Febeuakt 17, 1922] 



SCIENCE 



165 



though probably masked by absorption. The 

 exploration of this region by methods not 

 strictly spectroscopic lies beyond the scope of 

 this paper but the subject is so fascinating 

 that I cannot resist a brief digression. 



The production and study of very soft 

 X-rays has occupied the attention of many 

 investigators here, in England and on the con- 

 tinent. One of the researches typical of the 

 early development of the subject was that of 

 Sir J. J. Thomson'^^ in 1914; one of the more 

 recent is that of Halweck.^" As the method 

 of this second investigator seems to trace a 

 rather direct path across the frontier of the 

 extreme ultra-violet it merits our attention. 



Halweck generated his soft X-rays in a 

 Coolidge tube of special form separated from 

 an ionisation chamber destined to measure 

 the absorption eofScients of the radiations by 

 an extremely thin window of celluloid. He 

 observes that the absorption coefficient of gases 

 increases with wavelength between 40 A.U. and 

 100 A.U. following an exponential law similar 

 to that observed by Owen for ordinary X-rays. 

 For celluloid the coefficient at first follows the 

 same law, then increases less and less rapidly, 

 passes through a maximum near 320 A.U. and 

 diminishes toward the ultra-violet. This work 

 is of particular interest since it traces a 

 physical property of a solid, its absorption, 

 through the "no man's land" in question. All 

 estimates of wave length are obtained from the 

 potential difference in the tube and the use of 

 the relation Ve = hv. 



Holweck finds that a film of celluloid a 

 quarter of a micron thick transmits about 

 twenty per cent, near 1000 A.U., while but 

 three per cent, will pass at the maximum of 

 absorption, this result is of some practical 

 importance to the spectroscopist. Miss Laird" 

 has also made a preliminary investigation on 

 the transmission of thin membranes. 



The study of radiation and ionisation poten- 

 tials affords another means of bridging the 

 gap between optical spectra and X-rays. 

 Begun by Franck and Hertz, and continued by 

 many investigators here and abroad with 



10 Phil. Mag., 28, p. 620, 1914. 



^^Comptis Rendu, 171, p. 849; 172, p. 439. 



^T Physical Bev., 15, 543, 1920. 



results of such fundamental importance, work 

 in this field has led Mohler and Foote^^ to the 

 discovery of characteristic soft X-rays pro- 

 duced by ares in vapor. They interpret their 

 results to mean that the critical potentials 

 which they have measured correspond to the 

 first L absorption lines of X-ray spectra of 

 the substances in question : for if the square 

 roots of the frequencies, computed from these 

 potentials, be plotted on a Moseley diagram, 

 the points will be very nearly on a continua- 

 tion of the straight line typical of the L X-ray 

 spectra of the heavier elements. The longest 

 wavelength they have measured in the L series 

 is for sodium, 353 A.U., magnesium yielding 

 268, and phosphorus 98. A critical potential 

 for potassium is interpreted as an M series 

 limit with a wavelength of 536 A.U. 



When we return to purely spectroscopic 

 investigation we find that the most striking 

 results have been recently obtained by Milli- 

 kan.^^ In an address before the National 

 Academy, in April, he reports the discovery 

 of certain lines of aluminium, magnesium and 

 sodium at 144.3, 232.2, and 372.2 A.U., 

 respectively, which he identifies as the La lines 

 of the X-ray spectra of these elements. The 

 square roots of the corresponding frequencies 

 lie very nearly on the straight line connecting 

 La frequencies and atomic number on the 

 Moseley diagram. The wavelength of the La 

 line for sodium is in qualitative agreement 

 with the value found by Foote and Mohler; 

 for aluminium and magnesium, however, the 

 case is not so clear. As Duane has pointed 

 out the difficulty arises in this way: it is a 

 fundamental property of X-ray spectra that 

 the La line lies on the long wavelength side of 

 the L absorption; now the L absorption for 

 almninium and for magnesium can be accu- 

 rately calculated; they lie at 173 and 257 

 respectively, but the La lines for these sub- 

 stances chosen by Millikan are at 144.3 and 

 232.2. Thus the values are on the wrong side 

 of the position of absorption. 



The spectra of aluminium, magnesium and 

 sodium occupy but a very limited region in 



isjow. Wash. Acad., 11, p. 273, 1921. 



See also Kurth, Fhy. Bev., 17, 528, 1921. 

 laProc. Nat. Acad., 7, 289, 1921. 



