348 



SCIENCE 



[N. S. Vol. XLVI. No. 1189 



We have shown, by experiments at Har- 

 vard, that this equation is not, in general, 

 true, but that it does hold for particular 

 eases. Mr. Hunt and I investigated the gen- 

 eral X-radiation from a Coolidge tube, ex- 

 cited by a high potential constant voltage 

 storage battery, using an X-ray spectrom- 

 eter, and found that although the effective, 

 or average, frequency does not obey the law 

 represented in the equation (1), the equa- 

 tion does give the maximum frequency ob- 

 tainable with a given electron energy. Dr. 

 "Webster then examined the characteristic 

 X-radiation, and discovered that the kinetic 

 energy of the electrons required to produce 

 the alpha and beta lines of the K series is 

 larger than is represented by equation (1), 

 but that the gamma line (the highest fre- 

 quency line in this series) approximately 

 obeys the law. It appears, therefore, from 

 our experiments, that equation (1) gives 

 the maximum frequency of the radiation 

 due to an electron's hitting an atom, but 

 does not, in general, mean that the entire 

 amount of the electron's energy is radiated 

 at frequency v. 



I have recently shown that it is not neces- 

 sary to assume that energy is radiated in 

 quanta "h" in order to deduce equations 

 for the distribution of energy in emission 

 spectra similar to the equations represent- 

 ing black body radiation, so that we are not 

 compelled to believe that because Planck's 

 radiation law fits the facts of black body 

 radiation more or less closely, therefore 

 energy must be radiated in quanta hv. In 

 attempting to explain why this constant 

 "h" enters into the radiation law and in 

 seeking for a physical conception of the 

 mechanism of radiation, we are not there- 

 fore compelled to explain the emission of 

 radiation in quanta hv, but rather the fact 

 that an electron with a given kinetic energy^ 

 when it hits an atom, can produce radia- 

 tions of frequency up to but not greater 

 than that given by equation (1). This is 



the fundamental fact that needs explana- 

 tion. 



According to the modern conception of 

 the constitution of matter, an atom pos- 

 sesses a complicated electro-magnetic struc- 

 ture in which the electrons play an impor- 

 tant role. The electro-magnetic forces in 

 this structure are greater near its center 

 than at the periphery, and therefore the 

 high frequency vibrations of the electrons 

 must be associated with parts of the atom 

 near its center. Hence, the reason why an 

 electron can not produce a high frequency 

 radiation unless it possesses a certain ki- 

 netic energy lies in the fact that it does not 

 penetrate far into the atom unless it has a 

 sufficient speed. This presupposes a force 

 of repulsion between the electron and the 

 atom. The theory of atomic structure 

 seems to demand such a force in order to 

 explain why atoms do not collapse; so that 

 we have confirmation of the existence of 

 such forces from two sides: the radiation 

 and the structure of matter. 



Before discussing further the nature of 

 this force and the laws it must obey, I 

 would like to present to you a conception 

 of the difference between line spectra and 

 the general, or continuous spectra. The 

 frequencies of the characteristic lines de- 

 pend upon the nature of the atoms struck 

 by the electrons, whereas the frequencies 

 of the general radiation depend upon the 

 kinetic energy of the electron that does the 

 striking. This suggests that the character- 

 istic lines are due to vibrations of parts of 

 the atoms themselves (of electrons in the 

 atoms, for instance) whereas the general 

 radiation or continuous spectrum is due to 

 the vibrations of the electrons that hit the 

 atoms. 



The question now arises "How can an 

 electron vibrate with all possible frequency 

 so as to give a continuous spectrum?" 

 The electron moves in the strong electro- 

 magnetic field of the atom, and when an 



