MOUNT WILSON SOLAR OBSERVATORY, 225 



where many low-temperature lines are observed. The distribution 

 of line intensities through the visible spectrum bears little resem- 

 blance to the intensity gradation observed for the spectrum of an 

 incandescent solid, though the extension of the spectrum into the 

 ultra-violet increases with the temperature. 



An important class of furnace lines is found to be relatively much 

 weaker in the arc spectrum. For titanium, most of the lines of this 

 class appear at temperatures from 2300° C.to 2600° C, being absent 

 or faint at lower temperature and weak in the arc. They thus appear 

 especially sensitive to a certain degree of excitation, for which the 

 furnace is most suitable. 



The furnace, as regularly operated in a partial vacuum, is found 

 to give only the line spectrum for titanium, the bands given by the 

 arc in air not appearing in the furnace. This affords additional evi- 

 dence that the banded spectrum is due to titanium oxide. 



"Tube- Arc" Spectra. 



Some experiments by Mr. King with thin-walled tubes forced to a 

 high temperature led to the development of a new form of arc, the 

 spectrum of which shows many interesting peculiarities. The hori- 

 zontal graphite tube used in the furnace chamber was heated to high 

 incandescence and made to burn apart, an arc forming between the 

 ends which carried for a short time a current of about 800 amperes at 

 a potential of 30 volts. This "tube-arc" is essentially a high-current 

 vacuum arc inclosing a mass of luminous vapor somewhat removed 

 from the direct path of the current between the ends of the broken 

 tube. The spectrum is very different from that of the furnace and of 

 the ordinary arc in air, the most striking feature being the great 

 strength of the enhanced lines, which usually require the spark dis- 

 charge to show them strongly. In this case these lines appear in a 

 low-potential arc, and, further, have been found to be strongest near 

 the center of the tube's cross-section, away from the direct path of 

 the arc. In addition to the enhanced hues of metals, the tube-arc 

 gives with great briUiancy the hydrogen lines from Ha to He and the 

 entire line spectrum of carbon. The structure of the important 

 spark line X 4481 of magnesium is better defined than in any other 

 source in which it has been examined. The line was found to be 

 double, and the evidence is against this being caused by reversal. 

 The pair of lines which make up X4481 has been measured, the 

 violet line being about twice as strong as the red. 



When the image of the luminous interior of the tube-arc was 

 projected on the slit of the plane-grating spectrograph, interesting 

 variations were observea in the strength of the hnes of different ele- 

 ments between the center and wall of the tube. This led to an 

 extended series of experiments, in the course of which lines of hydro- 

 gen, carbon, titanium, vanadium, iron, chromium, silicon, tin, cal- 

 cium, strontium, and barium were examined. A region near the 



