SOME CONTEMPORARY ADVANCES IN PHYSICS-III 269 



how iiTtain properties of the waves vary along the spectrum. Enough 

 is known about the origin of eiectrotnagnelic waves to justify using 

 it as a basis of ciassitiration. Classifying the rays, therefore, by 

 mode of production, we can distinguish at least four sharply-contrasted 

 types: first, rays emitted from atomic nuclei in process of disintegra- 

 tion; second, rays emitted from atomic electron-systems in process 

 of rearrangement; third, rays due to atoms vibrating to and fro about 

 their positions of equilibrium as constituents of molecular groups or 

 of space-lattices; and finally, waves generated by oscillating electrical 

 circuits.' For each of these classes there is a region of the spectrum 

 which is particularly, although not exclusively, its own. 



The rays emitted from disintegrating nuclei lie at the topmost end of 

 the frequency -scale; they overlap the rays of the second class, but do 

 not approach either of the gaps. The rays resulting from rearrange- 

 ments of the electron-systems surrounding atom-nuclei e.xtend over 

 an enormous range. The minimum wave-length of this range is 

 .1075A, the X-frequency of the uranium atom; it is and will almost 

 certainly remain the definitive limit, unless someone should succeed 

 in discovering a substance further up the periodic table than uranium, 

 or in removing some of the deepest electrons from the electron-system 

 of some heavy- atom. As maximum wave-length we might take 

 that of a line 4050OA lately recognized by Brackett as belonging to 

 atomic hydrogen; but this is certainly not the definitive limit. Emis- 

 sion-bands due to atoms vibrating within molecular groups are found 

 in and beyond the "near infra-red" (and indeed in the ultra-violet 

 around 3000A, if we include bands of "compound" origin, resulting 

 from processes occurring together which if happening separately 

 would produce rays of the second and third types, respectively); 

 while the "residual rays," which are ascribed to atoms vibrating within 

 the gigantic molecular group which is a crystal lattice, extend as far as 

 0.152 mm. (residual rays of thallium iodide). Between 0.1 mm. 

 and 0.4 mm. rays have been discovered emanating from the mercury 



' This classification is obviously not an exhaustive one. Continuous spectra have 

 been omitted — thermal emission spectra of solids, and continuous X-ray spectra, 

 which may be as<rihcd to random accelerations of free electrons. The continuous 

 bands in gas spectra, of which one has just Ix^en e.xplained by Gcrlach (ZS. f. Phys., 

 18, pp 2.?'>-248; 1923) and others fjy Bohr (Phil. Mag. 26, p. 17; 1913)'. can \x 

 included in the second class by a slight generalization; and so, probably, can some 

 fluorescence and phosphorescence sf>ectra, at least if we extend "atomic electron- 

 systems" to include "electron-systems of grouped atoms." There is also the possi- 

 bility of rays due to changes in rate of rotation of molecules, not compounded with 

 changes in oscillation or electron-arrangement. 



' Meaning an atom with a large nuclear charge, which would have heaviness, or 

 more properly massiveness, as a secondary characteristic. A short and simple 

 adjective to descritte where an atom stands in the scale of nuclear charge, i.e. in 

 the periodic table, would be very welcome. 



