U UKTVE 8 



xxv.] "REVERSING LAYER" DOUBTED. ^361 



" The observations made by the Government Eclipse Expedition 

 which went to India in 1871, showed that this flashing out of lines 

 was a real phenomenon ; but as the observation was a general one, 

 and as, during the eclipse, the Fraunhofer lines were invisible, 

 there was no absolute demonstration of the identity of the two 

 spectra. 



" The facts now beyond question, that qud the same element, the 

 spectra of spots and flames differ, and that the spectra differ widely 

 among themselves, throw great doubt upon the conclusion to which 

 reference has been made. 



" First they seem to indicate that some of the absorption takes 

 place at a higher level than that occupied by the so-called reversing 

 layer. 



"Secondly, they seem to indicate that many of the brightest 

 lines seen during the flash to which reference has been made may 

 be those seen thickened in spots or intensified in the prominences, 

 although they do not occur except as excessively faint lines among 

 the Fraunhofer lines. 



"In short, in 1870, the fact that the spot and prominence spectra 

 are so widely different from the ordinary solar spectrum, had not 

 received the attention it must receive in the light of the most 

 recent inquiries, and it was taken for granted that because a large 

 number of lines was seen, that, therefore, they occupied the same 

 positions as the large number of lines which compose the ordinary 

 solar spectrum. 



" The recent work seems to show that the complete absorption 

 spectrum of any one element is produced, not at one level, but at 

 various levels, the absorption of all the levels being added together 

 to give us the complete result. 



" If this be so, the lines seen in the flash will not be Fraunhoferic 

 lines with the ordinary intensities." 



It follows then, from the considerations thus set forth, that 

 the lines seen best before or after totality, or brightest and 

 shortest in the flash at the beginning or end of totality, 

 should be those seen in prominences, and not in spots, and 

 relatively brighter in the spark than in the arc; while the 

 longer lines should be those lines affected in spots, and not in 



