LYMAN. — THE FALSE SPECTRA FROM DIFFRACTION GRATINGS. 43 



B was illuminated by light from a spark between iron terminals, and thus 

 the standard iron spectrum was photographed under the false niagueS|iura 

 lines. To determine in Angstrom units the amount by which the one 

 spectrum was shifted with respect to the other, both A and B were illu- 

 minated with light from the iron spark and the relative position of two 

 known lines was measured. As an example of the method, the false 

 line, which by the theory should be the 47th order of the grating of m 

 groups, was compared with the line in the iron spectrum 3541.2* and 

 found to have the relative position 3541.0. Next, the line 2737.0* in 

 spectrum A was compared with the line 4404.9 * in spectrum B. The 



o 



diifereuce between them was .5 Angstrom units. Thus the shift of 

 spectrum A with respect to B was [(4404.9 — .5) — 2737] =: 1667.4, 

 and the apparent wave-length of the false line was 3541.0 — 1667.4 = 

 1873.6. In this way the apparent wave-lengths of the four false lines 

 were determined. In order to check the accuracy of the method, slit B 

 was moved toward slit A and the grating was re-adjusted. In this way 

 the shift of one spectrum with respect to the other was altered and the 

 unknown spectra were compared with a new position of the iron spec- 

 trum. The following table gives some idea of the accuracy of the method 

 and of the agreement between the observed and the calculated values. 



The method gives an average error of 0.17 of an Angstrom unit. The 

 average difference between the observed and the calculated values is .14 

 units. Thus the difference between theory and practice is within the 

 errors in observation. 



It might seem that values different from n = 70 and a period of 3 

 would satisfy as well the conditions when substituted in the equations. 



* Exner and Haschek, K. Akail. der Wiss. in Wion, 106, Abth. II. (1807). 

 Compare also Kayser and Runge Arc Spectrum. 



