POLARIZED SPECTRA OF DOUBLE CHLORIDES. 



109 



thetical position was taken as the starting-point of the corresponding 

 absorption series, the hypothetical positions of the subsequent mem- 

 bers being found by addition of the weighted average for the observed 

 interval of that series. 



By reference to figure 73 and to tables 35 and 37, the reader can note 

 the general agreement between observed and calculated positions ; also 

 the occasional discrepancies. In the spectra of uranyl ammonium 

 chloride and uranyl rubidium chloride for example, the B w and A w 

 series are spaced at such an interval that aw and bw can not coincide, 

 as will be seen from figure 73. The observed bwaw bands occupy posi- 

 tions between the assumed positions of the b and a series, which tends 

 to show that the ba band is a narrow doublet. The fact that a few of 

 the observed absorption bands do not appear to be in their proper 

 places can be readily explained when it is remembered that there is 

 sufficient experimental evidence to lead to the belief that many of the 

 absorption bands are doublets, consisting of a strong and a weak com- 

 ponent. The breaks in a few of the absorption series, as in the e g and 

 a g series of the uranyl ammonium chloride, are undoubtedly due to the 

 sudden increase in strength of one component, accompanied by a 

 corresponding decrease in the other component. 



Table 35 contains the observed positions of all the fluorescence and 

 absorption bands at +20, measured in our determinations of the 

 spectra of the four salts. Figure 73 is a map of only the central portion, 

 extending, as already stated, two groups into the fluorescence on the 

 one side and two groups into the absorption on the other. 



TABLE 35. Polarized series at +20 C. 



URANYL POTASSIUM CHLORIDE. 



