206 CARNEGIE INSTITUTION OF WASHINGTON. 



LABORATORY INVESTIGATIONS. 

 THE ULTRA-VIOLET SPECTRUM OF TITANIUM. 



King has made a study of the titanium spectrum in the ultra-violet with 

 the electric furnace, beginning at X2600 and extending to X3900, the limit 

 of a former investigation. The incompleteness of existing wave-length 

 tables, together with the requirements of a search for related lines, has neces- 

 sitated an especially thorough examination of the effect of different light- 

 sources and the measurement of many new lines. The furnace temperatures em- 

 ployed ranged from 2000° to 2700° C. for both emission and absorption spectra. 

 According to the usual method, the intensities at three furnace temperatures 

 and in the arc were employed as a basis for the temperature classification. 

 About 780 lines have been classified in this way, 321 of which either have 

 been measured for the first time or have had their wave-lengths materially 

 improved by the more favorable character of the furnace lines. The spec- 

 trum is especially rich in lines which are much more prominent in the furnace 

 than in the arc, while other lines which are very diffuse in the arc are sharp 

 in the vacuum furnace. Experience with other elements indicates that lines 

 relatively strong in the furnace arise from the simpler vibrations and should 

 be the first material to be used in a search for regularities in the spectrum. 



The method of mixing in the furnace the element under examination with 

 one having a lower ionization potential has been used with titanium, and the 

 quenching of enhanced lines by this means has been observed, as in the case 

 of the other elements which have been tested. In accordance with Saha's 

 theory, this effect, in the absence of known series relationships, offers proof 

 that these lines arise from the ionized atom. While the furnace gives only 

 the stronger enhanced lines, it is found to emit these more readily in the 

 region of shorter wave-lengths. Photographs of the spark spectrum made 

 on a large scale show a variety among the lines of the ionized atom, as regards 

 ease of production, structure, and reversal phenomena similar to that found 

 among the lines of the neutral atom. 



The method of treating furnace, arc, and spark spectrograms in the ultra- 

 violet has been extended to X5000, to include the region most used in stellar 

 spectra. Aside from the measurement of some furnace lines not previously 

 listed, this has provided a check upon the earlier classification. 



TITANIUM-OXIDE BANDS IN THE ELECTRIC FURNACE. 



The green, yellow, and red bands of titanium oxide, prominent in the 

 spectra of M-type stars and of sun-spots, have been examined by King at 

 various furnace temperatures, both in emission and absorption. The bands 

 appear at a lower temperature than the lines in this region, and the relative 

 strength of lines and bands can be controlled closely by the supply of oxygen. 

 Titanium oxide, when placed in the furnace, seems to be dissociated rapidly, 

 and is much less effective in producing the bands than a stream of oxygen led 

 through the tube over metallic titanium. In connection with these observa- 

 tions, the bands have been photographed on a large scale for a study of their 



s1^ n 1 c\ 1 1 rp 



GENERAL ELECTRIC-FURNACE INVESTIGATIONS. 



The arc Hnes of silicon, X4103, and the enhanced pair X4128 and X4131, 

 which are important in certain stellar spectra, have been measured by King 

 on high-dispersion plates of the "tube-arc" spectrum. The spark triplet 



