354 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 1910. 



to the polygons formed by the vortex cells in liquids so well studied 

 in France by Bernard.^ 



Since the vapor descends over the flocculi and rises at the inter- 

 stices, each one of these solar jDolygons is thus probably a vortex cell. 

 Other flocculi in the same image show polygons less sharp and less 

 complete and sometimes, although more rarely, of wholly irregular 

 forms. 



May 20, 1909. June 15, 1909. 



PS 



November 27, 1909. 



Fig. 4. — Drawings of the upper layer of the solar atmosphere showing the characteristic 

 black filaments and especially the polar filaments. These images, obtained with the aid 

 of d'Azambuja, have been made from the monochromatic pictures of the sun taken with 

 the central portion of the Ha line of hydrogen and the K line of calcium. They show 

 only the dark filaments without the alignments. The bright regions above the faculae 

 have not been represented. 



Moreover, the filaments and lines are probably the limits of vortex 

 cells yet greater, superposed upon the preceding in the upper stratum, 

 and of which the spots are the centers. This is in accord with the 



1 This arrangement in juxtaposed polygons is at times very distinct over nearly the 

 whole sun. The K3 plate of Sept. 18, 1908, shows in the Southern Hemisphere, near the 

 center, several of these polygons joined by their sides and comers ; but a larger and 

 sharper image is necessary to show them well. 



