ATMOSPHERE OF THE SUN DESLANDRES. 351 



line amounting to between forty-seven one-hundredths and sixty- 

 two one-hundredths of an Angstrom, we get the result of 1893 ; that 

 is to say, the regions of faculse appear black relative to the back- 

 ground. 



When the middle of each side is used at a distance of from ten 

 one-hundredths to forty-two one-hundredths of an Angstrom from 

 the center, the result is entirely different. It shows the principal 

 characteristics of the spectro-heliograms taken in America in 1908 

 and particularly the groups of small lines which Hale has called 

 " solar vortices." 



Finally, with the center of the line we get a third, yet different, 

 aspect from the other two, much paler and simpler and correspond- 

 ing to the upper layer of hydrogen. 



Now, and this point is important, the new image shows the dark 

 filaments of the K 3 layer of calcium. As to the regions of faculse, 

 they are bright, never dark; they cover a smaller region than with 

 the K 3 stratum and correspond to the maxima of brightness of the 

 similar regions in the K 3 stratum, maxima which differ from those 

 of the K 2 and K x strata. The darkest and the brightest parts are 

 the same. (See the annexed pictures taken with K 3 and with Ha 

 the 11th of September, 1909, and the 21st of March and the 11th of 

 April, 1910.) 



And yet further, we have isolated the various parts of the blue H/3 

 line of hydrogen, showing a lower elevation in the solar atmosphere 

 than the Ha line, and so obtained images which show almost exclu- 

 sively the dark regions of facula? such as we found in the shaded 

 portion of the red Ha line and which therefore correspond to a low 

 level. 



Finally, we are led to conclude that hydrogen gives, like calcium, 

 at least three distinct superposed strata which are now for the first 

 time clearly distinguished. 



Now, in what just precedes I have treated the different portions 

 of the same line and the different corresponding images by the 

 ordinary laws of emission and absorption by gases, admitting nat- 

 urally that the density of the gas and the width of the corresponding 

 line diminish as we go upward in the solar atmosphere. Now, the 

 theory of anomalous dispersion has been brought forward as coming 

 into play here, and, at least in part, explaining the peculiarities of 

 these images. But it seems to me that anomalous dispersion, while, 

 of course, to some extent it must come into play, does so only to a 

 minor extent and may be neglected in this preliminary study. The 

 real reasons for making such an assertion would take too long to 

 develop here. However, anomalous dispersion has been found in the 

 laboratory with the lines H and K of calcium, and recently with the 



