole-effects at positive and negative poles with 

 the relative luminosities of these regions in different parts of the spec- 

 trum shows no apparent relationship between the two quantities. 



The possibility of an increase of pressure locafized in the core of the 

 arc and very near to the pole was soon recognized as a hypothetical 



