938 Prof. Horton and Miss Davies on Electron Velocities 



viz. those which follow the Ritz interpolation formula with 

 great exactitude, without any modification of the combination 

 principle, and those which require either a modification of 

 this principle or a modification of the Ritz expression. In 

 the paper already referred to, Paschen gave the extension of 

 the Ritz interpolation formula, in the case of each series of the 

 second tvpe, which was necessary in order that the combination 

 principle might hold. In a more recent paper * he has shown 

 ■that it' the combination principle is modified by the addition 

 of a constant to the combination values of all terms of ihese 

 series, the values of the terms so obtained can be expressed 

 by formulae of the exact Ritz type with greater accuracy than 

 by the formulae given earlier. The series to which \6678 

 and X5298 belong are of the type which requires this modi- 

 fication of the combination principle : but the series to 

 which the other lines belong were satisfactorily expressed in 

 the earlier paper. Thus both types of series are represented 

 in the spectrum of the luminosity produced at 20'0 volts and 

 in the additional lines which appear when the ionization 

 velocity of 22*8 volts is passed. 



The fact that the combination principle holds without any 

 modification between some of the lines excited at 20*0 volts 

 and some of those excited at 22'S volts, rules out the possi- 

 bility of these two ionization velocities corresponding to the 

 ionization of two distinct elements. "We" are thus forced to 

 the conclusion that the different ionization velocities which 

 have been detected correspond to the removal of differently 

 situated electrons from the neon atom, ;md hence that lines as 

 closely related as those of a principal and the corresponding 

 subordinate series arise from the return of electrons removed 

 from different positions within the atom. 



From the position of argon in the Periodic Table of the 

 elements, its atom would be expected to have an arrange- 

 ment of external electrons bearing some resemblance to that 

 occurring in neon. We should therefore expect that argon 

 would have more than one ionization velocity corresponding 

 to the removal of a first electron from its atom, and that its 

 spectrum could be excited in stages. We have not yet 

 investigated the minimum electron velocities required for 

 the excitation of spectrum lines in argon, but our expe- 

 riments on the ionization of this gas gave no indication 

 of the existence of more than one critical electron velocity 

 for the production of ionization, and nothing is recorded 

 in the work of other investigators which suggests that 



* F. Pasclien, Ann. der Pfojsik, vol. Ixiii. p. 201 (1920). 



