( 79 ) 
$ 4. Comparison of the different groups of hands of the same salt. 
I The photographs of PI. I and II show better than any description 
the spectra of a number of uranyl compounds. The data for the 
separate spectra are given with the explanation of the figures. 
| The spectrum of a definite salt consists of different groups (we 
(observed from 7 to 8), which though not quite the same, still show 
; a very small difference between successive groups. Only the most 
^refrangible group, some bands of which, as we shall see later, belong 
both to the emission and absorption spectrum, is somewhat different 
from the others. We shall call those bands homologous which occupy 
the same relative position in the different groups. Such, for instance, 
are the bands whose wave-lengths are marked on fig. 7. (Double 
acetate of uranyl and sodium). 
The difference in the appearance of the different groups is the 
result of. successive ascending changes in the relative intensities of the 
bands in these groups. This can be seen in all the figures, especially 
in fig. 7. The band 566^.34 (fig. 7) is the most intense in the least 
refracted group visible in the figure; in the groups succeeding this 
one towards the side of the smaller wave-lengths, the homologous 
bands 540. u “45. 516. ^78, 495.' /u 02 become steadily weaker, and 
in the blue group the band which should be homologous with the 
foregoing seems to have disappeared. 
On the other hand, the bands which occur on either side of the 
above-mentioned homologous series, preponderate in the more refracted 
portion of the spectrum, but become less distinct as the greater wave¬ 
lengths are approached. 
The spectra of uranyl sulphate has the same general appearance 
quite independently of whether it is combined with other sulphates 
or not. (See figs. 1, 8, 9,10). It was noticed long ago’) that the 
structure of the spectrum depends chiefly upon the acid with which 
the salt is formed, and is little influenced by the other base in the 
double salt. But this is shown much more clearly by observations 
at temperatures at which the bands no longer overlap. 
§ 5. Law of succession of the bands. One of the most important 
problems is to ascertain the law connecting the frequencies of suc¬ 
cessive bands. Observations at ordinary temperature had already 
shown that the broad bands (such as those of fig. la) were at such 
distances apart, that the difference between the frequencies of successive 
*) Edmonu Becquerel, Mem. de 1’ Acad. d. Sc. t XL p. 8 and 15. 
