700 Relatlonsliip heliceen Spectra and Atomic Welglits. 

 Again, if we assume the following doublets as homologous 



Ea. 



(1) 

 (2) 

 (3) 

 (4) 



4826 

 4436 



4826 



4;j4i 



3815 

 3650 



4f;82 

 4436 



lA. 



207206 

 22540-4 



20720-6 

 23037-1 



26215-2 

 27399-2 



21356 8 

 22540-4 



Hg. 



5819 

 4078 



5819 

 3790 



3544 

 3039 



4916 

 3984 



1\. 



17184-9 

 24521-5 



17184-9 

 2()382-7 



2^219-5 

 32908-9 



203400 

 25099-9 



We get by Dr. Watts^ second method : 



(1) 



(2) 

 (3) 

 (4) 



99-6 

 100-4 

 100-5 



99-7 



Mean 100-05 



Instead of 100 I might just as well have chosen any other 

 value between say and 300, showing the method to be 

 entirely futile. 



Had Dr. Watts confined his calculation of the atomic 

 weight of radium to lines that we know to be homologous *, 

 he would have found that the two laws enunciated by him 

 are far from being fulfilled in the case of radium and the 

 related elements. The first law cannot even be said to be 

 approximately true for the following homologous lines : — 



Mg. 



Ca. 



Sr. 



Ba. 



Ra. 



2803 



3969 



4216 



4934 



4682 



2796 



3934 



4078 



4554 



3815 



2791 



3159 



3381 



3892 



3650 



For if it were true the radium lines ought surely to lie on 

 the less refrangible side of the barium lines, while in reality 

 one of the radium lines is even more refrangible than the 

 homologous line of calcium. As to the second law the 

 differences of the oscillation -frequencies of homologous 

 doublets are : — 



* See the paper bv J. Precht and the writer, Physikal ZeitscJiriff, 

 4 Jahrgang,, p. 285 (1903) ; Phil. Mag. [6] vol. v. p. 476 (1903). 



