89 



373 



(13 — 12), (22 — 21) in which the angular momentum of the electron round the 

 nucleus decreases by ft/'ir. From table XI the first of these components would be 

 expected to be stronger than the second and the second again stronger than the 

 third. Paschen, however, characterises their intensities by the numbers 7, 7,5, 3 

 respectively i.e. he linds the component (04 -'03) a little weaker than (13 > 12). 

 This seems to indicate again that the a-priori probability of spontaneous transition 

 between two circular orbits in the region of small n's is less than it would be 

 expected from the numbers obtained by the method of estimating intensities by 

 means of the values of the am[)litudes of the harmonic vibrations occurring in the 

 motion of the atom, these values giving, in a singular case as this, an exaggerated 

 picture of the intensities (compare page 60). In Paschen's spark discharge image 

 the three components in question also appear, but they have become more diffuse, 

 and I has become stronger than II, the relative intensities being now characterised 

 by 7, 6, 0,0 respectively. This might obtain anexplanation if we assume that, in the 

 case of a spark discharge, perturbing electric fields have been acting on the atoms 

 of such intensity that the new components (22 12) and (13 ^03) have appeared 

 with considerable intensity. In fact, these com[)onents lie so near to (04^*03) that 

 Ihey may be assumed together with the latter component to contribute to the intensity 

 of the component I observed by Paschen, while on the other hand the new compo- 

 nent (04-^12) which lies very near to (13-* 12) will, as seen from the table, only 

 [jossess a very small intensity in comparison with (22^12) and (13 -'■03), and 

 cannot therefore be expected to contribute essentially to the intensity of component 

 II. In this connection it must, however, bo remarked that the amplitudes of the 

 harmonic vibrations of frequencies which correspond to the original components, 

 in general, owing to the influence of the perturbing field, will have changed by small 

 amounts proportional to ^',0-, and that as a consequence of this we may be prepared 

 to find that the intensities of these components themselves have varied by amounts 

 which are of the same order of magnitude as the intensities of the new components. 



The weak original components (22-^03) and (31^12) which correspond to 

 transitions for which the angular momentum increases by '» 2;7, and the theoretical 

 distance between which would be equal to 0,034 A, are in the continuous discharge 

 spectrogram, as well as in the spark discharge spectrogram, recorded by Paschen 

 as a single line, which in figure 10 is indicated by IV. 



Moreover Paschen has observed se])arately the new components (31-^03), 

 (31-^21) and (13^21), indicated in the figure by V. \'l and VII respectively. In 

 table XI the values of s(R'-) and s(/?"-) corresponding to (31 -»03) are equal to 

 zero but, as pointed out in the analogous case in the Stark effect (see page 5S, 

 compare also page 99), it is not permissible from this to draw the conclusion that the 

 a-priori probability for this transition is zero. Also the component (31 -* 2 1) appears 

 both on Paschen's continuous discharge image and spark discharge image and is 

 stronger than (31 03), in agreement with the table. Finally, as regards (13 ^21), no 

 corresponding component is observed in the continuous discharge image, bul on the 



