a Mantis of the kind called "floral simulators." 179 



rose-colour, that I should certainly have overlooked it, if it had 

 remained quiet. The legs were widely outspread, the head and 

 thorax were directed downwards, somewhat like the stalk of a 

 flower, the delicately coloured and veined wings slightly spread 

 out over the abdomen directed upwards. In this position the 

 creature at each approach began to move the wings to and fro, 

 evidently to act like a flower swinging in the wind. As however 

 a perfect calm prevailed this manceuvre led to the detection of 

 the Insect. Whenever I removed to a distance the wings and the 

 abdomen were depressed. My companion and I several times 

 tempted the creature to make use of this mimicry, and always 

 with the same result 1 ." 



The habits of Gongylus gongylodes in India have been 

 described by Dr Anderson, and his account is quoted at length in 

 the Cambridge Natural History, Vol. V. I need therefore only 

 say that in it the dilatation of the thorax has a petaloid coloration, 

 and that the insect puts itself in a position that causes it to look 

 like a flower, and that it makes swaying movements like a flower 

 shaken by a breeze. 



The three genera, Empusa, Gongylus and Idolum, are closely 

 allied. They all simulate flowers, but by means of colours placed 

 in different positions on the body, and by structures that are also 

 different in position and shape. All that the three have in common 

 is the instinct or habit of using the different characters they 

 possess for a similar purpose. 



To the considerations I have adduced we may add another — 

 subordinate, but bearing directly on the point under considera- 

 tion — viz. that in each of the numerous little dramas for the 

 capture of its prey that an Idolum goes through in the course of 

 its life, the attitude must be the first act. 



This study strongly suggests that in the evolution of the 

 peculiarities of this species the attitude has played a primary 

 part. 



Let me now briefly resume the points of this examination. 

 Idolum diabolicum, though apparently highly modified for its 

 special kind of life possesses no peculiarity but such as we may 

 assume from a study of other Mantidae to be capable of acquire- 

 ment independent of their advantage for the purpose to which we 

 see them put. The plate-like dilatations of the legs are of the kind 

 alluded to by de Saussure as probably connected with physiological 

 conditions of which we at present know little or nothing 2 . The 



1 Zool. Jahrbuch Syst., ix., 1897, p. 527. 



2 There is now some reason for supposing that the various sacs and expansions 

 (of which wings are the highest condition) so frequent in insects are connected with 

 the metamorphosis, and facilitate the accomplishment of that act, which is now 

 known to be one of great difficulty and danger. 



