Pottok—Lines of the Spark-Spectra of the Elements. 187 
Of course, if silver or copper is present in the electrodes, they 
cannot be used for testing solutions for those metals. 
‘When the point at issue is the presence or absence of 
some impurity, or foreign substance, in a particular sample, 
a very convenient method is to photograph a pure sample of 
the substance, using a long slit, then shorten the slit, and 
photograph the suspected sample, when any new substance 
present will be seen at once by the short lines; and if no short 
lines are seen, the identity of the two samples is demonstrated. 
Spectrum 7 shows a sample of pure aluminium chloride, 
photographed against gold; and spectrum 8 shows a sample 
of aluminium chloride, containing beryllium, photographed 
against pure aluminium chloride and gold, the short lines 
indicating, at a glance, the nature of the new substance 
present. 
Plate XI. gives the spectra of a number of other elements 
photographed against gold. The spectra extend from A 5896°2 
to about A 2500; but the plates were not very sensitive below 
gold line No. 9. X 4792°8. The two lines just before gold line 
No. 13 are the H and £ lines of calcium, and are derived from 
the air, and also from the capillary glass tube placed round the 
gold electrode to supply the solution. They are, practically, 
always present, and are very useful, as they serve to at once 
identify aluminium, lanthanum, and a number of other elements, 
having lines that lie between them. 
Unfortunately, the spectra are somewhat indistinct in the 
plates; but they are sufficient to indicate the general method 
-of procedure. 
Intensity. 
The intensities given in the Tables refer, in some cases, to the 
spark-spectrum of the metal, and in others, to the spark-spectrum 
of a solution of the chloride. Some of the intensities are marked 
from 1 to 10, while others are numbered from 1 to 100; and there is 
a general want of uniformity in the standard of intensity adopted 
by different observers ; but it will involve much labour, and take 
a long time to reduce the spectra of all the elements to a uniform 
‘standard of intensity. In the red and yellow end of the spectrum 
