V8 



ture is more complex than in 1900. In particular that the polar streamers 

 instead of being radial, are bent and interlaced, and in every case long 

 coronal streamers are above the prominences. 



The plates exposed for the intra-mercurial planets are heavily fogged, 

 as one vs^ould expect from a sky covered with bright clouds, but not so 

 badly as to obscure faint star-images. I believe that a plate of the sen- 

 sitiveness of the Seed's 27. which we used, can be exposed three minutes 



Fig. IV. The Intra-Merciirial PlanetlC'aineras. 



without serious fog at a time of a total solar eclipse. Our sky Avas so 

 cloudy that it is unreasonable to expect star-images on these plates. 

 We examined two of them hurriedly (the ones on which Regulus should 

 have appeared), but found no star-images. The photograph of the corona 

 on one of the intra-mercurial plates showed longer extension than on any 

 other plate we exposed — due perhaps to the shifting of the clouds during 

 the long exposure. 



The corona impressed me as being brighter than in 1900. The effect 

 on the clouds of the light from the eclipsed sun was peculiarly striking, 

 and from a spectator's point of view was very beautiful. 



