( 172 ) 
intensity both increase the number of electrons ejected from the 
metallic atom (after intermediate returns), and thus also increase 
the probability that amongst these there will be found some whose 
direction is the proper one. In the same way, it can be explained 
that the momentary process, the glow during the illumination without 
afterglow, in the lower momentary phase is not entirely wanting in 
the permanent part, for it is just the return of the electrons which 
have not reached the place where they are stored that causes this glow. 
§ 8. Non-resolution of bands in amorphous substances. The 
supposition that in consequence of heat motions the atoms of the 
centra temporarily assume different configurations also affords an 
explanation of the broadening and narrowing of the emission bands 
at high and low temperatures respectively, analogous to that given 
by P. Lenard in explanation of fche broadening of the metal-lines in 
flame-spectra*), and to that given by H. and J. Becquerel and 
H. Kamerlingh Onnes at the end of their paper for the modification 
of the positions of the bands of uranyl under the influence of 
the proximity of other atoms which form with it a molecule. For 
the different configurations will cause varying changes in the periods 
of the inherent vibrations of the centra. Without heat motion, 
at 0° K., each centrum would have a definite time of vibration, 
invariable with the time. It still depends, however, upon a particular 
circumstance whether these inherent and time independent vibration 
periods of all centra (of the same kind) in the whole preparation 
should in each case be constant, which would be necessary in order 
to obtain a spectrum consisting of fine lines. For this all centre 
would have to take up the same position with respect to the neig - 
bouring molecules. For these molecules in a solid substance act wi 
different forces upon the centra according to their distances an 
orientation, and these must cause different modifications in the pen • 
These equal distances and resulting equal periods for all P arts ® 
the preparation independent of time and position are to be expec 
only in crystalline substances such as those investigated by the above- 
mentioned experimenters, and not in the amorphous ‘ phosp ors 
investigated by us; this affords an explanation of the fact that ev<m 
at 14° K. we did not yet see the spectra resolved into lines. ^ 
we saw was — according to the image just developed the ov ^ 
apping of bands consisting of the sharp lines which each cen ru ^ 
gives at very low temperatures, and which by mutual interac 1 
!) Lenard. Ann. d. Phys. 17 p. 232, 1905. 
