INTIMATE STRUCTURE ON COOLING. 



79 



of the bands. No attempt has been made to denote the width of the 

 bands. Below each spectrum the absorption bands are sorted out into 

 their respective series. The figure is necessarily on a greatly reduced 

 scale. Our working maps of these spectra are about 2 meters in width. 

 From these maps some of the statements already made can be 

 verified at a glance; e. g., the increased number of bands and series 

 at 185; the greater extent of absorption toward the red at +20 

 than at 185, and that there is in general a greater degree of resolu- 

 tion of absorption than of fluorescence. It may also be noted that the 

 known absorption spectrum is of greater extent than the fluorescence 

 spectrum and that the absorption, considered as a unit, suffers a nar- 

 rowing on cooling which is more marked on the side toward the red. 



TABLE 28. Average intervals of absorption series at 185 C. 



REVERSALS AND THE REVERSING REGION. 



The phenomena of the reversing region, where fluorescence and 

 absorption overlap, are complicated. Some points applicable particu- 

 larly to the double chlorides are, however, discussed here. 



The early observers of uranyl spectra were of the opinion that some 

 connection or relation must exist between the system of bands of 

 fluorescence and absorption. Becquerel and Onnes, who first studied 

 these spectra at low temperatures, were able to confirm the impression 

 of Stokes that the two systems overlapped and that there was actual 



