160 Mr. R. Meldola on Evolution. 
its food-plant, in which attitude it ‘has almost exactly the 
appearance, in form, colour, and size,” of one of these shells, 
“which greatly increases the difficulty of finding them when 
thus feigning death”*. ‘The habit of rolling up into a ring 
when alarmed is common with many caterpillars which are 
found in situations where mimicry of shells cannot possibly be 
adduced as a reason for the habit. Hence in the case of L. 
caniola I am inclined to believe that natural selection has 
taken advantage of and improved upon a habit originally 
acquired for a distinct purpose. 
The most remarkable case referable to the present class 
that has recently been published is that of Gongylus gongy- 
lodes, Limm., an Indian Mantis which simulates a flowery. 
When exhibiting some of these insects at a meeting of the 
Asiatic Society of Bengal, Dr. Anderson remarked that when 
seen from above “ they did not exhibit any very striking fea- 
tures beyond the leaf-like expansion of the prothorax and the 
foliaceous appendages to the limbs, both of which, like the 
upper surface of the insect, are coloured green ; but on turning 
to the under surface the aspect is entirely different. The leaf- 
like expansion of the prothorax, instead of being green, is a 
clear pale lavender-violet, with a faint pink bloom along the 
edges of the leaf; so that this portion of the insect has the 
exact appearance of the corolla of a plant—a floral simulation 
which is perfected by the presence of a dark blackish-brown 
spot in its centre, over the prothorax, and which mimics the 
opening to the tube of a corolla. A favourite position of this 
insect is to hang head downwards among a mass of green 
foliage; and when it does so it generally remains almost 
motionless, but, at intervals, evinces a swaying movement as 
of a flower touched by a gentle breeze ; and while in this atti- 
tude, with its fore limbs banded violet and black and drawn up 
in front of the centre of the corolla, the simulation of a papilio- 
naceous flower is complete. ‘The object of the bright colouring 
of the under surface of the prothoracic expansion is evident, 
its purpose being to act as a decoy to insects, which, mistaking 
* Newman’s ‘British Moths,’ p. 473. 
t Proc. As. Soc, Beng., Aug. 1877. For an analogous case see a paper 
by Mr. Wallace in Macmillan’s Mag. for Sept. 1877. The Mantis re- 
ferred to resembles a pink orchid, and is stated to be attractive to butter- 
flies. Prof. J. Wood-Mason informs me that the floral resemblance of 
the above and other species of Gongylus has been known to him for years; 
but its object had remained unexplained till 1875, when he received from 
Assam some larvee of Hymenopus bicornis, Serville, in which species the 
resemblance to a flower is, according to Prof. Wood-Mason, even more 
erfect than is the case with the Gongylus. See also Proc. Ent. Soe., 
Noy. 7th, 1877, p. xxix. 
