218 
then, when cold, agitating in a test-tube with bisulphide of 
carbon. ‘This sinks to the bottom, holding nearly the whole 
of the chlorophyll in solution, whilst nearly all the xantho- 
phyll remains dissolved in the alcohol, along with all the 
various substances soluble in water. Leaves having an acid 
juice must not be used, for that would change the normal 
chlorophyll into another modification, which gives an entirely 
different spectrum; nor, in any case, should the solution be 
left long in contact with them, for then the separation of the 
chlorophyll and xanthophyll is comparatively very imper- 
fect. 
My reason for thinking that there is a mixture of two sub- 
stances in the normal chlorophyll of leaves is, that when the 
spectrum of the light transmitted through them is com- 
pared with that of the light transmitted by certain species of 
Oscillatorie, it is seen to differ in having a very narrow ab- 
sorption-band in the red, which is absent in the case of the 
Oscillatorie. The chief absorption-band is the same in both; 
and their relations may easily be explained by supposing 
that the chlorophyll of the Oscillatorie is a single substance, 
and that in most green leaves it is mixed with another, which 
gives a somewhat different spectrum, but has such very 
similar properties that they have not yet been separated. 
When dissolved in bisulphide of carbon there is an analogous 
difference in the spectra, but the above-named narrow ab- 
sorption-band is not nearly so distinct as in the spectrum of 
the light transmitted by the well-illuminated leaves. 
2. The xanthophyll group also contains several distinct 
species, but only two are common in leaves, one being more, 
and the other less, orange. These respectively may be 
obtained separate by heating in alcohol the orange exterior 
and the yellow interior of some carrots, and agitating with 
bisulphide of carbon, which removes much of them from the 
alcohol. ‘They are insoluble in water, but soluble in alcohol 
or bisulphide of carbon ; and, when dissolved in this latter, 
their spectra show two rather indistinct absorption-bands at 
the blue end, but the red, yellow, and yellow-green rays are 
freely transmitted. In preparing xanthophyll from yellow 
leaves, it is well to digest them for some time in considerably 
less alcohgl than would dissolve the whole of it. This 
removes the greater part of the unchanged chlorophyll, and 
on digesting them in fresh alcohol, and agitating with bisul- 
phide of carbon, or on evaporating to dryness and dissolving 
in it, a tolerably pure preparation may be obtained. By ex- 
amining the colouring matter thus extracted from various 
leaves, I have been led to conclude that the most common 
