[ 231 ] 



VIII. On the Ultra-Violet Spectra of the Elements. Part III. Cobalt and Nickel. 



By G. D. LIVEING, M.A., F.R.S., Professor of Chemistry, and J. DEWAB, M.A., F.R.S., 

 Jacksonian Professor, University of Cambridge. 



Received February 27, Read March 15, 1888. 



[PLATES 9-14.] 



IN our first communication to the Society on this subject we gave a reference map of 

 the iron lines in the ultra-violet region, based on measurements of the wave-lengths 

 of nearly a hundred of the principal linea These measurements were made by means 

 of a diffraction grating by RUTHERFURD. Since that time we have obtained some of 

 ROWLAND'S fine gratings, having a much larger ruled surface, though a somewhat 

 smaller number of lines to the inch, than the RUTHERFURD grating. With larger 

 gratings it was advantageous to use larger telescopes, and it was a matter of some 

 interest to determine whether the improved appliances led to the same numerical 

 values of the wave-lengths as before. In each case we determined the constant of the 

 grating that is, the distance between successive lines by ANGSTROM'S scale. We 

 had, for instance, measured the deviations by the RUTHERFURD grating of several 

 lines namely, C, D, b, F, and H for as many orders of spectra as we could on both 

 sides of the collimator, and then calculated what must be the distance between the 

 lines to produce these deviations, on the assumption that ANGSTROM'S wave-lengths of 

 these lines were correct. We have since used the solar lines E, or the corresponding 

 lines of the spectrum of the spark taken between iron terminals, for the purpose of 

 gauging our gratings, because ANGSTROM had bestowed the greatest pains in the 

 measurement of the wave-lengths of these lines. PEIRCE and BELL have corrected 



o 



ANGSTROM'S measures, but, for convenience in comparisons, we have retained the old 

 scale. In order to make our numbers correspond to the new scale they must all be 

 multiplied (according to ROWLAND) by 1 '00016. 



The first comparison of measurements with the different gratings was made with 

 respect to the cadmium lines, which have frequently been used as lines of reference. 

 Four pairs of photographs of the cadmium line No. 17 in MASCABT'S notation had 

 given us the following wave-lengths : 



3.8.88 



