( 220 ) 



as to the order in which the elements will come into evidence, on 

 account of possible condensation, and because the pressure of radiation 

 counteracts gravitation to a degree that depends on the size of the 

 particles, and, therefore, on numerous unknown conditions prevailing 

 in the sun. 



Yet for each element a certain level must exist, in which its per- 

 centage in (he mixture is a maximum. Accordingly, the refracting 

 properties of successive layers will be governed by different elements. 

 A photograph, made with the spectroheliograph in a hydrogen line, 

 shows a structure which, of course, depends on the distribution of 

 all the hydrogen present in successive layers, but is chiefly determined 

 by the density gradients in a rather high level ; whereas a photo- 

 graph, made with an equally strong iron-line, reveals more especially 

 the structure of lower regions. This explains the difference in character 

 between iron- and hydrogen plates. 



It must be possible, on the other hand, to obtain almost idenlical 

 photographs with different lines, provided they belong either to the 

 same element, or to elements that are most in evidence at about 

 the same level in the sun ; but then another condition has also to 

 be fulfilled, viz. that the camera-slit transmits rays of the same 



( A'.oAi)) 



Fig. 3. 



refrangibility in both cases. If for instance Fig. 3 represents the 

 dispersion curve near H y and near He, the width and the position 

 of the camera-slit ought to be so chosen, as to let in only waves 

 corresponding to parts of the curve, enclosed between equal ordinates 

 in the two dispersion bands l ). 



*) Waves, lying more or less symmetrically en either side of an absorption line, 

 and answering the relation n— 1 = 1 — n' between the indices of refraction n and n' 

 of the medium for them, must give nearly the same spectroheliograph results on 

 the greater part of the disk. This follows from a discussion of the various possi- 



