302 



REPOET 1880. 



taken by M. Oornu by means of a quartz grating ; the lines ruled on the 

 quartz numbered 60 to the millimeti'e. The photographic process em- 

 ployed was the ordinary one with wet collodion and a developer of feri-ous 

 sulphate. The collodion was salted with cadmium iodide and bromide, 

 there being 4 parts of the foz-mer to 1 of the latter salt. M. Cornu's 

 observations regarding that portion of the solar spectrum more refrangible 

 than H include the following remarkable facts. 



1. No single or isolated line is met with among the principal groups, 

 such as we are familiar with in the luminous portion of the spectrum, as 

 for instance the single lines C, D, h and F. 



2. The groups of rays are always made up of twin lines, triplets or 

 multiple groups, lying very close together. This occasions a certain con- 

 fusion which enhances the difficulty of accurately measuring their wave- 

 lengths. The confusion of lines is increased when diffraction-gratings are 

 used instead of prisms, since their dispersive power in comparison with 

 that of prisms decreases with the refrangibility of the rays. 



3. The greatest extent of the solar spectrum can be photographed only 

 in the spring-time of the year and at mid-day ; at any other time the most 

 refrangible rays are intercepted by the atmosphere. 



4. Nearly all the solar rays more refrangible than H are due to the 

 spectrum of iron. A photograph of the spectrum of this metal is a fairly 

 accurate representation of the solar rays, and may be used as a spectrum 

 for comparison. 



5. Without exception all these rays belong to matter which enters 

 into the composition of meteorites. 



The wave-lengths of the principal lines in the solar spectrum have 

 been measured by M. Oornu. He adopts the designation O, P, Q, R, 

 r and S, given by M. Mascart to indicate certain rays. 



The following numbers are the wave-lengths of the principal solar rays 

 determined with a diffraction- grating made by Brunner. 



Solar Line. 

 G' 

 h 

 H 

 K 

 L 

 M 

 N 



The numbers are the results of five different measurements which 

 agreed well with each other, the means agreeing very satisfactorily with 

 numbers obtained by M. Mascart. 



Comparing the spectrum of iron with the solar rays, it was found that 

 the incandescence of the metal caused by the action of fifty-five Bunsen's 

 elements was much more intense than sunlight. The lines in the 

 spectrum thus observed were for the most part coincident with the solar 

 rays L, M, N, 0, P, Q, Sg T and U; their identity was easily recognised. 

 The following table shows the wave-length of iron lines compared with 

 those in the solar spectrum. The ray E. is due to calcium, while other 

 important lines belong to nickel, aluminium, magnesium and titanium. 

 The line P cannot be recognised as belonging to any known metallic 

 spectrum. 



Plate XI. is a copy of a beautiful plate of the normal spectrum ex- 

 tending beyond the line H, published by M. Cornu.' 



' Annales de VEcole Kormale, 1880. 



