188 



Spectrum of Chromosphere, 



strong as the majority of those which are the representatives of the 

 enhanced lines. 



In the case of iron, all the well-enhanced lines are represented in the 

 eclipse spectra, but they are not of quite the same prominence as the 

 titanium enhanced lines. They are, so far as their intrinsic intensities 

 in the iron arc spectrum are concerned, quite insignificant lines as 

 compared with the majority of other iron lines, but their importance 

 lies in the fact that they are a class of lines of special behaviour, being 

 relatively stronger in the spark spectrum than in the arc. In the 

 eclipse spectra they are undoubtedly represented by stronger lines 

 than are the great majority of unenhanced iron lines, however strong 

 the latter may be in the iron arc spectrum itself. 



Owing to the great number of iron lines in the solar spectrum, a 

 comparison similar to that given for titanium over the whole region 

 covered by the eclipse lines would necessitate the compilation of a 

 very lengthy list. But whatever evidence there is either one way or 

 another should be revealed by a comparison over a limited region, so 

 it is proposed to take that between X 4500 and X 4600, since the 

 proportion of enhanced to unenhanced iron lines is there greatest, 

 and therefore a better opportunity is afforded of a fair comparison of the 

 behaviour of the two classes of lines. The table given on p. 187 is 

 arranged in exactly the same way as in the case of titanium, with 

 the exception that there is an additional column showing the inten- 

 sities in the arc spectrum, as recorded by Kayser and Runge. 



It will be seen that the unenhanced lines are here also unrepresented 

 in the eclipse spectra, with the possible exception of three, which are 

 recorded as very weak lines in one of Professor Frost's spectra, but are 

 missing from the other. All the enhanced lines, however, although 

 they have the weakest arc intensities, appear in each of the eclipse 

 spectra, and have abnormal intensities compared with those corre- 

 sponding to the unenhanced lines. It must be pointed out that only 

 four of the nine enhanced iron lines in the part of the spectrum con- 

 sidered appear in the above list, because they are the only ones which 

 are given in Rowland's origins for solar lines. At least four out of 

 the remaining five— those at XX 4515-51, 4522-69, 4556-10, 4576-51, 

 probably correspond to the solar lines 4515-51, 4522*69 (or possibly 

 4522*80), 4556*06, and 4576*51, to which Rowland has assigned no 

 origin. The outstanding line at A. 4541*40 is doubtfully present 

 in the solar spectrum. The first three of these five have correspond- 

 ing lines in the eclipse record ; the other two have not. In the 

 Kensington reductions of eclipse spectra there are, however, lines 

 agreeing (within 0*3 tenth-metre) with every one of the enhanced 

 lines mentioned. 



