352 ANNUAL KEPORT SMITHSONIAN INSTITUTION, 1910. 



Ha line.^ But as the center of the line does not suffer anomalous 

 dispersion, this theory can not apply to the images of the upper 

 strata with which we are now interested. 



The black filaments which are found similarly in both calcium and 

 hydrogen are in fact a very characteristic element of the upper 

 stratum. Hale had already a glimpse of them in the earlier, really 

 composite spectro-heliograms taken with K and Ha light, and noted 

 them under the name of long dark flocculi and suggested that 

 they very probably belonged to the high strata. Indeed, under such 

 conditions one may often obtain the most important filaments which 

 appear as very broad, dark lines. But for a complete knowledge of 

 the filaments and their properties we must have recourse to the 

 images solely of the upper stratum. 



Another important element of the upper strata is the bright re- 

 gions of faculae wdiich are found at the same Dositions as on the sur- 

 face though of different form. 



To sum up, if we examine the four layers formed by the surface 

 and the atmosphere of the sun, the brightest portions are above the 

 facula?. But the darkest regions are placed very differently at the 

 surface and in the upper strata. Below they occur in the spots; 

 above, in the filaments which occupy a total black surface greater 

 than that of the spots. The area covered by these filaments should, 

 be measured as exactly as that occupied by the spots. 



7. RESEARCHES ON THE MEVEMENTS OF THE ATMOSPHERE AN INSTRU- 

 MENT FOR REGISTERING SPECTRUM VELOCITIES. 



The black filaments especially attract attention, and indeed justly, 

 for, as we have just said, they have an importance at least equal to 

 that of the spots. What, then, is their origin and what the nature 

 of these long dark lines? An accurate answer is very difficult; it is 

 significant to recall our uncertainty as to the spots which have been 

 studied 300 years. However, with the filaments the inquiry may be 

 more easy. The surface which carries the spot lies between the inte- 

 rior of the sun, which escapes our vision, and the lower complex 

 strata of the atmosphere ; on the other hand, the upper layer, with 

 which the filaments are connected, is more free, more disengaged, and 

 may have a structure and movements more simple. 



Indeed, at Meudon, several results regarding the filaments have 

 been obtained worthy of note by virtue of a special device developed 

 and used as yet only at Meudon and which is an instrument for 

 registering or indicating spectrum velocities (spectro-enregistreurs 

 des vitesses). This apparatus, used since 1892, was greatly improved 

 in 1907. It reveals, as its name indicates, the radial movements of 



1 Some months after the Royal Institution lecture we found at Meudon that the anomalous 

 dispersion was actins on the pdjres of the K^ line ; so that the displacement toward the red 

 is increased and the displac(>ment toward the violet diminished. But the image of the 

 center of the line with the spcctroheliograph is always unchanged. 



