140 RUMFOiiD SPECTROHELTOGRAPfi. 



sphere lies beneath the h)west level that can be observed at the limb. 

 On the basis of Kirchhoff's laAv, the comparative darkness of these 

 bands in the solar spectrum would be ascribed to the fact that the 

 calcium vapor which they represent is cooler than the photosphere 

 below it. With increasing elevation, in a region of lower pressure, 

 the density of the vapor decreases, and to this decrease of density 

 there corresponds a decrease in the width of the bands. In the lowest 

 portion of the chromosphere that can be observed at the sun^s limb 

 the density of the vapor is so far reduced that the broad and diffuse 

 bands are replaced by fairly well-defined lines (Ho, Ko), which main- 

 tain their width up to a certain elevation in the chromosphere and 

 then grow narrower, thinning out to much narrower lines (H,,, K3) 

 in the upper chromosphere and prominences (fig. 2a, pi. 11). On 

 the disk H,. and K3 appear as fairly narrow dark lines at the center 

 of the broad H^ and K^ bands. They occur in practically all parts 

 of the disk, but differ greatly in intensity in different regions. Every 

 bright calcium flocculus on the disk is characterized by the presence 

 of bright H. and Ko lines at the center of H^ and K^, with narrow 

 dark H3 and K3 lines, due to the absorption of the cooler and rarer 

 vapor in the upper chromosphere supei-posed upon them. Intensity 

 curves showing these peculiarities of the H and K lines are given by 

 Jewell in the Astrophysical Journal, iii (189G), page 100, where the 

 displacements of the lines are also discussed. They are also illus- 

 trated in the photographs reproduced in plates 11 and iii. 



From a strict application of Kirchhoff's law it would appear that 

 the calcium vapor in the lower chromosphere is actually hotter than 

 the calcium vapor which lies above and below it. It seems improb- 

 able that the law can be rigorously applied in this case, and hence it 

 may be necessary to attribute the strong radiation of the intermediate 

 layer to causes other than temperature alone. 



In view of the composite character of the calcium lines, it should 

 be ])ossible with the spectroheliograph to photograph sections of 

 the calcium flocculi at levels corresponding to their several elements." 

 If, for example, tlie second slit were set at the extreme edge of K^, 

 the resulting photograph should show only that calcium vapor 

 which is dense enough to proihice a line of this })readth, i. e., a sec- 

 tion across the base of the calcium flocculus, should be obtained. 

 Under no circumstances could the upper and rarer portions of the 



n Experiments of this liind were not nndertalien witli tlie Kenwood spectro- 

 lieliograpli, since the Instrnnient was not well adapted for worlc with darlc lines. 

 Some of tlie photographs, however, apparently show low-level (K, ) phenomena, 

 and Mr. Evershed uiforms me that his plates do likewise. 'SI. Deslandres 

 made photographs with K, and K, in 1804, hnt T have seen no statements of 

 conclusions derived from a study of the K, plates, and do not know whether 

 the method has since heen employed at Meudou. 



