346 TROPICAL NATURE v 
This caterpillar feeds upon the orange tree, and also upon 
a forest tree (Vepris lanceolata) which has a lighter green 
leaf; and its colour corresponds with that of the leaves it 
feeds upon, being of a darker green when it feeds on the 
orange. The chrysalis is usually found suspended among the 
leafy twigs of its food-plant, or of some neighbouring tree, 
but it is probably often attached to larger branches ; and Mrs. 
Barber has discovered that it has the property of acquiring 
the colour, more or less accurately, of any natural object it 
may be in contact with. A number of the caterpillars were 
placed in a case with a glass cover, one side of the case being 
formed by a red brick wall, the other sides being of yellowish 
wood. They were fed on orange leaves, and a branch of the 
bottle-brush tree (Banksia sp.) was also placed in the case. 
When fully fed, some attached themselves to the orange 
twigs, others to the bottle-brush branch, and these all changed 
to green pupx, but each corresponded exactly in tint to the 
leaves around it, the one being dark, the other a pale faded 
green. Another attached itself to the wood, and the pupa 
became of the same yellowish colour, while one fixed itself 
just where the wood and brick joined, and became one side 
red, the other side yellow! These remarkable changes would 
perhaps not have been credited had it not been for the pre- 
vious observations of Mr. Wood; but the two support each 
other, and oblige us to accept them as actual phenomena. It 
is a kind of natural photography, the particular coloured rays 
to which the fresh pupa is exposed in its soft, semi-transparent 
condition effecting such a chemical change in the organic 
juices as to produce the same tint in the hardened skin. It 
is interesting, however, to note that the range of colour that 
can be acquired seems to be limited to those of natural 
objects to which the pupa is likely to be attached, for when 
Mrs. Barber surrounded one of the caterpillars with a piece 
of scarlet cloth no change of colour at all was produced, the 
pupa being of the usual green tint, but the small red spots 
with which it is marked were brighter than usual.+ 
1 Mr. E. B. Poulton has since greatly extended these observations, both in 
pupe and larve, with very remarkable results. See Proc. of the Royal 
Society, No. 243, 1886; Transactions of the Royal Society, vol. clxxviii. B., 
pp. 311-441. These are briefly described in Darwinism, p. 197, and more 
fully in a volume by Mr, Poulton on The Colours of Animals, 1890, 
