MEANING OF SHAPES AND COLOURS OF THE MEMBRA CIDER. 
277 
The larvae (Plate II. Fig. 4 a) of the species of this genus are described by Canon 
Fowler in the Bioloyia Cent rail- Americana as “ very curious, being of much the same 
shape as the perfect insect, but formed of separate upright narrow plates of different 
heights.” That these larvae protectively mimic the leaf-carrying ants is highly 
probable, far more so than in the case of the mature insects; for we have here the 
testimony of a biologist who observed the living insect in its natural habitat. 
Mr. W. L. Sclater, on returning from his journey to British Guiana in 1886, told me 
that on one occasion while collecting insects by shaking the branches of a tree over a 
sheet, his native servant, whom he described as a very acute observer, mistook one of 
these Membracid larva for a “ Cooshie ant ” carrying its fragment of leaf. 
Mr. Sclater brought the larva home, and it is figured in a short paper communicated 
to the Zoological Society (P.Z.S. 1891, p. 462, Plate XXXVI.). In this case we 
know that the thin flattened body is of a green colour like a leaf, while beneath it 
the legs and head are brown like the part of the ant which is not concealed by 
the leaf. 
It is of great interest that the remarkable forms of larva and perfect insect— 
although superficially alike—are produced in entirely different ways. In the larva 
the thin flattened shape is due to compression of the whole of the body rings behind 
the head, and every one of them contributes to form the sharp dorsal line which so 
much resembles the serrated margin of a leaf or a jagged edge gnawed by the 
mandibles of the ant. The same sharp line, forming a smoother sweep, is, in the 
perfect insect, made up by the edge of the pronotum alone. If, therefore, both larva 
and imago resemble leaf-carrying ants, the part representing the leaf is made up by 
all the segments in the one, and by the pronotum alone in the other. Both larvae 
and imagos probably live in the trees which the ants frequent for the purpose of 
cutting the leaves. 
At first sight it seems very difficult to account for the origin of such a case of 
protective mimicry, if indeed the interpretation here suggested be correct. It 
is, however, probable that the thin green body-form was gradually evolved to promote 
concealment among leaves, and that the few special details which suggest the ant 
were subsequently added. 
It is also of much interest that forms superficially resembling Membracis 
should be found in the Orthopterous genus Xerophyllum (Plate I.) where the dead- 
leaf-like appearance is not confined to the pronotum but is further carried out in the 
legs and head. The resemblance is clearly incidental and syncryptic. 
The appearance of the genera Phyllotropis and Cryptonotus (Plates III. and IV.) 
is not unlike that of Membracis. In the genera Enc/tophyllum and Enclienopa (Plates 
IV. to VI.) the pronotum is prolonged into a horn anteriorly, in some species bent, 
2 o 
