MOUNT WILSON SOLAR OBSERVATORY. 227 



GENERAL ELECTRIC-FURNACE EXPERIMENTS. 



In addition to the investigation just described, 114 photogi-aphs of 

 furnace spectra were made by Mr. King, chiefly of the spectra of iron, 

 calcium, strontium, and barium, under various conditions of tempera- 

 ture, vapor-density, and surrounding atmosphere. The iron spectrum 

 from X 2800 to X 6800 was photographed with the 15-foot concave 

 grating spectrograph at temperatures of 1650°, 2000°, and 2350° C, 

 these temperatures having been found to give the most decided differ- 

 ences in the spectrum. This material, with that previously obtained, 

 will furnish a homogeneous classification of iron lines according to 

 temperature change. 



The use of the "furnace flame," produced by passing oxygen through 

 the furnace tube during the vaporization of a metal, showed that oxida- 

 tion does not decidedly modify the line spectrum of iron. The effects 

 were more readily explained as the result of an altered distribution of 

 the vapor. A comparison with pubUshed data on the flame spectrum 

 of iron showed that temperature changes in the furnace can produce 

 effects similar to those ascribed to chemical action in the flame. 



The new photographs of the furnace spectra of calcium, strontium, 

 and barium provided material as to temperature changes of important 

 lines, and in the case of barium furnished interesting data as to line 

 structure and the conditions which govern it. In this spectrum the 

 lines given by the arc in air between X 3000 and X 3800 are in general 

 so diffuse and unsymmetrical that no accurate measurements of wave- 

 length can be made. Their structure in the vacuum furnace is so 

 different from that in the arc as to present the appearance of a different 

 spectnmi. The diffuse arc lines become very sharp in the furnace and 

 permit measures of high precision. Some, which in the arc are so 

 hazy that no measurement has been attempted, become well defined 

 in the furnace. Striking changes in structure also appear, a pair or 

 triplet sometimes showing in the furnace instead of the single arc line. 

 In case the furnace hne is single, it is usually at the violet edge of the 

 arc line, the displacements corresponding to a furnace difference of one 

 atmosphere being very large. In the region in question the wave- 

 lengths of 67 barium furnace lines were measm-ed from iron standards, 

 with a large gain both in number and accuracy over measurements 

 possible in the arc spectrum. Tests under various conditions showed 

 that the vacuum furnace at different temperatures gives always the 

 same lines, but that at high temperature the stronger lines become 

 unsymmetrical in the direction of the dissymmetry of the arc Une. 

 Increase of pressure in the furnace, however, gradually suppresses the 

 extra components peculiar to lines in the vacuum spectrum. It 

 appears that the structure of the arc line is due to a combination of 

 high excitation and pressure, the former being the chief agent in causing 

 the dissymmetry which gives rise to the wave-length differences. 



