258 



SCIENCE. 



[N. S. Vol. XV. No. 372. 



10 the strongest line. Wave-lengths were 

 compared with Rowland's measures of the 

 .solar spectrum. Of the 363 flash lines, 269 

 were identified with lines on Rowland's map. 

 Although we cannot directly compare the 

 intensities of the bright lines of the flash 

 (scale 0-10), with those of the dark lines 

 given in Rowland's tables (scale 1-1,000), 

 we can arrive at certain theoretical con- 

 siderations if we compare the average in- 

 tensities for the different elements, i. e., 

 Flash intensities , , 

 Solar intTn"site ' ^^'^ ^^'^ *^^ ^^^^"^ of the 

 number of lines of each element identified 

 to the whole number of solar lines for that 

 metal. Forming these ratios and arranging 

 them, we are at once struck with the system- 

 atic variations, not only in the ratio of 

 intensities, but also in the per cent, of lines 

 identified. The meaning of these sys- 

 tematic differences will be understood if 

 we consider these ratios in combination 

 with the atomic weights of the various ele- 

 ments, as in the following table, where also 



is put down the number of the lines in the 

 flash due to each metal. 



These lines naturally fall into three 

 groups, as given in the table below. 



To these may also be added the following 

 lines : 



La, atomic weight, 138.5 3 lines 



Ba, " " 137 1 line 



Ln. " " 65 1 line 



In Group I. would also fall Al if we con- 

 sider the relative intensities of the two lines 

 ^ 3,944.160 and x 3,961.674; and un- 

 doubtedly Na if our plate took in the D 

 lines. The remarkable variation of the 

 relative intensities in the flash and Fraun- 

 hofer spectra, as Evershed has pointed out, 

 is undoubtedly due to the heights to which 

 the vapors of the different metals ascend in 

 the chromosphere. A gas with an intrinsic 

 brightness 1 and a layer 100 miles in thick- 

 ness, would give a photographic line in the 

 flash spectrum just as bright as a gas of 

 intrinsic brightness 100 and only 1 mile 



Group 11. — Lines Strong in Flash, Weak in Solar Spectrum. 



