91 



wandering insect, which is instantly lowered into the regions below, its 

 body sucked dry, taken up to the top, and jerked clear out of the way. 



So far I had been able to watch above ground, but I wanted to see 

 what was going on below,— /z^ze/ the larva supported its body in its 

 vertical burrow. To enable observation to be carried on, I obtained 

 a number of glass jars, which I filled with sani. Then close to the 

 side I made a hole a quarter of an inch diameter, ramming the sand 

 somewhat hard. I withdrew the small rod, and carefully cleaned the 

 inside of the glass, whose side form.ed part of the burrow. I next 

 turned out a larva from the home colony, and transferred it to the 

 glass jar, coaxing its tail first down the hole, which, when it had 

 entered, it descended to the bottom, examining the sides, evidently 

 not liking the transparent side. I watched it for some time, when 

 it deliberately took a mouthful of sand, moistened it, and then 

 plastered it over the glass. This act it repeated until it had plastered 

 the sand up an inch and a half high, or a quarter of an inch beyond 

 its own length. It then crawled up and took a look at me, as much 

 as to say, " Now you cannot see what I am doing behind the screen," 

 — which was quite true. So I turned all out, starting a fresh hole, 

 into which I once more put the larva, and once more it covered 

 itself in — was turned out and in majiy times, until at last it became 

 resigned to living in a hole exposed to the light on one side. After my 

 repeated failures I could now see how it used its legs and tail to climb 

 up its burrow ; also, that after fitting its head in at the top it drew 

 its abdomen up, bent it at the third and fourth segment, until the 

 swollen fifth was brought right up and almost touching the thoracic 

 plates. Then I observed the use of the two vertical spines on the 

 apex of the hump ; these were driven into the sand at right angles. 

 The remaining four abdominal segments curled round until the 

 anus reached the opposite side, into which the spines were driven ; 

 the larva now resting in a perfect zigzag form, which at the slightest 

 movement of any object above would in a moment straighten the 

 body and drop instantaneously to the bottom, and there remain until 

 danger had passed, when it would re-ascend sitting in this zigzag 

 position, waiting for hours and often all night (as I have proved by 

 watching with it), until some insect came within reach. I had not 

 been able to discover the use of the other pair of diagonal spines on 

 the fifth segment ; but I found out the larva could open them laterally 

 until they stood almost at right angles to the body, so bringing the 

 curved points into such a position that they would easily penetrate 

 the sand. 



Having noticed the ease with which the larva could throw its head 

 backward, I caught a fly and held it between the pliers, and gradually 

 bringing it near to the larva as it sat at rest. When within half an 

 inch of the hole it was suddenly seized and snapped from the pliers, 

 disappearing down the hole, but so rapidly taken that I could not 

 detect more than an instantaneous movement and the disappearance 

 of the fly. After repeated experiments I found that as soon as the 



