162 THE ZOOLOGIST. 



shown in the accompanying sketch. To these spines strings 

 of buds were attached by silk, the buds themselves being 

 fastened to each other by silk. The buds were all green and 

 fresh, and it was quite obvious, after a few days' observation, 

 that the caterpillar took care to provide itself with a new supply 

 whenever the old one withered and turned brown, for strings of 

 such brown and withered buds were frequently found below the 

 food-plant, whilst the caterpillar itself was nearly always covered 

 with fresh green specimens. Once or twice the experiment of 

 removing by hand all the buds from the caterpillar was tried, 

 and in every case the larva at once set to work to cover itself 

 again ; a bud would be shorn off with the mandibles, then held 

 in the two front pairs of legs, and covered all over with silk 

 issuing from the mouth of the larva ; the larva then twisted 

 round the anterior part of the body, and attached with silk the 

 bud to one of the spinous processes ; another bud would then 

 be attached to this, and so on, until a sufficiently long string 

 (generally three or four buds) was made, when operations on 

 another spine would be commenced. 



The larva fed on the buds of the inflorescence, scooping 

 out the interior and leaving an empty shell; with its head 

 and perhaps two thoracic segments buried in a bud, it pre- 

 sented a comical appearance, reminding a non-entomological 

 friend of a cat with its head in a milk-jug. Very often the 

 empty shell would not be dropped, but attached to a spine ; in 

 fact, on some days it was noticed that there were more shells 

 attached to the spines than buds, and it was only under the 

 unnatural conditions induced by the experiments above noted 

 that the larva hastened to cover itself entirely with uneaten 

 buds ; the empty shells naturally withered more quickly than 

 the buds. 



When irritated the larva curled up in the attitude repre- 

 sented in the sketch, and it would remain in this position for 

 fifteen to twenty minutes ; it rarely moved much about, but 

 when it did, it was with a curious swaying motion, which may 

 have been part of the protective design, since it then looked 

 like a branchlet blown by a breeze, or else the insect was top- 

 heavy with the superposed buds. 



The larva pupated on the 28th, spinning up a silk cocoon 



