156 THE INDIAN ECLIPSE, 1898. 



third contact, these two points being separated by an angle of 

 39 degrees only. The lowest strata of the layer would of course 

 only be revealed at the moments of the contacts, while at 

 mid-eclipse only the highest limits would be seen. But the 

 comparatively slow rate at which the excessively thin flash 

 spectrum layer would be occulted by the moon and then again 

 uncovered, would evidently afford an excellent opportunity for 

 obtaining a long series of photographs of the spectrum at the 

 various stages; those taken at mid-eclipse giving high-level 

 spectra only, wiiilst photographs taken fifteen seconds earlier 

 and later would give high- and low-level spectra combined. 



But even under these favourable conditions it will still be 

 difficult to separate the high- and low-level spectra unless 

 instruments of great focal length and considerable aperture are 

 employed. The radiation coming from a stratum at the base 

 of the layer, say within 100 miles of the photosphere, may be 

 very intense, yet it will be difficult to photograph on account 

 of the extreme fineness of the spectrum lines ; and small 

 apertures would probably entirely fail to get anything but the 

 spectra of the more extensively diffused gases. 



At all stations in the Spanish peninsula where the duration 

 is one-third that on the central line, it will happen also that 

 one of the contacts occurs at, or very near, one of the poles 

 of the sun. If north of the central line, third contact will be 

 at the north pole : and if on the southern border of the shadow 

 track, second contact is at the south pole. In either case, 

 therefore, the spectrum of the polar regions would be obtained, 

 and it would be of interest to learn whether the flash spectrum 

 at the poles differs in any way from that at low latitudes 

 hitherto observed ; or whether it has the same composition in 

 all parts of the sphere. It w T ould be important from this point 

 of view to carefully compare the spectra obtained at second and 

 third contacts, since these points being separated by 39 degrees 

 would give this range in latitude. 



Another important point which future eclipse w r ork will deter- 

 mine is the composition of the flash spectrum with reference to 

 the sun-spot cycle. A comparison which I have made between 

 the flash spectrum photographed by Mr. Shackleton in August 

 1896, and that obtained by me in January 1898, shows that 

 in the visible part of the spectrum between D and H the two 

 spectra are identical. The former does not extend far enough in 

 the ultra-violet to make the comparison complete, but below H 

 I can find , no line on the one which is not also indicated on 

 the other, and vice versa ; and the relative intensities seem 

 to be the same. But it may well be that photographs taken 

 at opposite phases of the spot cycle would show differences. 

 Judging by the changes which seem to take place in the 

 coronal spectrum, we should expect the flash to be richest in 



