266 
ME. J. NORMAN LOCKYEE ON SPECTRUM-ANALYSIS 
NOTES TO TPIE MAPS*. 
The lengths of the lines are given only in reference to those of the same element ; no 
relation is intended to be indicated between the lines of the various spectra as repre- 
sented on the Maps. The lengths of the lines of one spectrum as compared with another 
are liable to great variation — the lines of barium being, for instance, considerably 
longer than those of zinc, though they are represented as of the same length on the 
Maps. 
The poles were also much further apart in some cases than in others. In the case of 
barium, for instance, unless the poles were widely separated all the lines would have 
stretched across the spectrum, while in the case of manganese or nickel, unless the 
poles were close, no spectrum would be produced. The widths of the winged lines are 
mere approximations to the relative widths of the lines of the particular spectrum to 
which they belong, the width of a line in one map only relating to the lines of that 
map. No comparison can hence be made between widths given in different maps. 
Nor must the widths assigned to the lines be regarded as true; they have been 
purposely exaggerated. For instance, wings of the lines 4924 and 4911 (zinc) cover 
a length of 8 millims., equivalent on the scale to i0)0 ^ 00 -f millim. ; but it is not intended 
to assert that the wings extend over so large a space. 
Plate XI. Strip II.— LITHIUM. 
Whether re- 
Whether in 
Wave-length. 
Length. 
versed 
Solar 
trum. 
in the 
Spec- 
the Spark- 
Spectrum of 
the Chloride. 
Remarks. 
6705-2 
6102-0 
0 
5 
m 
Present, j 
These lines invariably stretch across the spectrum 
when metallic lithium is used as the electrode. 
4602-7 
1 
.3 
5? 
This line is nebulous hut brilliant; it is much 
o 
& 
shorter than either of the others. 
Plate XI. Strip III.— SODIUM. 
6160-0 1 
6154-2 J 
3 
Reversed. 
Not present. 
5895-0, \ 
5889-0 J 
4 
5? 
Present. 
5687-2 1 
5681-4 
3 
5154-8 
5152-5 
1 
Not present. 
Thauen has represented these lines as extending 
across the spectrum and longer than 5687-2, 
5681-4 ; hut this is not the case. 
The I) lines of the solar spectrum ; their tops have 
never been seen, as they invariably stretch from 
pole to pole. 
These lines are slightly longer than those at 
6160-0 and 6154-2. 
These lines are very short and faint. 
* Added during the printing of the paper, 
t Four one-millionths or forty ten-millionths of a millimetre. 
