148 CERURA VINULA. 



sides. The saddle has a distinct white line as an edging, 

 but this becomes more yellow round the flaps and 

 spots ; the hump on 4 is coloured pink on the sides ; 

 the sides of the larva are rich green, of a somewhat 

 yellow-green tint; the belly full green, with a purplish- 

 brown streak on 5 — 7, and two purple-brown marks 

 on 11 — 13 ; on 11 and 12 a pair of purple spots, on 

 13 two pairs. The face is brownish- buff, the lobes 

 bordered with black, the mouth black, the fold of the 

 neck rose-pink all round, the places of the horns are 

 two velvet black spots ; the spiracles white ringed 

 with black, and segments 3 and 4 have indications of 

 false spiracles ; the thoracic legs black, with two yel- 

 lowish-green rings, the ventral prolegs green, with 

 black transverse marks halfway down, and some brown 

 bristles on the lower mark ; the tails are pale-green 

 set with black points ; the whips are pink. 



When full-fed the larva assumes a sad brown tint, 

 and if supplied with wood will soon set about its 

 cocoon. I have found the cocoon on the bark of poplar 

 trees, and once I took a cocoon, which I still possess, 

 from a stone wall close to a poplar tree ; the stone was 

 a coarse red sandstone, and the cocoon, thinner than 

 usual and not so hard, was coated all over with little bits 

 of this. I have also a cocoon made on cardboard, but 

 this is very stout and hard ; those I have on touch- 

 wood are very hard, being more than 2 mm. thick in 

 their substance, besides the outer covering of bits of 

 the wood. This summer I watched several larvaa mak- 

 ing their cocoons. Having chosen a spot on the wood, 

 a larva would begin by spinning some coarse sticky 

 threads in a random manner from side to side ; it then 

 appeared to get its head under them, and to bite out 

 with a jerk little bits of wood and stick them here 

 and there on the threads. In this way it would proceed 

 until just half the body was covered with a loose tangle 

 of threads and bits of wood, from half an hour to three 

 quarters being the time thus spent ; it would then turn 

 its head to one side, and commence widening the base 



