Pottok— Lines of the Spark-Spectra of the Elements. 189 
indicated Hartley’s results by the use of the last letters of the 
Greek alphabet :— 
@ = seen with 1 part in 100 of solution. 
x = seen with 1 partin 1,000 of solution. 
w = seen with 1 part in 10,000 of solution. 
w = seen with 1 part in 100,000 of solution. 
ll 
Thus the letter ~ after the intensity of an element indicated 
that this line is still visible when the solution contains :01 per 
cent. of the element, but that it is not visible for less quantities, 
such as ‘001 per cent.; whereas w would indicate that even with 
this minute quantity in solution the line could still be seen. 
Of course, some elements do not give any w lines, and some 
do not even give ~ lines. To determine whether an element 
is present or not, it is only necessary to look for the most per- 
sistent lines, and if these be absent the element is not present, even 
if there be several lines closely agreeing with a few of its lines; as 
already explained, they must be attributed to some other element. 
To determine the presence of an element, one need only look for 
the six most persistent lines; these lines we will call the residuary 
lines of the element, but they may be marked x, W, or w, accordingly 
as they are seen in a ‘1 per cent., ‘01 per cent., or ‘001 per cent. 
solution. 
Identification. 
To facilitate the identification of lines, I have marked strong 
electrode lines at convenient distances with numbers as shown 
on the gold cadmium spectrum on Plate X. These lines are 
similarly numbered in the index to enable lines that lie between 
them to be identified by inspection. ‘Thus, if a line is observed 
a little beyond the gold line No. 20, the first of a very 
characteristic group of three lines, and its distance from the 
last of the group is just about equal to the space between each 
member, you can at once conclude that it is Al or Yt, and could 
not well be any other element. All the gold lines, and such 
silver lines as are usually shown by gold of an ordinary degree 
of purity, are printed in heavy type; and a list of those lines 
is given in a preliminary table, the wave-lengths of the gold 
