68 Proceedings of the Royal Society 
a series of upper partials each, in an exactly similar inharmonic 
ascending scale, the tones of the heavier — potassium — would all he 
of lower pitch than the lighter — lithium. An absorption line of 
potassium in the visible spectrum will correspond in this case to 
one of lithium in the ultra-violet. Thus all attempts to establish 
exact mathematical relationships between the wave lengths of similar 
metals are for the present out of the question. We can but hope to 
find rough points of similarity, and these are forthcoming. The 
chloride of the alkaline earths are nearly related. The lithium and 
sodium spectra have some points of similarity, especially in their 
extreme simplicity; also potassium and rubidium, each with their 
five groups of lines. Then again the chlorides, bromides, and iodides 
of calcium and barium are nearly related, the lines shifting towards 
the red end of the spectrum in a way which is nearly proportional 
with the increase of atomic weight. 
Many salts absorb light of different wave lengths, and are in 
consequence of a definite colour. The colour is an index to this 
absorption and to the state of molecular vitiation of the salt mole- 
cules. In a paper on the colour of chemical compounds, recently 
published in the Philosphical Magazine ,* Professor Carnelley de- 
monstrated a relationship between salts of metals of the same group 
in respect to colour. The salts, say the chlorides, of a group of 
metals are much of the same colour, except that as you pass to the 
higher members of a group this changes somewhat, shifting uni- 
formly towards the end of the spectrum. This is illustrated by the 
following diagrams taken from his paper, in which only these metals 
of a group are given in alternate series : — 
Metal. 
Cl. 
I. 
Metal. 
Cr0 4 . 
As0 4 . 
Na 
White. 
White. 
Mg 
Lemon Yellow. 
White. 
Cu 
White. 
Cream. ' 
Zn 
Yellow. 
White. 
Ag 
White. 
Light Yellow. 
Cd 
Orange Yellow. 
White. 
Au 
Yellow White. 
Golden Yellow. 
H g 
Red. 
Yellow. 
The striking analogy between the above tables and those of taste 
already alluded to needs scarcely to be pointed out; it is obvious. But 
what does this shifting towards the red end of the spectrum indicate 
* Phil. Mag., July 1884. 
