223 
towards the red end by partial oxidization, and they are then 
changed into colours of the group described below. 
The best way of preparing these chrysotannin colours is to 
digest leaves as free as possible from red colour, for a few 
days in enough cold ether to dissolve all the contained water, 
and proceed as described when treating of erythrophyll; but 
it is also very desirable to evaporate to dryness the solution in 
ether, separated from the water after agitation, and dissolve 
what may be soluble in water, filter, and evaporate to dryness. 
We thus obtain two different preparations, both soluble in 
water, and on examining and comparing them, they may be 
found to differ very considerably, and we may thus be able to 
establish the existence of two or more colouring matters, 
which were dissolved to a variable extent in the water and in 
the ether. In each case the aqueous solution should be 
evaporated to dryness at as low a temperature as is con- 
venient, and attention must be paid to any change in colour 
that may occur then, or subsequently when kept dry. 
I have found the following methods extremely useful in 
studying mixtures of different colours of this group, more 
especially in ascertaining the relation between the chrysophyll 
and the tannic acid sub-groups. Taking two of the experi- 
ment cells, | add to each so much of the aqueous solution of 
the two different preparations under examination, as will, for 
example, give exactly the same depth of colour with a little 
bicarbonate of ammonia, which is employed in order that one 
may not be more acid than the other, and that there may thus 
be a perfect equality as far as the colour is concerned. I 
then add a small quantity of iron alum along with some double 
tartrate of potash and soda, to prevent the precipitation of 
ferric oxide, and observe the relative depth of the dark 
colour due to the presence of any tannic acid. Perhaps it 
may then be found that one is made four or five times as deep 
as the other, as proved by dilution with water, from which 
we may at once conclude that some yellow colour is present, 
independent of the tannic acid, for, if not, equal amounts 
would have been made equally dark. Two other equal tubes 
may then be taken, and carbonate of soda added to each. At 
first they may be almost equally changed by the stronger 
alkali, but when left exposed to the air for a day or more, 
under a bell-glass lined with wet paper to avoid evaporation, 
both may gradually change, from the top downwards, to a 
fine red; but that preparation which had been found to con- 
tain four or five times as much tannic acid may become four 
or five times as deep a red. In such a case we may conclude 
that this red colour is due to the action of the air on the 
