379 
In very many other cases bisulphide of carbon is a most 
valuable reagent to separate different colours. An excess 
should be agitated with the alcoholic solution, when it sinks 
to the bottom, carrying with it the whole of some colours, 
and leaving the alcohol quite colourless. Other colours are 
only partially removed, whilst others are not abstracted in 
the least degree. These last are usually, but not always, 
those soluble in water, whereas those easily removed are 
usually, if not always, insoluble in the liquid. 
The cells for the examination of the spectra of solutions 
in bisulphide of carbon, ether, chloroform, or benzol, should 
be fixed to the glass plate by means of a mixture of glue and 
honey made so as to be quite stiff, but yet to melt when 
warmed. Cells so cemented must be washed out with the 
liquids named above, and never with water or hydrous alcohol. 
The spectra of colours dissolved in the bisulphide are far 
more characteristic than w'hen dissolved in any other liquid. 
So far, I have met with between two and three dozen different 
vegetable colours soluble in this reagent. The absorption 
bands are much farther removed from the blue end, when 
they occur in that part, so as to be far more distinct ; and 
their position is not subject to variation from any change 
produced by evaporation, as in the case of ordinary alcohol, 
w r hich may thus become more weak and give the bands in a 
very different situation. 
On the whole, the spectra are so characteristic and uniform, 
that different colours, soluble in bisulphide of carbon, may 
usually he recognised in a very satisfactory manner by means 
of the spectrum of that solution alone, and this is fortunate, 
since such colours can seldom be distinguished by their 
behaviour wfith reagents. The dried colour must be dissolved 
in fresh bisulphide, since when it separates from the alcohol 
it contains some of that liquid in solution, which alters 
the position of the bands. As an illustration of what may be 
done I give a few examples : 
Cheese. 
Cheese of orange colour was digested in bisulphide of 
carbon, the solution washed by agitating with alcohol, and 
evaporated to dryness. On redissolving in alcohol, diluted 
with a little water, a considerable amount of oily matter was 
separated ; and after evaporating the clear solution to dryness, 
and redissolving in bisulphide of carbon, it gave the spectrum 
5f 7-ib which corresponds exactly with that of red annatto. 
In all such experiments it is requisite to use deep experi- 
ment cells (I use one two and a half inches deep), so that the 
