( 82 ) 
bands was almost constant 1 ). But as the bands at this temperature 
are broad, diffuse, and complex, it was not possible to determine the 
accuracy of the relation. In the spectra obtained at low temperatures 
the wave-lengths could be measured; this was done by comparison 
with the iron-spectrum which was photographed upon the same plate 
as the phosporescence spectrum (see, for instance, figs. 6, 7, 8, 10). 
A high degree of accuracy is possible in these measurements, 
particularly at the temperature of liquid hydrogen : in the case of 
the double sulphate of uranyl and potassium (fig. 6) and of the 
double acetate of sodium and uranyl (fig. 7) the error in the position 
of the principal bands is not more than ± 0^,03. The table given 
on pp. 80 and 81 contains the results of the measurements for 
the bands of certain salts, homologous bands being placed upon the 
same line. 
The values of wave-lengths given above were measured in air, 
but the differences between the reciprocal wave-lengths are reduced 
to vacuum, so that the values given for these differences are pro¬ 
portional to the frequencies. 
From the table it appears that the difference between the frequen¬ 
cies of two successive homologous bands is practically constant, not 
only for the same, but also for all series of homologous bands of 
the same salt. Further , the values of this constant for the various 
salts difjer but slightly from each other. 
As regards the position of the bands, the spectrum of a uranyl 
salt is thus represented on the scale of frequencies by successive 
identical groups at an equal distance from each other. The deviations 
from this law are, as a general rule, so small, that they can be 
ascribed to errors in the measurements of the wave-lengths; never¬ 
theless there is a tendency towards decrease in the frequency-diffe¬ 
rences with increasing wave-length, and it is possible, and even 
probable, that there is a deviation from the law in this direction, 
although in that case, the deviation is extremely small. 
In the foregoing we have recalled the influence of the acid upon 
the structure of the spectrum; now we must call attention to the 
fact that although these spectra are also spectra of molecules, which 
differ in the various salts, yet their character is due to the uranium> 
which subjects the spectra of all the salts to the law just mentioned. 
The law according to which the homologous bands correspon 
with equal differences between the frequencies is the simplest t a 
could connect successive lines in the spectrum. 
*) Edmond Becquerel loc. cit. § 4 etc. Henri Becquerel Compt. R* d - u 
V-.1 1885 p. 1259. 
