k 



107 



near the Atlantic, and the latter at Desierto de las Palmas, on 

 the Mediterranean. Allowing for difference of position in viewing 

 the sun, the photographs taken on that occasion have a most 

 satisfactory accordance, and show the prominences gradually 

 disappearing on one hmb and coming into view on the other, as 

 the moon advances across the solar disc. Hence Secchi con- 

 cluded that the prominences were real solar phenomena. For, 

 in the first place, the perfect agreement between the pictures 

 obtained by two observers separated the one from the other by 

 a hundred leagues, precluded all possibihty of the appearances 

 being merely optical delusions of the mirage type. Then, too, 

 they were not mountains on the moon, for they were gradually 

 covered and uncovered by that body. The only possible theory 

 was that they were true solar appendages. 



If the methods adopted in the eclipse of 1860 were a step in 

 advance, a much greater step was taken in the Indian echpse of 

 1868, when the camera was supplemented by the most won- 

 derful instrument of modern times, the spectroscope. Before 

 proceeding to give any account of the eclipse of that year, it 

 will be necessary, omitting details which would be out of place 

 in the present instance, briefly to recall to mind a few of the 

 great principles which form the basis of the science of spectrum 

 analysis. If we pierce a hole in the shutter of a darkened room, 

 as Newton did in his experiments, and admitting the light of the 

 sun, receive it on a screen after passing through a prism, we 

 shall have a coloured image of such a hole painted in every one 

 of the colours of the spectrum. But the size of the hole will 

 cause the images to overlap, and render the spectrum impure. 

 If, however, we use as an aperture for admitting the light a thin 

 hne, we shall get the image of the line painted in all the spectral 

 colours ou our screen, but in this second case the images wUl not 

 overlap so much, and the resultant spectrum will be purer. The 

 finer the line, the less will its images interfere one with the other. 

 Sohd bodies, liquids, and gases under great pressure, when at a 

 white heat, give such a continuous spectrum as has been de- 

 scribed, in which no image of the Hne is wanting. If however 

 we take as our source of light the vapour of some metal, as for 

 instance that of sodium, one of the constituents of table-salt, we 

 no longer get the sht-images painted in every colour, but only in 

 a few. In the case of the metal selected as an example, under 

 ordinary conditions of temperature and pressure we shall get but 

 one image of the sUt, forming upon the screen a bright yellow 

 hght. The second kind of spectrum then which is produced is 

 one of bright lines, and is due to the incandescent vapomrs of 

 metals. When we view the analyzed light of the sun, in a 

 spectroscope, we see the continuous band of colours, whence we 

 draw the deduction that the core of our luminary is composed of 



