November 3, 1923J 



NA TURE 



655 



The Origin of Optical Spectra. 



AMONG the many remarkable communications 

 made this year to Section A (Mathematics and 

 Physics) of the British Association, which, grouped 

 together, will probably mark it off as an outstanding 

 meeting, the address by the sectional president. Prof. 

 McLennan, on the origin of spectra, was not the least 

 interesting. From among the many subjects he 

 surveyed it may be of interest to select some, and to 

 try to give a not too technical account of these, show- 

 ing the sort of progress that is now being made under 

 the stimulus of Prof. Bohr's theory. 



We agree now that all spectra are emitted by 

 atoms or molecules during the process of return to 

 their normal state after a more or less violent dis- 

 turbance, and that any particular spectrum is emitted 

 only by a particular atom or molecule after a suitable 

 disturbance. We agree too (partly for theoretical, 

 partly for experimental reasons) that spectra can be 

 divided into two distinct types-^-line spectra or series 

 spectra and band spectra or many-line spectra — which 

 have their origin in the reconstruction of atoms and 

 molecules respectively. It is with recent advances 

 in the more advanced and more important study of 

 these atomic or line or series spectra, emitted during 

 the reconstruction of atoms, that the president dealt, 

 and with these only shall we be concerned here. 



Physicists will agree that an atom consists of a very 

 small massive nucleus of positive electric charge Z 

 units, the unit being the charge on the electron, sur- 

 rounded by a planetary system of Z electrons. These 

 move, when undisturbed, as a conservative system in 

 a set of orbits which must have a definite structure, 

 controlled by laws of which we are not yet masters, 

 to which, however, the present quantum theory gives 

 the most complete expression yet achieved. The 

 number Z is called the atomic number of the atom, 

 and specifies its place in the periodic table and all its 

 physical and chemical properties. We can agree 

 further that the orbits of the Z electrons are not all 

 essentially different. They can be classified in groups, 

 orbits of which are characterised by the same values 

 of certain integers (three to each orbit), commonly 

 called quantum numbers. 



There are a variety of disturbances to which such 

 an atom can be subjected. By suitable means supply- 

 ing sufficient energy we can shift one or more of its 

 electrons from their normal orbits, either right out of 

 the atom, or into other possible orbits characterised 

 by different quantum numbers. In the subsequent 

 reconstruction the atom will emit a spectrum of sharp 

 lines of definite frequencies characteri.stic of itself 

 and the particular disturbance it has suffered. Each 

 separate line is emitted during the return of an electron 

 from one particular permissible orbit to another of 

 less energy, and its frequency is related to these orbital 

 energies by the most fundamental equation of the 

 quantum theory Yjy-Y.^ = hv. After the partial 

 removal of a particular electron we merely get part 

 of the spectrum corresponding to complete removal 

 of the 3am(; electron. We can therefore, speaking 

 generally, classify the complete line spectrum of a 

 given atom into a number of separate spectra, each 

 of which is associated with the recapture of one electron 



NO. 2818, VOL. I I 2] 



by an atom after the removal of any specified set of 

 its original Z electrons. Classified thus, an atom's 

 spectra will divide into two well-marked types — those 

 in which one or more of its deeper lying electrons 

 have been removed and those in which the electrons 

 removed, whatever their number, are entirely those 

 most lightly bound. In the first type we can and do 

 find internal reorganisations taking place before a 

 new electron is captured. These are the X-ray spectra, 

 with which we are not here concerned. In the 

 second type no such reconstruction can occur, except 

 while the new electron is being brought in. These 

 spectra, which theoretically must all be of the same 

 general series type, are called the optical spectra of the 

 atom. 



The typical optical spectrum (the so-called arc 

 spectrum) of an atom is agreed to be that which is 

 emitted during the return of the last (Zth) electron 

 to an atom in which the rest of the system is in its 

 normal state. When such a spectrum is fully analysed 

 it is found that the lines can be arranged in series 

 which display a certain fundamental constant R, 

 Rydberg's constant. The value of this constant and 

 its perpetual occurrence in all arc spectra is (as is well 

 known) properly predicted by the theory. But this 

 is not all. If we call the ordinary arc spectrum Z(I) 

 and its Rydberg's constant R, the theory we have 

 outlined predicts Z optical spectra in all, of which the 

 Qth spectrum Z(Q), with constant Q^R, will be emitted 

 by the atom with its first (Z-Q) electrons in their 

 proper orbits as it catches its (Z - Q + i)th electron. 

 The characteristic frequencies of these spectra will, 

 of course, get higher and higher as Q increases, and for 

 the later " optical " spectra of a heavy element will 

 lie in the X-ray region. It is not the frequency range 

 but the type of spectrum which remains characteristic- 

 ally optical. 



The predicted second optical spectra Z(II), with 

 Rydberg constant 4R, haye been known for some years 

 for a number of elements, under the general name of 

 spark spectra ; until recently we have had no experi- 

 mental confirmation for values of Q greater than 2. 

 In the last year there has been a great advance, for 

 the third optical spectrum of aluminium with constant 

 9R has been obtained by Prof. Paschen, and the fourth 

 and parts of the third optical spectra of silicon with 

 constants 16R and 9R respectively by Prof. Fowler. 

 These spectra are known by the very convenient 

 notation of AlIII, Si IV and Silll. It will be seen 

 that the spectra Si IV, AlIII, Mgll and Nal are all 

 concerned with the capture of the eleventh electron 

 by an atom (of varying Z) which has already bound 

 its first ten electrons in their permanent orbits. These 

 four spectra should be and are of the greatest similarity 

 in their finer details. Their further detailed compara- 

 tive study should be fruitful. 



Prof. McLennan also pointed out that this successful 

 study should throw light on the various optical spectra 

 of the analogous series of elements, lithium, ber)'llium, 

 boron, and carbon. In this difficult and very important 

 region little progress has liitherto been made, but Prof. 

 McLennan seemed hopeful that, with the theoretical 

 and comparative guides now available, a renewed 



