219 
kind is similar to that found in the yellow interior of carrots, 
but sometimes mixed with more or less of the orange species, 
as may be proved, not only by the character of the spectrum, 
but also by the fact of our being able to partially separate 
them by the careful use of bisulphide of carbon and alcohol. 
Iam much inclined to believe that the more orange kind may 
be formed naturally by the partial oxidization of the other, 
for it is sometimes only developed after the leaves have turned 
yellow, and then undergo further change ; and the lowering 
of the absorption-bands towards the red end of the spectrum 
entirely agrees with this explanation, in accordance with the 
principles already described. 
3. ‘The erythrophyli group comprises a number of different 
colours soluble in water, in alcohol, and in ether, but not in 
bisulphide of carbon. ‘Those met with in leaves are usually 
more or less purple, made bluer by alkalis and redder by 
acids, and thus very similar colours might make a plant with 
a very acid juice bright red, and another with a more neutral 
juice, dark purple; whilst, on the contrary, two quite dif- 
ferent substances might give rise to exactly the same general 
tint. The erythrophyll may be obtained free from chloro- 
phyll and xanthophyll by heating the reddest leaves in alcohol, 
evaporating to dryness, redissolving in water, filtering, and 
evaporating to dryness at a gentle heat; but it then contains 
more or less of the colours of the following group. It is, 
however, much better to digest the leaves for a few hours in 
so much cold ether as will dissolve all their contained water. 
If the plant be very succulent, and too little ether used, the 
water is displaced and sinks to the bottom, with nearly all 
the colour in solution, and the ether is left almost free from 
erythrophyll. On agitating the ethereal solution in a test- 
tube with water, it sinks to the bottom, with nearly all the 
erythrophyll, whilst the chlorophyll and xanthophyll are left 
in the ether, along with a considerable part of the colours of 
the chrysotannin group. This ethereal solution should then 
be removed, as far as possible, by means of a suction pipette 
with an elastic top, and the rest by blotting paper, the 
aqueous solution evaporated to dryness at a gentle heat, re- 
dissolved in water, filtered, if turbid, and again evaporated 
to dryness. For this purpose the small saucers used by 
artists are very convenient, and the dry colour may be kept, 
sometimes for years, without any material change. It is, 
however, difficult, if not impossible, to effect a complete 
separation of the chrysotannin colours, and we must always 
be prepared to find the reactions of the various reagents 
modified by the presence of yellow colours belonging to 
