90 BARNARD 



apparent that they were hatched from egg-bearing females of the first 

 population, since they were of the same size as hatched juveniles observed 

 in the laboratory. The effect of this change in character of the early popu- 

 lations (from an entirely adult to a mixed population including juveniles) 

 on the average body length of the animals at each collection is illustrated 

 in fig. IB. 



One may consider two hypotheses concerning migration in chelurids : 

 (1) juvenile chelurids attempting to migrate are unable to survive on 

 the new wood or are subject to predation by other animals; (2) adult 

 chelurids only are subject to migration pressures, possibly in conjunction 

 with mating behaviour. The first hypothesis was tested when fresh 

 wooden blocks were placed within two inches of chelurid infested wood, 

 eliminating any long migration path; but no juveniles appeared until 

 some time after the arrival of the first adults. The facts that more 

 juveniles than adults are present in an established population of chelurids 

 and that juveniles, unlike adults, are small enough to invade individual 

 limnoriid holes are evidence that juvenile chelurids do not migrate. Un- 

 fortunately, as the writer has been unable to observe chelurids mating, 

 the second hypothesis must remain unproved. It is possible that migra- 

 tion is a passive result of mating by the chelurids while swimming in 

 the water outside the burrows and that after mating the animals seek 

 a protective niche, which may or may not be the same wood from which 

 they came. Laboratory experiments show that chelurids are unable to 

 return to the wood from which they swim (unless by accident) if they 

 are more than three inches from it. When farther away than this they 

 swim in irregular paths until within three inches of some large, opaque 

 object, toward which they then swim. 



CHELURID BURROWS IN NATURE 



On test blocks exposed for short periods of time (2-3 months) chelu- 

 rids are found in the uncovered and abandoned limnoriid tunnels and in 

 the large caverns formed from the combined action of limnoriids. The 

 laboratory experiments show that chelurids, unlike limnoriids, do not 

 bore discrete, circular burrows but engage in a browsing type of erosion 

 resulting in hemicylindrical furrows. This same kind of furrowing is 

 found in nature in the enlarged and unroofed limnoriid burrows, some 

 of which must be produced by chelurid activity. Blocks of wood infested 

 only with limnoriids were dried to kill the gribbles, then introduced into 

 laboratory aquaria containing from 20 to 50 chelurids. Within a month 

 of exposure, many of the discrete limnoriid burrows had been unroofed 

 and interconnected to form furrows. 



