4 Mr. H. Eamage Relations of Spectra, Densities, [Nov. 7, 



(<?.) The effect on the lines of a change of atmosphere surrounding 



the source of light.* 

 (/.) Shift of lines under changes of pressure.! 

 (g.) Harmonic series of lines. 



The formulae of Eydberg and of Kayser and Eunge should be used 

 in connection with the last of the above. 



The experimental data are still incomplete, especially in the red, 

 infra-red and ultra-violet regions. We now believe that the greatest 

 accuracy of measurement and the fullest details of the ultra-violet lines 

 are most necessary for the completion of this investigation. As far 

 as can be seen from the data now available, all the above methods lead, 

 and will lead, to the selection of the same lines as being produced by 

 corresponding motions of or in the molecules. Much experimental 

 work must yet be done, and it is hoped that the present investigation 

 will indicate the direction and character of the work most needed. 



Flame Spectra. 



JBunsen Flame Spectra. — Eder and Valenta have studied these over 

 the greatest range of spectrum. They, by heating the chlorides of 

 lithium, sodium, potassium, calcium, strontium, and barium in the 

 Eunsen flame, obtained spectra composed of lines and bands, which 

 they attributed to the metal, oxide, or chloride in each spectrum, j 



Oxyhydrogen Flame Spectra. — Hartley has studied these in the ultra- 

 violet region and in the more refrangible portion of the visible 

 spectrum. § The author assisted Erofessor Hartley in this work, and 

 from the summer of 1893 until 1899 the investigation was con- 

 ducted as a joint work. The spectra of nearly all the elements were 

 photographed in the course of the investigation : some of them gave 

 only continuous spectra, and these cannot be included in the present 

 work. Of the others, the following nineteen have been selected and 

 studied in detail : — 



(1.) Lithium, sodium, potassium, rubidium, and caesium. 



(2.) Copper, silver, and gold. 



(3.) Magnesium, zinc, cadmium, mercury. 



(4.) Calcium, strontium, barium. 



(5.) Aluminium, gallium, indium, thallium. 



The following metals also yield line flame-spectra, but the spectra 

 have not been grouped as perfectly, nor studied as fully, as those of 



* Crew, ' Phil. Mag.,' November, 1900. 



+ Humphreys and Mohler, ' Astrophjs. Jour.,' toI. 3, p. 114, 1896. 

 X < Sitzber. Kais. Akad. Wien,' vol. 60, 1893. 

 8 « Phil. Trans.,' A, 1894, 161-212. 



