INTRODUCTION. XXVU 



they flash off on the instant ; even the falHng of a shadow disturbs them, 

 though its removal appears more startling. Banchi slightly disturbed will 

 stand motionless for one-and-half minutes with antennae straightened and, 

 even after the resumption of feeding, still keenly on the alert ; they alight 

 on Heracleum without hesitation, feed like Exetastes, with head deflexed 

 and abdomen elevated above the long hind legs, in which position they 

 may usually be secured between the fingers. They nearly always choose 

 the highest flower-heads, and are capable of emitting a defensive odour 

 resembling that of pomade. Exetastes cinctipes, Retz. { = osculatorius, Fab.) 

 and Perilissus praerogator alight gently, after hovering with pendent legs 

 and backward flowing antennae. Sometmies these insects (^e.g., Amblyteles 

 palliatoriiis) are found, apparently gorged, beneath the flowers. They 

 never crawl backwards, but are capable of turning in a wonderfully small 

 space ; thus I have seen a female Ephialtes, which measured, all parts 

 included, fifty-eight millimetres, turn completely round in an eight-and-a-half 

 millimetre tube, and the much '^X.ovlX.q.x IchiietDuon lantatorius of twenty-five 

 millimetres is capable of the same feat. 



The caution employed before feeding upon the nectar is most rigorous 

 in the protection of the mouth. First, each stylopod is investigated with 

 the tips of the perpendicularly curved antennae, which are then usually, 

 though not always, held forward, and not far apart, to ward off frontal 

 attacks, and the inspection is continued more closely with the maxillary 

 palpi. Satisfaction being attained, the stylopod is at length licked with the 

 ligula, much as a cat laps milk, the while the maxillary protect the flanks 

 and the labial palpi are trailed backwards to intercept all attempts from 

 the rear. Should the antennae meet with a Braconid, or other object 

 already in possession, the insect remains motionless for a moment with 

 antennae keenly alert, then moves gently off to another part of the flower. 

 These precautions would seem unnecessarily elaborate were it not that 

 upon the disc of the tables there sometimes lurks a rather large spider, 

 Misuinena vatia, Clk., so exactly resembling in colour and in its wide- 

 spread legs the flower on which it rests as to be practically invisible, at 

 least to human eyes. In its meshes I have found Ichneumon lautatorius 

 entangled and destroyed, having been attacked from beneath. The 

 vulnerable point in the Ichneumon's chitinous tegument is said to be at 

 the junction of the head and collar, and it is here that the Wolf Flies 

 (Dioctria) are enabled to pierce the membrane and suck the vital juices. 



It is now generally acknowledged that the sting of the Aculeates is a 

 modified terebra, but for long the faculty of Ichneumons for stinging was 

 a moot point. I cannot here enter into all that has been written upon the 

 subject, but I may say that beyond all doubt some kinds do and some do 

 not possess this faculty, though whether it be at certain periods in, or 

 throughout, their perfect existence does not appear plain. Anatomy shows 

 the females to be, at least sometimes, furnished with a linear and branched 

 poison gland {cf. Bordas, in Zool. Anz., 1894, p. 131), whose liquid is 

 stored in a vesicle adjoining the terebra. Westwood truly says that Ich- 

 neumons held in the fingers always try to sting, but usually unsuccessfully. 

 William Kirby states he has been stung by a species of Ephialtes and 

 by another Ichneumon with a concealed terebra. I have often allowed 

 them to do their worst, with no result, though once a Paniscus, with which 

 I was not experimenting, succeeded in puncturing the skin and the pain in 

 this case was but momentary. Butler tells me that the large Ophioninae 



