Mat 21, 1920] 



SCIENCE 



505 



B. Quantum Emission Phenomena-Electrons, 

 by R. A. MiLLiKAN, University of Chicago. 

 Experiments on the potentials necessary to 



apply to just prevent the escape of photo- 

 electrons from different metals, combined with 

 measurements on the contact E.M.F.'s be- 

 tween the same metals demonstrate, not that 

 the energy absorbed by an electron which is 

 being photoelectrically ejected is hv, but 

 rather that the hinetic energy with which the 

 electron escapes from the atom under the in- 

 fluence of incident radiation of frequency v is 

 hr, and hence that the absorbed energy must 

 alivays exceed hv by the amount of the worTc 

 necessary to detach the electron from the 

 atom. In other words the absorption of 

 energy can not take place quantum-wise at all 



The energy which must be absorbed to 

 cause the escape of an electron must always 

 be more than a quantum and may exceed that 

 amount by any fractional part thereof. It 

 is only the kinetic energy of the escaping 

 electron which is abvays an exact quantum. 



The emission of electromagnetic radiation 

 may or may not take place quantum-wise. 

 Characteristic or fluorescent radiation appears 

 to be emitted quantum-wise but scattered 

 radiation is not so emitted. 



C. Spectrum Series, by William Duane, Har- 

 vard University. 



There is presented for discussion the fol- 

 lowing four topics : (a) The combination law 

 applied to the series spectra of ordinary light, 

 and a somewhat similar law for characteristic 

 X-rays; (6) the meaning of these laws in 

 terms of the energy of the radiating atom 

 from the point of view of the theory of radi- 

 ation in quanta; (c) the law of constant 

 angular momentum as used by Bohr in ex- 

 plaining some of the details of series spectra; 

 (d) Bohr's theory applied to characteristic 

 X-rays, with special reference to critical ab- 

 sorption data. 



(a) It has been found empirically that the 

 frequencies of vibration (or the wave num- 

 bers) corresponding to the lines in series 

 spectra may be represented as the differences 

 between any two of a set of terms, which may 



be denoted by symbols of the forms (IS), 

 (IP), (mP), (mD), {mF), (Is), (m/), etc. 

 These terms may approximate to, but do not 

 exactly equal a certain constant divided by 

 the square of a whole number. The differ- 

 ences between the correct values of various 

 pairs of terms, however, appear to represent 

 the wave numbers with great precision. 



Turning to X-rays we flnd that the form 

 of the " terms " is much more complicated. 

 It is possible, however, to obtain empirically 

 a relation between the X-ray emission and ab- 

 sorption frequencies that resembles the above 

 combination law. Mr. Shimizu and I recently 

 published^ the results of experiments, which 

 indicate that an emission frequency equals 

 the difference between two absorption fre- 

 quencies. From the data for the K and L 

 series of tungsten, it appears that the law is 

 correct to about one fifteenth of a per cent. 

 Since we presented this research to the Physr 

 ical Society some measurements by Dr. Sten- 

 strom of the absorption frequencies in the M 

 series of uranium and thorium have come to 

 hand, and these together with de Broglies' 

 values for the L absorption frequencies and 

 Seigbahn's values for the L emission fre- 

 quencies furnish data by which the law can 

 be tested. The calculations indicate that the 

 differences between the L and JVI absorption 

 frequencies equal the frequencies of some of 

 the L emission lines to within one per cent. . 



Theories of the mechanism of radiation 

 such as that suggested by Bohr lead to laws 

 similar to the combination law, and Kossel 

 has deduced from these conceptions relations 

 between the emission frequencies themselves. 

 One of these relations is that the difference 

 between the 'KjB and the Kot frequency equals 

 the Let frequency. This relation is not exact, 

 however, for Ko; represents a group of lines 

 and recent experiments have shovtm that K|8 

 also is not a single line. We get a much 

 closer agreement, if we take the frequencies 

 of the individual lines in the groups. 



(6) For a long time spectrum analysis re- 

 mained a purely descriptive science, contain- 

 ing data of extraordinary accuracy, it is true, 



1 Physical Beview, July, 1919. 



