178 H. G. SMITH. 
packed so closely that hardly any space separated them; they 
were feeding on the liquid as it exuded from the hole in the bark. 
In no instance were the ants eating the darker exudation, nor 
did they appear to be partial to the white when it had solidified, 
only one or two stray ones being near it when in that condition. 
The darker substance contains both tannic acid and eudesmin, 
which J presume is the reason why it is objectionable to them. 
The punctures or bores into the bark were often small, and 
appeared to be directly inwards towards the centre of the tree. 
From the same tree from which both the white and dark saccharine 
materials were obtained, some pure kino was also found exuding 
at the same time. It is remarkable that three substances differ- 
ing so much in composition, should be exuding from the one tree 
at the same time, and I set myself the task of attempting to 
solve the problem. From the specimens taken from the trees, 
it will be seen that :— 
(a) When the puncture has not penetrated entirely through the 
bark and a flow is set up, the exudation is quite white and 
consists largely of raftinose (melitose). 
(6) When the puncture has just penetrated through the bark 
the exudation is contaminated with tannic acid and eudes- 
min, showing that eudesmin is present in the cells of the 
tree with the tannic acid. 
(c) Also, that when the puncture or boring of the larva has 
continued into the wood of the tree, pure kino is produced, 
providing the sugary sap of the bark is not exuding at 
the same time. 
This indicates that the kino is not obtained from the bark 
directly, but from the wood of the tree, and that the sap of the 
bark of this tree does not contain tannic acid, but consists prin- 
cipally of the sugar raffinose (melitose). Although tannic acid 
could not be detected in the white manna, yet, when the bark 
was boiled in water, a small quantity of tannic acid was found in 
the solution. In all instances it was seen that: whenever kino 
was exuding, it was coming directly from the wood of the tree 
