214 
PROFESSORS G. D. LIVEING AND J. DEWAR ON 
lines (with the exception of the blue line mentioned below) are single, and one set of 
those of magnesium triplets. We now record a second harmonic* series of potassium 
lines which appear to he pairs, and the violet pair, and possibly the red pair too, belong 
to this series. Lithium shows a second harmonic series of single lines high up on the 
scale. Calcium gives a long series of well marked triplets; zinc likewise gives 
a series of triplets ; aluminium gives pairs, and in the highest region triplets; thallium 
gives a series which seem to be quadruple groups with two of the four lines in each of 
much greater intensity than the rest. The alternations of sharper and more diffuse 
groups are generally apparent and are very marked in the cases of calcium and zinc. 
The diminishing distance and intensity and increasing diffuseness of successive repeti¬ 
tions of the same group as the wave-length diminishes, are in all the cases mentioned 
very plain. In all these cases the different lines forming a group are tolerably close to 
one another, so that successive repetitions of a group do not overlap one another, but it 
may be that in other cases the lines forming one group may be so far apart that the 
most refrangible line of one group may be more refrangible than the least refrangible 
line of the next repetition of the group ; the groups and their sequence will thus he 
much less easily recognised. 
Potassium. 
The ultra-violet spectrum of potassium, so far as we have observed it, is apparently 
one harmonically related series of which the first member above the visible spectrum 
is a double line just below the solar line 0 ; the next falls between Q and II, and the 
others follow at decreasing intervals, the seventh and last that we have observed 
falling just above U. It is only in the case of the line near O that we have been 
able to make sure that it consists of a pair of lines, but it is very probable that 
all are pairs in reality; all are strongly reversed, as might be expected from the 
volatility of the metal, and expanded when a fresh quantity of the metal or its 
compounds is introduced into the arc, so that the separation of the pairs, if such 
they be, could not be seen, while the more refrangible lines die away and are not 
recognisable as bright lines amongst the many lines which come out in the arc, as 
the alkali metal is dissipated. The line between Q and It, which is a strong fine, 
happens to he in a region where the lines of iron, manganese, and chromium lie very 
closely, so that we cannot pronounce with certainty that it is a double line. 
* By an “harmonic series” of lines we merely mean a series of overtones of a fundamental vibration 
we do not mean that they follow the simple arithmetical law of an ordinary harmonic progression, but 
are comparable rather with the overtones of a bar or bell than with those of a uniform stretched string. 
