IN CONNEXION WITH THE SPECTRUM OF THE SUN. 
259 
the metal being hammered into it.] One of these cups with the chloride replaced the 
lower pole in the spark-holder, the upper one being composed of copper, that metal 
Fig. 4. 
Plan and section of cup used with salts. 
A. Aluminium wire. 
B. Cup-sliaped cavity drilled in it. 
C. Platinum wire. 
D. Flattened and split portion of the 
aluminium wire. 
Fig. 5. 
Aluminium cup placed in the spark-stand as in use. 
being selected as being a good conductor and giving a very simple and easily recognized 
spectrum. 
Chlorides observed. Lithic Chloride, Li Cl. — The wave-lengths of the longest lines of 
this metal are as follows : — 6705-2, 6102-0, and 4602-7, the latter (in the blue) being a 
wide and winged line. When the spectrum of the chloride is observed, the red line 
6705*2 is seen right across the spectrum, the orange, 6102-0, is faintly visible for about 
half the distance ; 4602-7 has vanished altogether. In the case of lithium this extinc- 
tion can be carried further in the flame reaction with an ordinary Bunsen burner, in 
which the red line 6705-2 is alone seen*. 
5895-0 i 
Sodic Chloride, Na Cl. — The D line ' I s by vei 7 l* ar ^ ie l° n g es t line in the 
sodium spectrum ; it is, in fact, the longest metallic line I have observed. After D 
* It is necessary in dealing with Li Cl and Na Cl to have the poles rather far apart (8 to 10 millims.), as, on 
account of the easy volatility of these chlorides, if the poles are close all the lines appear stretching across the 
spectrum. 
