22 Capt. E. H. Hills. [May 10, 



board the " Alecto." The resulting photograph, in the case of the 

 two-prism spectroscope shows a prominence spectrum on both sides 

 of the dark body of the moon, and outside these a corona spectrum 

 with a faint solar (dark line) spectrum on its extreme edge. The 

 H, K and some other lines extend over the dark moon and on both 

 sides beyond the limits of the corona spectrum. That of the one- 

 prism spectroscope shows the same general character, but there is a 

 prominence spectrum on one side only. Both these photographs were 

 over-exposed, better results would have been obtained if two or even 

 three exposures had been made in the same time. 



Measurement of the Photographs. 



The following is the method of measurement adopted. A very 

 accurate micrometer by Hilger, reading to O001 mm., was employed 

 throughout. 



The large number of bright lines in the prominence spectrum 

 rendered the use of the reference spectra unnecessary. 



The hydrogen series, together with the lines at wave-lengths 4215'3, 

 4471'2, and the 6 group gave a sufficient number of fixed points through 

 which to draw an interpolation curve. The micrometer readings of 

 these lines having been taken with the greatest possible accuracy, an 

 interpolation curve was constructed on a large scale, two curves being- 

 drawn for each photograph as a check on each other. The micro- 

 meter readings of the remaining prominence lines were then deter- 

 mined and their wave-lengths taken from the curves. 



The micrometer readings of the corona lines were next taken. It 

 was impossible to get both sides of the photograph in the field of the 

 microscope at the same time, so each side was taken separately, thus 

 getting four series of scale-readings representing possible coronal 

 lines. 



The wave-lengths corresponding to these scale-readings were then 

 determined from the interpolation curves, and lists were made first, 

 of lines common to both photographs ; second, of lines occurring on 

 both sides of one photograph ; third, of lines which had been ob- 

 served in previous eclipses. 



New measurements of the photographs were again made, with the 

 same care as the first, and all lines in the lists were struck out which 

 were not plainly visible in this second scrutiny. 



It is possible that this final list may contain some wave-lengths of 

 lines due to accidental marks ; this must be rare, however, as any 

 mark so treated must have been parallel to the lines. 



A comparison of the measurements of the two photographs will 

 give a good idea of the limits of accuracy of these results. 



