ATMOSPHERE OF THE SUN- — DESLANDBES. 355 



movements in the stratum next to the spots noted by the English 

 astronomer Evershed. We may easily explain why the spots are 

 pointlike and the filaments linear, sometimes very long. Our prob- 

 lem therefore through these researches has already received some 

 light; it will, it seems probable, be completely elucidated when we 

 have continuous measures of the radial velocities over the whole 

 disk of the sun, unfortunately necessary for a very long period of 

 time. 



8. THE DISCOVERY OF THE POLAR FILAMENTS. 



I will close with a new phenomenon of the filaments recently rec- 

 ognized at Meudon and already published. The observatory has so 

 far obtained pictures of the upper stratum for more than 20 en- 

 tire revolutions of the sun, and from them it is possible to study the 

 distribution of the filaments. They appear in all latitudes but at the 

 poles they are generally grouped on a curve, more or less circular, 

 surrounding the pole, although often not along a parallel of lati- 

 tude. This polar curve of filaments is at times clearly seen at both 

 of the poles, but in general it is distinctly visible only at one and 

 tends to move from one pole to the other. It was particularly dis- 

 tinct and strong during last April at the South Pole. (See the two 

 pictures of Apr. 11 and fig. 1, which show the filaments of four dif- 

 ferent days.) 



These polar filaments are accompanied by prominences and accord 

 with the secondary maxima of prominences at the poles which have 

 already been noted. They may thus be related to the special form 

 of the corona which appears during the minimum of the sun-spot 

 cycle and with the often-noted inclination of the coronal axis to 

 the ordinary solar axis of rotation. 



At times the polar curve is accompanied on the side toward the 

 equator by a line of parallel filaments which are reunited to the 

 curve by filaments or lines more or less inclined; and so we find a 

 disposition analogous to that of the bands on the planet Jupiter. 



Finally, the polar zone of filaments, where, as we have just seen, 

 the vapor is ascending, may be compared to the zone of spots and 

 faculse near the equator, where, contrarywise, the vapor is descending. 

 We are led to suppose a great meridian circulation in the upper 

 stratum, a vast general convection current analogous to that which 

 exists in each hemisphere of the earth between the latitudes 35° and 

 the poles. 



Time fails, unfortunately, for developing all the consequences of 

 these first observations, but the facts given suffice to show the great 

 interest connected with the study of the upper strata of the solar 

 atmosphere and the necessity of continuing it. 



