MOUNT WILSON SOLAR OBSERVATORY. 259 



be useful chiefly in the production of oxide band spectra, the relative 

 intensity of Unes in the ''furnace-flame" apparently being much the 

 same as in the vacuum furnace at the same temperature. 



Anomalous Dispersion. 



The electric furnace was adapted to the production of anomalous- 

 dispersion effects by cutting away the upper portion of the tube, so that 

 when a metal was vaporized in the trough thus formed the density 

 gradient was equivalent to a vapor prism. Light passing through the 

 vapor to the spectrograph gave, on each side of those absorption lines 

 capable of showing anomalous dispersion, the familiar curved spectrum. 

 The steadiness of the source made the use of high dispersion possible. 

 Observations by this method were made for lines of sodium, iron, cal- 

 cium, and chromium. The magnitude of the phenomenon was clearly 

 related to the tendency of the line to reverse, and it was found that 

 a certain temperature of the absorbing vapor is required to give the 

 most effective prism for a given type of line. Particular attention 

 was paid to the chromium hues XX 4254, 4275, 4290, which show 

 very strong anomalous dispersion. Sharp absorption lines of tita- 

 nium and calcium occur close to members of this group, and tests 

 of lines, so situated, were made for the mutual repulsion predicted 

 by the theory of Julius. No evidence of such an effect was found, 

 though it was sought by means of narrow lines of titanium and calcium 

 only 0.2 to 0.6 a from strong chromium lines. The same lines were 

 examined for a change of wave-length possibly affecting close pairs 

 from mixed vapors when one of the lines shows strong anomalous dis- 

 persion. The wave-lengths were found constant to 0.001 A, whether 

 the line of strong anomalous dispersion was present or not. 



Structure of X 6708 of Lithium and its Presence in Sun-Spot Spectra. 



An examination of the line X 6708 was made under high dispersion 

 with the object of determining whether it is sometimes produced by 

 calcium or whether it is in all cases the well-known red line of lithium, 

 and also for the purpose of comparing its wave-length with that of a 

 prominent sun-spot line. The observations led to the conclusion that 

 the line is always due to lithium and that its structure differs according 

 to the condition of the source, being apparently governed by the 

 amount of the material present. Two distinct sets of components may 

 be produced in either arc or furnace, or simultaneously in different parts 

 of the arc; but an excess of vapor gives a reversed line without structure. 

 The wave-lengths of the components were measured and their mean 

 value was found to agree within 0.01 A with that of the somewhat 

 diffuse sun-spot line, thus making it fairly certain that lithium is 

 present in the sun. X 6708, a low-temperature furnace line, does not 

 appear in the solar spectrum, but is apparently brought out in the 

 spot spectrum by reason of the reduced temperature of sun-spots. 



