HABITS OF THE ONTJPHIDJE. 309 



occasions, realised the position, and commenced biting 

 vigorously with their jaws at the lilting of the trough 

 in which they were placed, at the cut end of the 

 tube,- and at its outer skin, in which latter operation the 

 animal exposed fully 1^ inches of its body, bent U-shape 

 by crawling over the wall and down the outside of its tube. 

 All these proceedings differ entirely from the ordinary 

 behaviour of the worms as observed during several months, 

 both before and since. 



The worms were carefully watched for some hours, but 

 in neither case was a valve formed during that period, 

 though both worms, afterwards, when in the dark, soon 

 replaced them, and in one instance the damage was 

 actually repaired within 12 hours after the cutting of the 

 tube. Though I have not yet been fortunate enough to 

 see the complete process of forming these valves, I have 

 on several occasions during my experiments seen the worm 

 apparently operating with its mouth upon the newly- 

 formed membrane, on which, as well as upon the wall of 

 the tube itself, evidence seems to be left in the shape of 

 irregular coarse striations and cross-markings. It, there- 

 fore, appears probable that the unusual biting witnessed 

 immediately after the cutting of the tube, and subse- 

 quently, was an effort on the part of the worm to collect 

 material for the formation of new valves. A single pair 

 of very perfect, opposite valves was formed in each case, 

 which would suffice for the purpose of immediate protec- 

 tion. A few weeks later the tubes were slightly 

 lengthened, and a second pair of valves constructed. 



After an interval of ten weeks I experimented again 

 with the same two worms, being unfortunately unable to 

 get fresh specimens. On this occasion the newly-formed 

 anterior valves were cut off the tube of one specimen, and 

 the old posterior, or alternate valves from the other. The 



