These labours often come to an untimely end ; a sudden swirl in the 

 water and the insect, losing its foot-hold, is swept away and has to begin 

 all over again in another place. It is not the habit of the larva to 

 work with the whole body exposed, and this instinct would save a larva 

 from a catastrophe of this kind, but here having no case to speak of it 

 is compelled to act otherwise and take the risks. 



So far this description will apply either to C. pnsio or the stalk- 

 cased larva (No. i), but now differences begin to appear ; C. pusio has 

 only to build its arms, but the stalk-cased larva has a more elaborate 

 work to perform. This latter larva having built a straight tube about 

 3 mm. long on the surface of the dish, continues the tube at almost a 

 right angle to its former direction so that it projects freely into the 

 water ; this, the anterior end, is made more and more compact, while 

 from the posterior end the larva now continually tears away particles 

 to apply in front, and goes on in this way, robbing one end to add to 

 the other, until the whole of the horizontal tube has been destroyed. 

 The Stalk. Meanwhile by the transposition of these materials a stiff upright tube 



has been formed only attached to the dish at its base ; but now the 

 larva has to alter its tactics for if it continued to remove material from 

 the base the whole case would be dislodged and swept away. In 

 order to avoid this the larva is very careful in taking away 

 materials from behind, to leave a narrow strip uninjured on one 

 side ; and as further particles are taken away all round, this strip 

 gradually grows longer and longer until it appears as a stalk 

 supporting the short case at its free end. In order to strengthen 

 this stalk the larva turns round in its tube and, appearing at the 

 posterior end, plasters it vigorously with its labrum, paying special 

 attention to the point of attachment that all may be safe. When 

 removing material to build on in front, the larva comes out of the 

 anterior end and, bending over, stretches down outside the tube as 

 far as it can reach, without breaking its habit of keeping three or 

 four segments within, and tears off the most distant particle it 

 can reach (see fig. 35) ; this distance will of course be less than the 

 length of its body, so the case, although continually growing in 

 front, never becomes as long as the insect and is in fact always 



