SUMMARY OF RESULTS AND SUGGESTIONS. 157 



lines at a time of maximum sun spots. It is well known, too, 

 that when spots are numerous the chromosphere gives many 

 more bright lines than are seen at times of minimum activity, 

 the eruptive action which so frequently takes place in the 

 neighbourhood of spots elevating for the time being the low- 

 lying metallic vapours. 



A chance photograph of one of these metallic eruptions taken 

 during the moments when the flash spectrum is visible would 

 be of the greatest interest, as it would probably give valuable 

 information as to the constitution of the very lowest strata. 

 Such a favourable chance is perhaps almost too much to hope 

 for, at any rate until an eclipse occurs at a time of very great 

 solar activity. The observation could only be made at a station 

 near the central line of the eclipse, as eruptive prominences are 

 never seen in high solar latitudes. 



The experience gained at the recent eclipse with regard to 

 photographic plates is in one respect of great value to the 

 spectroscopist. The advantages attending the use of triple- 

 coated plates was clearly demonstrated by the successful 

 photographs obtained by Mrs. Maunder ; and in particular by 

 the coronal photograph taken forty seconds after the sun had 

 reappeared. 



Now, in the series of spectrum plates exposed by me, the first 

 was an ordinary single-film isochromatic, whilst the last was a 

 Sandell triple-coated plate : both received two images of the 

 cusp spectrum with exposures of about half a second, the single- 

 film plate just before second contact and the triple just after 

 third contact. Comparing the two results, it is evident that 

 there is a marked difference in favour of the triple-coated plate. 

 The final exposure on the latter was some twenty seconds after 

 the sun had reappeared, the full blaze of the brilliant cusp 

 falling upon the plate; yet no halation effects obscure the 

 delicate fringe of bright lines bordering the spectrum, and I 

 feel confident that longer exposures might have been made 

 without any ill effects, whilst the finer details would have been 

 more strongly impressed. Probably exposures up to two seconds 

 or more on these plates would give excellent results under the 

 same conditions. 



In prismatic camera work these out-of-totality photographs 

 are especially valuable, if only on account of the beautifully 

 defined Fraiinhofer spectrum which is impressed. This, in 

 effect, forms a most convenient wave-length scale, which greatly 

 facilitates the reduction of the bright-line spectra, and in all 

 future eclipse work with the prismatic camera it is most 

 desirable that these cusp spectra be obtained both before and 

 after totality. 



J. EVERSHED. 



