soMii coNTEMi'tu^.iKy .//>r./.V( 7; v /.v riiysus iiii a\» 



and ihtTi-forc I sli.ill Niolati' llu' ruli- I li.i\r luTfiufnrf followed, .ind 

 |M)stjH)nc the (U'scripiion of tlu- plifiionu'iia until tin- tlifory is st.iti-il. 

 Atoms of otluT kinds are atTerted in at least two ways; the stationarN- 

 states are displ.ired, .md tlie "missing lines" are csoked, as I have 

 siiid ulread\ . 



li I J. lulensily-Ralios 



The relative intensities of the various lines of a doublet, or triplet, 

 or multiplet are often equal within the (fairly large) uncertainties of 

 measurement to simple ratios, such as 1 : 2, 2: 3, !i : 4. This ha[)pens 

 tiK) often to be easily put down as a mere coincidence, and indicates 

 that the occurrence of transitions is governed by simple laws. Our 

 selection-principles are themselves indications of the same t>pe, since 

 they may be taken as signifying that the intensity-ratio of certain 

 lines to certain others is zero. This problem may be more difficult 

 than the ones I have stressed hitherto, since each line involves two 

 stationary states and is not a quality of one only. This applies to 

 other properties of lines, such as their sharpness or diffuseness. 



E IS. Excitation of Individual Frequencies 



So long as an atom is concei%-ed as a belfry full of bells of various 

 pitches, it would probably be argued that a shock to the atom would 

 set all the bells to jangling, and a gas bombarded by electrons would 

 emit all of its natural frequencies if any. The interpretation of 

 spectra to which these pages are devoted leads to a very different 

 idea. A spectrum-line of frequency t> is emitted when the atom 

 passes from a stationary state 5 to a stationary state A. The energy- 

 \alue of state B by itself does not determine v; this is controlled b%- 

 the difference between the energ>'-values of B and A, which is hv. 

 But the energ>-value of B has everything to do with whether or not 

 the frequency v is emitted under given conditions; for it will not be 

 emitted at all unless the atom is first put into state B. If the gas is 

 bombarded with electrons of energy insufficient to raise an atom from 

 its normal state to state B, then the line in question, and all of the 

 other lines which result from transitions from B to other levels of 

 lower energy-value, will fail to appear. If the energy' of the electrons 

 is raised past the critical value (the difference between the energy- 

 value of B and the energy-value of the normal state) all of these lines 

 suddenly appear. 



This is illustrated by Fig. 12, relating to magnesium. An electron 

 striking a magensium atom and having an energy equal to 3.2 eriuiva- 



