December 29, 1916] 



SCIENCE 



901 



spectral energy distribution of black-body 

 radiation from the assumption that the 

 radiation was emitted and absorbed by 

 electric oscillators in definite quanta, each 

 equal to the frequency of the oscillator 

 multiplied by a universal constant, h, the 

 wirkungsquantum. Later he modified this 

 theory so far as absorption is concerned. 

 Einstein and others went further in as- 

 suming that these quanta preserve their 

 identity in their propagation through 

 space, thus reviving a form of corpuscular 

 theory. This extreme view has been gen- 

 erally abandoned, but it has been found 

 impossible to explain away the wirkungs- 

 quantum h. It appears in too many rela- 

 tions to be the result of chance. The work 

 of Millikan in particular proves the exact 

 validity of Einstein 's relation V e = 

 h(v — v ) in the photoelectric effect, in 

 which Ve is the measure of the emission 

 energy of the electrons, v the frequency 

 of the incident light, and v the minimum 

 frequency which will cause emission of 

 electrons. A similar relation appears to 

 hold good in many cases of X-ray and light 

 spectra. It seems probable that this con- 

 stant depends upon atomic structure only, 

 and affects radiation through space only 

 in so far as emission and absorption are 

 determined by atomic structure. 



The theory of the nucleus atom is like- 

 wise of fundamental importance in spec- 

 troscopy. The work of Rutherford and 

 others leaves no escape from the conclu- 

 sion that the nucleus of the atom is a con- 

 centrated group of positive charges and 

 electrons, with an excess of positive ele- 

 mentary charges approximately equal to 

 half the atomic weight, while the same 

 number of electrons circulate about the 

 nucleus in rings. The spectroscopist must 

 try to fit his theories to these probable 

 facts, but he is met at the outset with ap- 

 parently insuperable difficulties in ac- 

 counting for the stability of such atoms and 



for the manifold complexity of spectra ac- 

 cording to accepted electrodynamical laws. 

 Bohr cut the Gordian knot by supposing 

 that the classic laws apply only to condi- 

 tions of stability, when no energy is radi- 

 ated, and that radiation attends the transi- 

 tion of an electron from one state of 

 stability to another, the frequency being 

 determined by the relation that h multiplied 

 by the frequency is equal to the difference 

 between the energies of the system in the 

 two stable states. In the case of hydrogen, 

 to which he assigns one radiating electron 

 and one nucleus charge, it is difficult to 

 account for the existence of so many stable 

 states, for the failure to radiate while sub- 

 ject to uniform radial acceleration, and for 

 monochromatic radiation while passing be- 

 tween two positions of stability. Never- 

 theless Bohr derived an expression like that 

 of Rydberg which locates accurately not 

 only the Balmer series, but also an infra- 

 red and an ultra-violet series predicted by 

 Ritz and found by Paschen and by Ly- 

 man, respectively. His attempt to apply 

 the same method to helium led to results 

 which are still in dispute, and which will 

 be referred to later. 



In reviewing recent progress we may be- 

 gin with that field in which this country 

 has taken a leading part — that of astro- 

 physics. This domain belongs as much to 

 the physicist as to the astronomer. The 

 heavenly bodies are laboratories on a vast 

 scale, in which nature has provided condi- 

 tions of temperature, pressure and elec- 

 trical state which we may never hope to 

 rival on the earth. The spectroscope gives 

 us data from which it may be possible to 

 form some idea of these conditions by com- 

 parison with our feeble laboratory imita- 

 tions of celestial phenomena, and con- 

 versely, the latter may aid in the interpre- 

 tation of terrestrial phenomena. 



One of the most fruitful astronomical 

 applications of the spectroscope is to the 



