96 PROCEEDINGS OF THE AMERICAN ACADEMY. 



dinary; states that they should have aroused Nutting's suspicion 

 and regards them as due to ghosts which become visible when the 

 intensity of the source is sufficiently great. He writes: — 



"Thus, according to Nutting, the red Cd line, the red and the blue 

 Zn lines form triplets; whereas, even with the greatest intensity and 

 the most varied sources of development, it is just these very lines 

 that have always been found to be unquestionably single by Michel- 

 son, Fabry and Perot, Hamy, Gehrcke and van Baeyer, and myself. 

 * * * Nutting's echelon had about the resolving power of the plane 

 parallel plate C and did not approach that of plate H, so that the 

 objection cannot be raised that he was able to make closer observa- 

 tions by reason of having a finer instrument. According to him all 

 five prominent silver lines are compound, and indeed, both triple and 

 quadruple, while the plate H even with the greatest intensity shows 

 no sign of satellites. * * * The characteristic line-structure remains 

 the same, no matter how the spectrum is produced. This is confirmed 

 by the agreement of the observations of the lines of Cd and Zw, where 

 it makes absolutely no difference with whatever instrument one 

 observes and no matter how the spectrum is produced. * * * That 

 the designation of the brightness of the satellites sometimes varies, 

 as in Cd X 4800, is immaterial, since the satellites are weak and the 

 differences in their intensity very slight." 



Here follows a discussion of unsymmetrical broadening noted with 

 the Rowland grating by Kayser, Rowland and others. The statement 

 is made that " a good Rowland grating would not resolve an unsym- 

 metrical reversal the components of which, like the chromium line, 

 are 0.043 Angstrom units apart, and the resultant apparent shift 

 about 0.02 Angstrom units." There follows a reference to the work 

 of the writer who, with Avery, made certain measurements upon two 

 titanium lines. He writes: — 



"They found an average shift of 0.019 and 0.018 Angstrom units 

 for the two titanium lines XX 3900.7 and 3913.6. In the mean taken 

 from both observers, the minimum and maximum shifts for the line X 

 3900.7 are found to be 0.009 and 0.038 Angstrom units. This very 

 circumstance seems to me to indicate that Kent and Avery were 

 dealing here with unsymmetrical reversals like those of chromium and 

 calcium, reversals which their grating would not resolve and which 

 appeared to them as line-shifts." 



