C84 ) 
We have investigated if a magnetic field has any influence upon 
the spectra of the uranyl salts. In a field that could exceed 20 Kilo- 
gauss in value , no influence even upon the narrow bands at 14° K. 
(solid hydrogen ) was to be observed. This negative result forms a new 
analogy between these spectra and the band spectra of gases, which 
likewise, in the majority of cases undergo no change in a magnetic 
field, although the strength of this is raised to the highest value at 
present possible *). 
§ 7. Relation between the emission by phosphorescence and the 
absorption of the uranyl salts. Reversible and irreversible bands. 
In the foregoing we have recalled the existence of groups which 
contain at the same time emission and absorption bands. Crystals of 
autunite (double phosphate of uranyl and calcium), a mineral from 
which very transparent plates may be obtained by cleavage, are 
especially suitable for the study of absorption. 
Fig. 12 shows the emission and absorption spectra in the blue and 
green region of the spectrum of a plate of autunite at a temperature 
of 14° K. (solid hydrogen). It can be seen that two bands lying very 
close beside each other are common to both spectra; a is well 
developed both as an emission and absorption band, is strong as 
an emission and weak as an absorption band. 
Fig. 13 gives the spectrum of autunite at 80° K. 
In can be shown experimentally in a striking manner that some 
bands can arbitrarily become emission or absorption bands. For this 
purpose an inteuse beam of light is transmitted through the crystal 
plate and the position of the absorption band fixed; the intensity of 
the transmitted light is then diminished and at the same time the 
plate is illuminated by violet light whose intensity always increases. 
The dark band is seen to be transformed into a light band occupying 
exactly the same position. This experiment, which becomes possible 
at the lowest temperatures, resembles the classical experiment of the 
reversal of the lines in the case of the sodium-flame. 
The bands in the less refrangible groups cannot be reversed; they 
all belong to the emission spectrum. 
The marked difference between the groups which are common to 
both spectra, and those which belong only to the emission spectrum, 
can be attributed to the reversible bands whose relative intensity, 
x ) This, however, is not generally the case. As is known, A. Dufoub has found 
numerous bands in the spectra of the chlorides and fluorides of the alkaline ear^ 
which show a Zeeman effect, and one of us has also observed this effect in 
yttrium-bands. 
