April 5, 1917] 



NATURE 



II 



about 0-3 mm., while the shortest Hertzian waves so 

 tar obtained are 2 mm. long. The study of line radia- 

 tion in this region is even more difficult, but Paschen 

 and his pupil, the American Randall, have succeeded 

 in measuring many lines extending to about 90,000 

 Angstrom units. 



In the ultra-violet Lyman has extended the region 

 tirst made known to us by Schumann to a wave- 

 length of about' 600 Angstrom units. Beyond this 

 point it is difficult to go, on account of absorption, 

 lack of sensitiveness of the photographic plate, and 

 small reflecting power of speculum metal. Gratings 

 ruled on silicon and photoelectric detectors may enable 

 - to bridge the gap between these waves and the 

 jch shorter ones which may be examined with the 

 1 of nature's diffraction gratings, crystals which 

 \e made the study of X-ray spectra possible. 

 Of all the discoveries of recent years, that of the 

 vsave nature of the X-rays and of a practical method 

 m examining their spectra is the most remarkable 

 and the most important, for it has revealed to us the 

 most fundamental radiations of the elements and has 

 :;iven us a glimpse into the very heart of the atom. 

 In quick succession Laue and his pupils demonstrated 

 '- diffraction effects produced by crj'stals, the Braggs 

 owed how reflection might be employed to isolate 

 vaves of different lengths by a principle similar to 

 that producing colours of thin plates, but of far greater 

 r'solving power by reason of the greater number of 

 etlective reflecting surfaces, and Moseley photographed 

 many characteristic spectra by an extraordinarily 

 simple method. He found that the principal lines in 

 the spectra of a large number of elements were con- 

 nected by a remarkably simple relation, namely that 

 the square roots of the frequencies are proportional 

 to the ordinal numbers, which increase by one in 

 passing from one number of a periodic group to the 

 xt. When there are anomalies between the atomic 

 ight and the place of an element in a group, this 

 omaly disappears when the atomic number rather 

 m the atomic weight is considered. This work 

 - been extended by others, notably by Siegbahn 

 A Friman, to include nearly all the known elements 

 'ween sodium and uranium, inclusive, with the 

 -Lilt that all the atomic numbers between hydrogen 

 aid uranium are accounted for, with the exception of 

 >;x gaps. As interpreted by Bohr's theory, the ordinal 

 number which determines the frequency is the excess 

 number of positive elementary charges in the nucleus, 

 and these results are, therefore, in complete harmony 

 with the theory of the nuclear atom develop>ed by 

 Rutherford, van den Broek, Soddy, and others. The 

 comparison, of the X-ray spectrum of lead obtained by 

 Siegbahn with the gamma-ray spectrum of radium B 

 obtained by Rutherford and Andrade shows the iden- 

 tity of ten of the principal lines. This strikingly con- 

 firms the accepted theory of isotopes, or elements of 

 different atomic weights, which are chemically and 

 spectroscopically alike beause they have the same re- 

 sultant nuclear charge. 



The positions of the principal lines are consistent 

 with Bohr's general formula, but perhaps this rela- 

 tionship is purely formal. But whether or not this 

 theory applies, apparently we cannot dispense with 

 the wirkungsquantum. In addition to the character- 

 istic X-radiation of an element, there is a continuous 

 spectrum, with a sharply defined boundar\- on the side 

 of shorter wave-'eneths. The investigations of Duane, 

 Hull and D. L. Webster have shown that this boundary' 

 is accurately defined by Einstein's relation Ve = hv 

 for fields up to 110,000 volts. Such a simple law 

 does not hold for the characteristic radiations ; but 

 Webster has shown that they do not aopear until the 

 voltage somewhat exceeds that demanded by the Ein- 

 NO. 2475, VOL. 99] 



stein relation. The k>ngest X-waves so far discovered 

 by Siegbahn are about 12 Angstrom units in length, 

 so that there is not a very great gap between them 

 and the shortest waves discovered by Lyman. The 

 investigation of this region is difficult, but undoubt- 

 edly means will be found to attain success. .Much 

 also remains to be done in the study of details of 

 X-ray spectra, which contain many weak lines, and 

 possibly bands, which have not so far been carefully 

 examined. 



During the past ten years great advance has been 

 made in our knowledge of spectral series. Rydberg, 

 Ritz, Paschen, Fowler and others have shown that a 

 I generalised form of the Balmer equation, with Ryd- 

 berg's universal constant and a few special constants, 

 is capable of wide application. Different combinations 

 of a few constants have been found to give a number 

 of related series, and many new lines so predicted have 

 been found. The common limit and other numerical 

 relationships between different series of the same 

 element indicates that the different emission centres 

 I have some dynamic coupling and Rydberg 's universal 

 j constant indicates a structural element common to all 

 I substances. According, to Bohr, this quantity is a 

 function of the electronic and atomic mass, the ele- 

 mentary electrical charge, and the wirkungsquantum 

 h, and should slightly increase with increasing atomic 

 weight. As it is commonly assumed that it is an 

 absolute constant, careful measurements may furnish 

 a test of the validity of Bohr's theor}'. 



The relationships of frequency to atomic number 

 found by Moseley recalls that Ramage, Watts, Runge 

 and Precht and Hicks have found linear relationships 

 between the squares of the atomic weights and the 

 frequencies or frequency differences of homologous 

 lines in the spectra of elements of the same group. 

 Ives and Stuhlmann have shown that in some cases 

 the results are improved by substituting atomic 

 numbers for atomic weights, but the relationship is 

 evidently not so simple as in the case of X-ray spectra. 

 The discovery of the Zeeman effect and the ex- 

 planation of its simpler forms by Lorentz was the first 

 step toward a rational spectroscopic theory. The later 

 discovered complexities and anomalies, while they may 

 defy mathematical analysis, do not lessen our con- 

 fidence in the theory, for they are what we might 

 expect as a result of complicated atomic structure. 

 The same intellectual satisfaction does not attend the 

 discovery of the analogous effect of an electric field, 

 because the simplest cases are so complex that they 

 cannot be adequately explained by any theory yet 

 proposed. The possibility of such an effect had long 

 been the subject of speculation, but Stark was the 

 first to realise and attain the necessary conditions for 

 its occurrence. Lo Surdo also discovered it in the 

 neighbourhood of the cathode in capillary tubes. As 

 in the case of the Zeeman effect, the phenomena are 

 different when viewed transversely and parallel to the 

 field. In each case the lines are split into a number 

 of components, the number being different for different 

 lines, even for those belonging to the same series. 

 In the transverse effect the components are plane- 

 polarised in hydrogen and helium, the stronger central 

 lines vibrating at right angles to the field, and the 

 stronger outer components vibrating parallel to the 

 field. A remarkable relation is found for the series 

 lines of hvdrot^en, helium and lithium. For each the 

 number of principal normal components appears to . 

 be equal to the ordinal number of the line in the series. 

 Higher dispersion shows that in the case of hydrogen 

 each component is double. If this rule holds good 

 throuerhout the series, the last know-n line, the twenty- 

 eighth, would have fifty-six such components, an eaual 

 number polarised at right angles to these, and a 



