CELESTIAL DISSOCIATION 201 



further fact that in sun-spots one set of iron lines is found, 

 and in this chromosphere, quite another. 



Furthermore, at the maximum sun-spot period the 

 widened sun-spot lines are nearly all unknown; at the 

 minimum sun-spot period they consist of iron and other 

 well-known substances, a fact only to be explained on the 

 assumption that the increased energy at the maximum sun- 

 spot period is adequate to break the " iron and other 

 well-known substances " into finer things. The most inter- 

 esting evidence, however, of the decomposition of iron un- 

 der the fervent heat of the sun is adduced from the fact 

 that while some of the " iron " lines in the sun show that 

 the substance giving rise to them is in rapid motion, other 

 different and adjacent " iron " lines show that the sub- 

 stance in causal connection with them is at rest. Now, if 

 the "up-rush" and "down-rush" of incandescent gas in 

 the sun were caused by iron vapour as iron, it should be 

 registered equally by all the iron lines alike. The fact that 

 it is not gives us every reason to suppose that in the sun, 

 " we are not dealing with iron itself, but with primitive 

 forms of matter contained in iron which are capable of 

 withstanding the high temperature of the sun, after the 

 iron observed as such has been broken up." 



Fig. 50, according to Lockyer, shows the variations, in the 

 line spectrum of iron as they existed in the reversing layer 

 of the sun, the electric arc, the high-potential electric 

 spark, in sun-spots observed at Kensington, and in solar 

 prominences observed at Palermo. It will be observed that 

 the normal Fraunhofer's lines of iron in the sun correspond 

 to the spectrum of iron as obtained at the temperature 'of 

 the arc, while the iron spectrum in sun-spots or solar storms 

 corresponds to the more strenuous conditions of the high- 

 potential spark. 



