THE EVOLUTION OF THE ELEMENTS. 
127 
Each (undoubtedly) compound body presents us with a 
^characteristic spectrum. The spectra of compounds are charac- 
terised by channeled bands and fluted spaces ; the spectra of 
elements (so-called), on the other hand, are characterised by 
lines, and are comparatively free from bands and channeled 
spaces. 
If calcium chloride be volatilised in a flame the temperature 
of which is insufficient to cause decomposition of the chloride, 
no lines characteristic of the metal calcium are seen in the 
spectrum; if the temperature be increased the calcium lines 
make their appearance, while the bands due to the chloride fade 
away, and at last disappear altogether. 
The behaviour of calcium chloride may be regarded as 
typical of that of a compound body when subjected to gradually 
increasing temperature 
If an element A, known to be mixed with relatively small 
-quantities of another element B, be volatilised, the spectrum of 
the flame exhibits the characteristic lines of that element (A) 
which is present in larger quantity, and, added to these, the 
lines of the impurity (B) ; the greater the amount of B the 
greater being the prominence of its lines in the spectrum. 
The two cases — viz. that of a compound undergoing decom- 
position as the temperature increases, and that of a mixture of 
two elements — may thus be partially differentiated by means of 
the spectroscope. 
Suppose that a compound substance is volatilised at a tem- 
perature at which dissociation just begins, and suppose that four 
lines (a,/3, 7 , 8 ) are apparent in the spectrum; now let the temper- 
ature be increased, dissociation proceeds, the line S disappears, 7 
thins out, while (3 becomes prominent ; let the temperature be 
further increased, a process analogous to that already sketched 
proceeds, until the line a alone remains prominently shown. 
On reversing the process — that is, on passing from higher to 
lower temperatures — the spectrum gradually becomes more com- 
plex until the original condition is reached. It is, however, to 
be noted that the line a which is the sole strongly-marked line 
at high temperatures, does not disappear, but only becomes 
thinner and less marked at lower temperatures. 
Now if we suppose the substance exhibiting the original 
spectrum (a, /3 , 7 , B) to be a so-called element, and further if 
we suppose this element to undergo decomposition as the tem- 
perature increases, we should have just such a series of spectra 
as that described. 
In the paper already cited, Mr. Lockyer gives a diagram of 
certain of the lines in the spectrum of calcium as seen at tem- 
peratures varying from that of the Bunsen lamp to that of the 
star Sirius ; and this spectrum he pronounces to be just what 
