Band and Line Speetra of the same Metallic Elements. 105 
AVE ele 
Further experiments devised for the purpose of ascertaining the conditions under which 
the bands of different characters are emitted by the flame spectra of the elements 
rubidium, caesium, calcium, strontium, barium, silver, and copper; and likewise 
to trace the origin of the bands. 
The coefficients of extinction of bands and lines respectively, in the spectra 
of the same elements, were obtained by burning gradually diminishing quantities 
of salts of the metals in the oxyhydrogen flame. For example, in the case of 
silver, 3°1 ers. of crystallized AgNO, (= 2 grs. Ag) were dissolved in 20 c.c. of 
absolute alcohol. Proportionate volumes of this solution were then dropped on 
to ashless filter-papers, and when dry were burnt in the flame in the usual 
manner. But it was necessary also to use larger quantities of some of the salts 
than could be satisfactorily ignited in this manner in one operation. In such cases 
the dry solid compound was employed: sometimes this was burnt on filter-paper ; 
in other instances it was introduced on slips of kyanite. Care was taken not to 
prolong the time occupied by the combustion or ignition, because this is tanta- 
mount to extending the period of exposure of the plate, and to elucidate the 
desired facts this would be inadmissible. It must be understood, howeyer, that 
when 2 ers. of the alkali salts were ignited, it was necessary to take four filter- 
papers, which quadrupled the period of exposure when only 0:5 gr. was ignited. 
This may be considered as the maximum of difference. 
All the following spectra were photographed on Wratten and Wainwright’s 
panchromatic plates, when not otherwise stated. 
RuBIDIUM. 
The salt used was RbCl. Photographed with Wratten and Wainwright’s 
panchromatic plates. 
Quantity of Description of Spectrum. 
Rubidium. ‘ : 
2 ers. There are three characteristic features in this spectrum: Ist, a con- 
tinuous spectrum, composed in part of fine lines, about the region of 
wave-length 4700; 2nd, a diffuse band appears in the red between 
7000 and \ 6700; 8rd, lines. The measurements here quoted are those 
made by H. Ramage on photographs taken from oxyhydrogen flame- 
spectra (Proc. Roy. Soc., 1902, vol. '70, p. 308). Several of the weaker 
lines do not appear on this spectrum, all the lines of which were 
independently measured, and are described. 
