230 REPORTS ON INVESTIGATIONS AND PROJECTS. 



which in the case of uranium glass was discovered by Burke, was subse- 

 quently investigated in much greater detail in the case of fluorescent liquids 

 by the present writers. In two recent papers Carmichel has denied the 

 existence of this phenomenon, but Miss Wick's measurements not only serve 

 to establish it, but afford an admirable quantitative study of the relations 

 involved. Her results will be incorporated in a final report to the Carnegie 

 Institution of Washington. Directly connected with the subject of fluo- 

 rescence absorption and dependent upon it is the increase of conductivity of 

 fluorescent liquids when under excitation, a phenomenon which was discov- 

 ered and described in an early portion of this investigation. This phenom- 

 enon has likewise been questioned by Carmichel, and arrangements have 

 been made with another graduate student, Mr. Percy Hodge, to repeat the 

 measurements upon this interesting subject. His results will likewise be 

 published in a final report. 



Preparations are at present being made for an extended investigation 

 upon the effect of high and low temperatures upon the position and intensity 

 of the fluorescence bands in the spectra of solids. 



Wood, Robert "W», Johns Hopkins University, Baltimore, Maryland. Grant 

 No. 248. Researches on the theory of light. (For previous reports see 

 Year Book No. 2, p. xxxix. Year Book No. 3, p. 128, and Year Book 

 No. 4, pp. 255-257.) $1,000. 



THE COMPLEX FLUORESCENCE SPECTRUM OF SODIUM VAPOR 

 AND ITS ANALYSIS. 



The study of the remarkable fluorescent spectrum of the vapor of metallic 

 sodium has been continued. The very complicated spectrum observed when 

 the vapor is illuminated with white light is being gradually dissected by 

 stimulating the medium with the monochromatic radiations from the electric 

 arcs between metal electrodes. The lamps of fused quartz made by Heraeus, 

 of Hanau, have been found most satisfactory, but unfortunately very few 

 metals can be used with them. For every type of monochromatic stimulation 

 we obtain a definite fluorescent spectrum, consisting usually of sharply defined 

 lines, spaced at nearly regular intervals measured along a normal spectrum, 

 both of longer and shorter wave-length than that of the exciting light. The 

 same series of lines can be obtained by excitation of various points of the 

 spectrum, providing that these points coincide in position with the lines of 

 the series. The series is not, however, always complete. For example, if 

 we excite the vapor with the blue cadmium line of wave-length 4799, we 

 obtain a series which is broken up into groups of two and four lines, owing 

 to the absence of certain lines. These missing lines can be detected in their 

 proper places by very long exposures, the complete set being made up of 

 lines almost exactly equally spaced along the spectrum. One of these lines 

 coincides in position with the line 5209 of the silver arc, and when the fluo- 



