TRANSITION TO A NEW THEORY 205 



ticular orbits, the classification of the spectral lines must 

 run parallel with the classification of the orbits by their 

 quantum numbers in the model. When the spectro- 

 scopists started to unravel the various series of lines in 

 the spectra they found it possible to assign an orbit 

 jump for every line — they could say what each line 

 meant in terms of the model. But now questions of 

 finer detail have arisen for which this correspondence 

 ceases to hold. One must not expect too much from a 

 model, and it would have been no surprise if the model 

 had failed to exhibit minor phenomena or if its accuracy 

 had proved imperfect. But the kind of trouble now 

 arising was that only two orbit jumps were provided 

 in the model to represent three obviously associated 

 spectral lines; and so on. The model which had been 

 so helpful in the interpretation of spectra up to a point, 

 suddenly became altogether misleading; and spectro- 

 scopists were forced to turn away from the model and 

 complete their classification of lines in a way which 

 ignored it. They continued to speak of orbits and 

 orbit jumps but there was no longer a complete one- 

 to-one correspondence with the orbits shown in the 

 model.* 



The time was evidently ripe for the birth of a new 

 theory. The situation then prevailing may be summar- 

 ised as follows : 



(1) The general working rule was to employ the 

 classical laws with the supplementary proviso that 

 whenever anything of the nature of action appears it 



*Each orbit or state of the atom requires three (or, for later refine- 

 ments, four) quantum numbers to define it. The first two quantum 

 numbers are correctly represented in the Bohr model ; but the third 

 number which discriminates the different lines forming a doublet or 

 multiplet spectrum is represented wrongly — a much more serious failure 

 than if it were not represented at all. 



