14 Mr. H. Eamage. Relations .of Spectra, Densities, [Xov. 7, 



Intensities. 



Wave -length.. | 



lli fr 



W~ o 



Gallium. 



4172 -21 \ 

 4033 -12 J 

 2943 1 



2/1 

 2873 J 

 27801 

 2718 J 



24961 broad 

 2447 J 



10 

 9 



10 

 9 

 3 

 4 

 4 

 3 

 2 

 2 

 1 



Intensities. 



WaTe-lensth. 



Thallium. 



5350 -65 \ 

 3775 -87 I 

 3529 -581 



19 "39 J 

 2767 -97 

 3229 -88 

 2580 -23 

 2921 -63 1 



18-43 J 

 2379 -66 



II 



10 



} = 



t>5 



M 



o 



10 

 10 



3 



4 



3r 



3 



lOr 

 10/- 



8r 

 lOr 

 10r 

 lOr 



8r 



6r 

 lOr 



8r 



Diagrams haye been drawn with other functions of the atomic mass, 

 but they have not given results of any great value. 



The diagrams reproduced in Plates 1 and 2 were drawn from the 

 oscillation frequencies of the lines and (1) the atomic masses, (2) the 

 squares of the atomic masses. The symbols indicate the positions of 

 the lines in the spectra. 



The following facts have been observed in the study of the 

 diagrams : — 



(1.) The metals considered may be classified into groups, according 

 to their spectra. The fact is not new, but the diagrams make 

 its truth most obvious. 



(2.) The connecting lines between the members of the groups are 

 not continuous : there are certain breaks in them. There is a 

 break in the group of the metals of the alkalies between sodium 

 and potassium. Another break occurs between the triplet of 

 magnesium and those of zinc, cadmium, and mercury. 



There are members of two subordinate series in the spectra of the 

 aluminium group — the sharp and the nebulous series. The break 

 between aluminium and the other metals is very marked in the diffuse 

 series, while there is only a slight irregularity in the lines connecting 

 the sharp series. This fact will be referred to again when considering 

 the physical properties of the metals. 



(3.) The cause of the displacement of corresponding lines in some 

 strictly homologous spectra is intimately connected with the 

 change in the atomic mass. The lines are moved towards the 

 less refrangible end of the spectrum with an increase in the 

 atomic mass. The shift of the lines due to this cause is most 



