WOOD. — SPECTRA OF SODIUM VAPOR. 245 



achromatic telescope objective with an aperture of 12 cms. and a focal 

 length of 2 metres. The dispersion of the prism was sufficient to sep- 

 arate completely the monochromatic images of the lamp, any one of 

 which could be thrown into the aperture, the light passing by the edge 

 of the 90'' prism by means of which the fluorescent light was reflected 

 through a lens and thence upon the slit of the spectroscope. 



The sources of light which have thus far been successfully employed 

 in stimulating the fluorescence of the vapor are the following : quartz arc 

 lamps containing cadmium, zinc, and thallium ; ordinary arcs between 

 lead, silver, bismuth, and copper electrodes ; lithium, sodium, and 

 barium arcs, and vacuum tubes containing helium and hydrogen. 



Unfortunately the quartz lamps are very expensive, and become 

 almost useless after a run of about thirty hours, the surface becoming 

 granular and an opaque black deposit forming on the inner walls. As 

 exposures of eight hours are sometimes necessary, it will be seen that 

 lamps at S30 apiece, with an average life of thirty hours, make the 

 investigation an expensive one. 



The photographs of the fluorescent spectra obtained with mono- 

 chromatic stimulation are reproduced on Plate 2. 



After each exposure the D lines were recorded on the plate, so that 

 the difi'erent spectra could be brought into coincidence for purposes of 

 comparison : the D lines will be found at the extreme right in each 

 spectrum. The photographs have been reproduced as negatives, and 

 the point or points coinciding with the wave-lengths of the stimulating 

 light are indicated by arrows. A large scale drawing or chart of the 

 most interesting of these spectra, together with others too faint for 

 reproduction, has been made on cross-section paper, the points of 

 excitation being indicated by arrows, as in the photographs (Plate 5). 



Drawings of the complex spectrum and the magnetic rotation 

 spectrum made from the photographs obtained "with the large concave 

 grating will be found at the bottom of the chart : the former is a posi- 

 tive. At the top I have given a composite drawing which represents a 

 superposition of all the simple fluorescent spectra thus far obtained. 

 Immediately below it will be found the spectra excited by the complete 

 radiation of the cadmium and zinc tubes. In each case there are three 

 difi'erent exciting radiations simultaneously applied, yet it is almost 

 impossible to find two fluorescent lines which coincide. The other 

 spectra are excited for the most part by a single monochromatic radia- 

 tion, the wave-length of which is indicated by the arrow. I have not 

 yet obtained photographs of the fluorescent spectra excited by the 

 separated radiations of the copper arc or by the separated zinc lines 

 468 and 472 ; consequently these have been drawn together. It is 



