MOUNT WILSON OBSERVATORY. 253 



made. Thus the low-temperature Unes are the more strongly absorb- 

 ing, which is in harmony with the phenomena of reversal. The present 

 experiments, however, show the relation with especial clearness, since 

 the emission spectrum proves that at a certain temperature the Une is 

 being radiated but that a higher temperature is required before it can 

 be absorbed. The change of absorptive power with wave-length is very 

 decided, the effect increasing rapidly toward shorter waves. 



The Zeeman Effect for Electric-Furnace Spectra. 



The use of a small tube-furnace in a magnetic field opens an extensive 

 line of investigation, of which a preliminary survey has been made. 

 A graphite tube 4 inches long, inclosed in a water-cooled jacket, was 

 placed axially between the poles of the large magnet, which, with this 

 gap, gave a field of about 6,500 gausses. The tube, of which 3 inches 

 were heated by a strong alternating current, readily gave the n-com- 

 ponents of the furnace lines, when observ^ed along the lines of force. 

 The spectra observed were of iron, from X3500 to X6700, and of vanad- 

 ium, from X5100 to X6700. Comparing the Zeeman components in 

 furnace and spark spectra, the lines common to both showed no differ- 

 ence, either in character or magnitude of the separations, so that in 

 future work the two sources can be used to supplement each other, to 

 obtain in the magnetic field the lines most favorably given by each. 

 A large class of the lines, hovv^ever, which, probably without exception, 

 are intensified in sun-spot spectra, are much stronger in the furnace 

 than in either arc or spark, so that the exposure time for such lines 

 in the magnetic field is a matter of minutes mth the furnace and of 

 many hours wdth the spark. Even in the limited work thus far done, 

 the furnace has given the magnetic separation of a considerable num- 

 ber of fines not yet obtained \\dth the spark. High fields are not 

 possible with the present apparatus, but the use of the furnace inside 

 a solenoid will remove this limitation, and the characteristic sharpness 

 of furnace lines will be of much advantage in defining the Zeeman 

 components when the furnace is inclosed in a vacuum chamber. 



Another important advantage of the furnace in the magnetic field is 

 the ease with which the inverse Zeeman effect is produced by using a 

 plug in the tube to give an absorption spectrum, as described above. 

 A number of spectrograms have been made with this arrangement. 

 The n-components appear to be quite the same in absorption as in 

 emission. The study of absorption phenomena at varying angles to 

 the lines of force with optical apparatus similar to that used in solar 

 work may, however, give data on the inverse effect vrhich will be 

 directly applicable to sun-spot phenomena. In any case, the magnetic 

 action on absorption lines can now be studied with the same facility 

 as for emission lines. 



