SOME CONTEMPOR.-IRY ADVANCES IN PHYSICS— III 281 



or to a moleculo. Tlu- oxiK'rimental material is al)un<lanl" cnongli 

 to k;ive practical certainty that "valence-electron rays" Ix-low lOOOA 

 occur in the six'ctra of only a few elenu-nts, and Ik'Iow oOOA in none. 

 Nevertheless, Millikan and Bowen, photographing the spectra of 

 all of the first twenty elements (neon and argon excluded, and chro- 

 mium and copper added) down to the extremity of the region ac- 

 cessible with the concave grating, discovered great numbers of lines, 

 of which they attribute dozens or scores to particular elements (for 

 example, some forty lines ascribed to potassiimi, though its ionizing- 

 potenti.il of four volts corresponds to a minimum wave-length exceed- 

 ing 2,")0()A). Some of these may be lines of compound origin, resulting 

 from two simultaneous changes in the electron-system of the atom, 

 one being a transition of the valence-electron and the other a re- 

 arrangement of the other electrons. (Saunders mentions such lines 

 in the s|)ectra of elements of the second column of the table.) Others 

 are due to rearrangements of internal electrons following upon a dis- 

 placement of one of these. Many others are attributed to displace- 

 ments of the valence-electrons of ionized atoms. Of this new field 

 of research, the spectroscopy of ionized atoms, I wrote briefly in the 

 first article of this series. In the more easily accessible regions of 

 the s()ectrum, F'aschen had discovered ra\s of doubly-ionized aluminium 

 and Fowler rays of trebly-ionized silicon."' Millikan and Bowen go a 

 step further by identifying certain rays of quadruply-ionized phos- 

 phorus: indeed they believe that, under the \iolent excitation provided 

 by their vacuum-spark, the waves emitted by atoms which have lost 

 all but o.ne of the electrons from their outermost electron-shells (the 

 three just specified are in this state) are especially abundant and 

 intense. 



'* In the pcrlotlic tabic of Fig. i the ionizing-potentials of the eleineiUs arc given 

 along with their atomic nunil}crs. Overlined figures are values calculated from 

 series-limits and confirmed by direct experiment; starred values are data of experi- 

 ment, for elements of which the series have not l)een worked out; the remaining 

 values are calculated from series-limits and have not been verified. The data are 

 from the cited sources, from Frxite and Mohlcr, and from Saunders' graphic tabula- 

 tion l.SViVncf, volume 50, pp. 50-51, January 18, 1924;). Interesting variations 

 with atomic numlier are observe<l which yield bases for estimating the values for 

 still other elements by interpolations. 



""It might l)e mentioned that the spectrum of silicon was first selected for in- 

 vestigation on account of its astrophysical interest. The lines representing suc- 

 cessive stages of ionization of this element ap|H'ar in stars which there is every 

 reason to Udieve arc at successively higher tem[)eratures. The complete series- 

 data for the four spi'ctra |trcbly,- doubly-, once- and non-ionized atoms] and the 

 ionization-|K>tentials de<luce<l from them, may Ik- exixxted to find an important 

 application in fixing the scale of stellar temperatures . . . .-Ml the series predicted in 

 (these four) sjKttra of silicon; in the spectra of doubly-, once-, and non-ionized 

 aluminum, of once-ionized and neutral magnesium, and of neutral sodium, have 

 been actually produced and have been found to have the character and constants 

 expected." — A. Fowler. 



