M. Fievez has recently sent me a map of the solar spectrum from C to A* inclusive, and as part of this region is one which I have been measuring, I have examined the new publication with great interest. Photography and eye measurements do not exactly coincide in the detail of the grouping of the little
a
group as far as A, and A itself is shown by M. Fievez’s map as wanting some details which appear in the photographs. Thus in the photographs there are some seventeen lines, whilst in M. Fievez’s map there are but thirteen. Between A and
a
there are several lines of marked intensity in the photograph which are not shown in the new map. The wave-lengths of the different lines from above “
a
” to A are not the same as those given by Fievez, when they are taken from comparison photo-graphs of the 1st order of the red and 2nd of the ultra-violet on the same plate, or when checked by photographs of the 2nd order of the red with the 3rd order of the green taken in a similar manner. In my paper, “Phil. Trans.,” Part II, 1880, I gave a method of using mirrors by which this could be effected, but since Professor Rowland introduced his concave gratings this is much more readily carried out. He has kindly furnished me with gratings for the purpose, having about 14,400 lines to the inch, with focal distances of 7 feet 6 inches and 12 feet 6 inches respectively. These have been employed in determining the wave-lengths of this part of the spectrum. Cornu’s map was used as a reference for the ultra-violet wave-lengths, and Ångström’s map for those in the blue and green. The two maps may be taken as equally exact. The determination of A has been made by Maseart, Smyth, and others, besides Ångström and Langley, with discordant results. I think the above may be taken as accurate as are Cornu’s and Ångström maps.
II. The wave-lengths of A,
a
, and of prominent lines in the infra-red of the solar spectrum
