216 
very convenient to divide animal and vegetable colouring 
matters into three groups, distinguished by the manner in 
which they behave when their solutions in water are treated 
with sulphite of soda. Thus— 
Group A.—Absorption removed when the solution is alka- 
line with ammonia. For example, changed from deep 
blue to pale yellow. 
Group B.—Absorption not removed when the solution is 
alkaline with ammonia, but removed when it is acid 
with citric acid. Thus, for instance, altered from 
deep pink to nearly colourless. 
Group C.—Absorption not removed even when the solution 
is acid with citric acid. No alteration in the colour. 
It is also sometimes most useful to arrange different 
colours into other groups, distinguished by optical and phy- 
sical peculiarities. This is particularly the case in the pre- 
sent inquiry, since it enables us to group together a consider- 
able number in such a manner as to make the subject far 
more simple and intelligible. ‘The individual species of these 
groups, distinguished by the number or position of the ab- 
sorption-bands in their spectra, are changed in an analogous 
manner, when acted on by reagents, so that the general rela- 
tion between the rays of light absorbed by the various modi- 
fications or new products is the same as that between those 
absorbed by the separate substances in their unaltered state. 
This sort of connection between the chemical and optical 
changes is of very great use in such inquiries as the present, 
since, when we know the character of the optical change, we 
may often form a tolerably satisfactory opinion on the nature 
of the chemical. This may perhaps appear hazardous, but is 
really only an extension of a principle constantly employed 
by chemists. Most vegetable blue colours belong to groups in 
which the broad absorption is raised by acids “from the red 
towards the blue end of the spectrum, and lowered by alkalis. 
On adding any solution to a blue vegetable colour, we there- 
fore conclude that, if it be turned red, an acid, but, if green, 
an alkali, is present. Now, I find fhat % in the case of very 
many kinds of colouring matters a slight amount of oxidiza- 
tion lowers the absorption towards the red end of the spec- 
trum, quite as uniformly as it is raised towards the blue end 
by acids, and lowered by alkalis, in the case of those just 
named ; and, therefore, when we find that such a change has 
occurred, it seems reasonable to infer that it has probably 
been due to this action of oxygen, if the circumstances of the 
case agree with such an explanation. 
