688 



SCIENCE 



[Vol. LVI, No. 1459 



to vacuum, Rydberg's values of N/(»» + \i.)^, 

 data for wave-lenigtihs and wave-numbers of 

 lines allocated to series, and the Hicks formula 

 constants, in Chapter V in which he considers 

 the effects of physical conditions, the various 

 Zeeman patteruB so far determined, illustrative 

 of Preston's law, are collected for oonvenient 

 reference and also the Stark data, and a teble 

 of radiation and iomzation potentials is 

 included. 



Chapters II and III give a clear introduc- 

 tion to types of series and to the series systems 

 now recognized in the different gi-oups of the 

 periodic taMe. In the following chapter Ryd- 

 berg's rules are discussed in the light of the 

 series systems now determined and of the avail- 

 able spectroscopic data. With the exception 

 of the constancy of N, he considers that ail the 

 values are valid; as to N he concludes that we 

 can not hope to determine exact values of N in 

 the various series and elements from deter- 

 mination of formulae constants alone, that it is 

 practically certain that the value in general is 

 larger than Rydiberg's value and nearer Bohr's 

 limit and that changes in N dn the different 

 sequences should not be unexpected. 



A distinctive contribution by Hicks to the 

 study of series is his modification of the Ryd- 

 berg formula which he writes « = A — N/ 

 (m + [I. + a/m)2 where A is the lim.it, 

 [J. + a/m the "mantissa," and N 109675 R. U. 

 In his notation the separations from the 

 limit in a series of lines form a "sequence" 

 of values and a particular value is the 

 mth "sequent." Separation is the differ- 

 ence in the wave-numbers of two lines, v = 

 donhlet separation, A = the difference in the 

 mantissse of doublet sequences and is shown to 

 be an integral multiple of the "oun." 



The mantissfe play an important role in the 

 author's development of the idea of the oun 

 and "linkages." He writes A = m g w^ where 

 m is an integer, q a universal constant, and w 

 the atomic weight, which for convenience he 

 divides by 100. From a consideration of the 

 doublets of Ag he finds that q is 361.78. 



The O'uin is considered to be a fundamental 

 constituenit in spectra and is wiiitten oun = 5, 

 = %S where S = 361'U)-. The evidence given 



foa- the dependence of the oun on the square of 

 the atomic weight is very strong, but the ex- 

 istence of isotopes introduces difficulties which 

 the author recognizes but does not succeed in 

 disposing of completely. 



When a multiple of the oun is added to the 

 limit mantissa, the modification is called dis- 

 placement and a displaced line a "collateral." 

 This leads to his theory of linkages, "that in 

 certain elements the spark spectra consist 

 almosit wholly of Ibng sets of lines differing 

 fix)m one another in succession by cei'tain spe- 

 cial separations which can be calculated from 

 the ordinary series limits 'and A, these separa- 

 tions may be called links, and a complete set a 

 linkage. These linkages appear to start from 

 ordinary series lines." The sutojedt of linkages 

 is too complicated for presentation in a review ; 

 it aims to relate the lines which do not heilong 

 to the regular series to lines occurring in the 

 ordinary series by a set of links. These links 

 may occur in any order forming chains and 

 meshes of lines. That "linkage spectra" repre- 

 sent realities and form a category coordinate 

 with series and band speotra would seem to be 

 a matter for further investigation. 



The reviewer's impression is that ;the author 

 over-estimaltes (the accuracy of the spectro- 

 scopic data and one wonders whether with more 

 accurate data .the evidence for linkages would 

 be increased or lessened. Professor Hicks is 

 convinced that links occur with much greater 

 frequency than a chance arrangement would 

 suggest, but eonsidei-s the most conclusive evi- 

 dence to be furnished by the large number of 

 regularities, repetitions, collocations of links 

 and meshes. It is, however, somewhait sur- 

 prising that lines in spark spectra should be 

 related to each other by separations calculated 

 from the limits and the A of the ordinary 

 series inasmuch as spark spectra are charac- 

 terized by their special types of series. 



Chapters VIII and IX are devoted to a dis- 

 cussion of the distinguishing properties of the 

 p and s and of the d and / sequences. For the 

 p sequence the march of the mantissee and 

 atomic volumes is so close that he concludes 

 that the p sequence depends directly on a quan- 

 tity equivalent to the volume of the atom. On 



