228 CARNEGIE INSTITUTION OF WASHINGTON. 



The fact that the furnace Unes permit of close measurement has made 

 possible a comparison with the solar spectrum. In the case of 31 lines, 

 the wave-lengths of barium lines in the furnace are in close agreement 

 with faint unidentified lines of the solar spectrum. In cases where a 

 line breaks up in the furnace into two or three components, the solar 

 lines appear to correspond to these rather than to the line given by the 

 arc in the air. This points to a reduced pressure and relatively weak 

 excitation in the sun for the level at which these barium lines originate. 



Analyses have been made for the spectra of two samples of rare 

 earths suspected of containing new elements. While positive results 

 have not been obtained, these and other experiments with furnace 

 and arc have shown several strong lines in the ultra-violet which have 

 not been identified in known spectra. The strongest of these agrees 

 closely with an unidentified solar line. Extensions of stellar and 

 nebular spectra in this region will be watched for evidences of the 

 presence of these fines. 



PRODUCTION OF SPECTRA BY CATHODE LUMINESCENCE. 



The prefiminary results on the excitation of metallic vapors through 

 the bombardment by cathode rays, reported b}^ Mr. King last year, 

 were much extended by the experimental work of IVIiss Carter during 

 the summer of 1916. Spectra of calcium and iron were obtained of 

 sufficient richness to show what is to be expected from a source of 

 this natm-e. The calcium spectrum was quite different in its features 

 from that given by the usual sources, the various ''single-line" series 

 being strongest, followed by the series consisting of line-pairs, while 

 the triplet series were relatively faint. The high intensity of X 4227 

 was of special interest. The iron spectrum obtained showed com- 

 paratively few lines, these being lines that are strong in the furnace, 

 arc, and spark. Stronger effects can doubtless be obtained by this 

 method with the high-tension direct current supplied by the new trans- 

 former and rectifier. The work is again being taken up by Miss 

 Carter, with the purpose of obtaining several of the more important 

 spectra, so that the characteristics of intensity and structure of lines 

 when thus excited may be definitely established. 



INVESTIGATION OF THE STARK EFFECT. 



The investigation of the Stark effect described in last year's report 

 has been continued by Mr. Anderson. A plane-grating spectroscope 

 giving intense spectra and fairly high dispersion (3 to 5 A per mm.) 

 was built and has been found very satisfactory. Displacements 

 amounting to 0.01 or 0.02 A can readily be detected, and the brightness 

 is such that the exposure time need rarely exceed 30 minutes. 



A vacuum chamber suitable for use with metals having high melting- 

 points presented some experimental difficulties, but one was finally 



