M. Huber on the Habits of a Saw-fly, 243 



fall by its own Aveight like a shred of the leaf. It will be sup- 

 posed that it is by means of its silk, skilfully employed, that 

 it succeeds in giving it the requisite position and fixing it in 

 its place. 



I shall give a full description of the method which it prac- 

 tises, a process in some degree analogous to that of the roll- 

 ing caterpillars, which consists in carrying threads from the 

 surface of the rolled portion to that part of the leaf which 

 they wish to join to their work. The weight of their body, 

 resting on the first threads, brings the leaf near to the centre 

 of the roll, and fresh threads tend to keep it in this position. 

 But the manner in Avhich our larva works every time that a 

 fresh portion of leaf is to be added to his case deserves greater 

 detail. 



There are generally three skeins or wefts * stretched from 

 the mouth of its case to the leaf. The first weft or skein is 

 situated on the body of the case ; this is the shortest, and 

 reaches the leaf by the shortest course ; the second proceeds 

 from the middle of the last whorl and also goes to the leaf; and 

 the third is fastened on still higher up, that is to say, at the 

 point where the rolling of the band begins. These wefts are 

 composed of threads parallel to each other, and nearly perpen- 

 dicular to the orifice. After the larva has bitten the leaf so far 

 as to separate enough for rolling up, it comes half out of its 

 case, mounts on the first weft and produces a fresh one which 

 I shall call No. 1 bis, similar, except that the threads are fixed 

 higher on the roll and on the leaf. After this operation it re- 

 enters its funnel, and comes out again at the space which lies 

 between the second and the third older wefts ; it mounts upon 

 the second, makes it bend with its Aveight, or perhaps by the 

 contraction of its body, and establishes a weft No. 2 bis, com- 

 posed of a score of silk threads ; lastly it places itself on the 

 third, and from tlience makes a new Aveft No. 3 bis. By this 

 means the work of rolling up goes on in succession, and the 

 whole portion cut off is rolled up at once, the spiral cone ac- 

 quiring a quarter of a whorl at each time ; it gains each day 

 a complete Avhorl. In fact, it is the cone which is rolled suc- 

 cessively on the bandage. Now the three Avefts bis become 

 fundamental, and serve as types for three new wefts, which in 

 their turn will be succeeded by others. Such is the tenor of 

 this process so far as I have been able to ascertain. The roll 

 turns during the operation of fixing the threads, and not at 

 all while the larva cuts out the bandage. We should mention 



* A row of parallel threads is obviously intended. The word trame of 

 the original has been rendered ivefl, with some doubt however of its beinar 

 the proper term. — R. T. 



R2 



