190 Scientific Proceedings, Royal Dublin Society. 
lines being taken from the measurements of Eder and Valenta.* 
As there are not many strong gold lines in the red, orange, yellow, 
and green part of the spectrum, one or two lithium, sodium, and 
nitrogen lines are added. 
The easiest method of identification is undoubtedly to prepare a 
set of plates with spectra of all the elements, common and rare, 
and then identify lines by inspection and superposition of plates, 
and this method should be used when possible; but in searching 
for new substances, and when traces only are present, this method 
is often impossible, and a table like the present becomes almost 
indispensable. 
Measurement of Lines. 
In measuring lines, their position should be measured 
from two well-known gold lines on either side, and their wave- 
lengths determined from an interpolation curve. In measuring a 
whole spectrum, all the principal gold lines should be measured 
at the same time as the others, their values compared with the 
curve values, and any corrections necessary applied to the values 
of the lines between. This will correct any errors caused by an 
alteration in the focus of the instrument, or inclination of the 
plate, or by the slit being slightly out of the perpendicular, or 
even by the curve being slightly erroneous; so long as the identity 
of the gold lines is absolutely certain, other errors will be elimi- 
nated. It is thus very desirable that an observer should make 
himself thoroughly acquainted with the position, intensity, appear- 
ance, and grouping of the gold lines, so that he may be readily able 
to identify any one of them. When this is achieved, it is possible 
by use of the index to identify most of the lines on a photograph 
without the necessity of measurement. 
1 Eder and Valenta: ‘‘ Denkschr. kaiserl. Akad. Wissench. zu Wien,’’ Bd. lxvy., 
1896; Bd. lxvili., 1899. 
