842 WU. NELSON ANNANDALE ON THE [Dec. 4, 



flower was not a flower at all, but a Mantis. Even then it betrayed 

 itself by turning round and staring me in the face in tbe 

 manner characteristic of the Mantidse. When I held the branch 

 on which the insect had established itself in my hand, I could not 

 tell exactly where animal tissue commenced and where flower 

 ended, so perfectly was the one assimilated to the other, not only 

 in colour but in surface texture, and perhaps even to some degree 

 in structure. The Mantis had ensconced itself in the very centre of 

 the inflorescence, a position which it never assumed in the three days 

 during which it remained alive in my hands : it will be noticed in the 

 figures (pp. 840, 841), which are from photographs taken from life at 

 Aring, that the insect is clinging to the edge of a bunch of flowers. 

 The attitude which it adopted did not change with its position in the 

 inflorescence. In all cases, so long as the insect was on the watch 

 for prey, the abdomen was bent backwards until its dorsal surface 

 almost touched that of the thorax. The head and thorax were 

 held upright, and the fore limbs in front of them, in the ordinary 

 "praying" attitude of the Mantidse. The other two pairs of legs, 

 by which it clung to the flowers, were disposed round the upturned 

 abdomen, moi*e or less at right angles to the main axis of the body. 

 Sometimes this was the same as that of the inflorescence, but often 

 it was at an angle to it ; for the Mantis seemed indifferent as to 

 whether its head or one of its sides was uppermost. When once 

 a position had been taken up, it was never changed so long as the 

 insect remained on that particular inflorescence. Perfect as was the 

 concealment thus effected, I cannot say that the Mantis imitated a 

 single flower or part of a single flower with any great accuracy. 

 Perhaps the upturned abdomen might be taken to represent a bud 

 not yet opened, while the darker expansions on the femora showed a 

 greater resemblance to petals which had already unfurled themselves. 

 The petals of Melasioma, however, are more elongated than these 

 structures on the legs of the Mantis, which are almost circular in 

 shape, though their diameter is as great as, if not greater than, that 

 of the floral structures which they represent. What I can say 

 with certainty is, that a most marvellous resemblance is produced 

 between the insect Hymenopus bicornis and part of an inflorescence 

 of Melastoma polyantlium when the flower and the insect are 

 combined, as they sometimes are in nature. I cannot say that 

 they are always found together. An interesting point in con- 

 nection with this simulation of the flowers is the part played by 

 the green bar across the thorax of the Mantis. This bar divided 

 the prothorax from the rest of the body, and apparently separated 

 the insect into two parts, which appeared to have no connection 

 with one another on the inflorescence. In no ordinary in- 

 florescence of this plant are the flowers large enough to afford an 

 expanse of uniform coloration of the size of Hymenopus. 



During an unsuccessful search for further specimens in the 

 buffalo-lawn at Aring, the Mantis was placed, together with the 

 inflorescence to which it still clung, in a large box lined with dead 

 leaves. On the lid being taken off from this, after about half an 



