231 



During the course of this exploration of the walls of the burrow, the animal 

 was observed to turn about its long axis 260 degrees in one direction and 220 degrees 

 in the other, or a total of 480 degrees. The viscera were obviously twisted, owing to 

 the animal's being attached to the burrow at the posterior end, but this twisting 

 appeared to occasion it no inconvenience. Thus is solved the problem of how the shell 

 can be brought into the various positions necessary for excavating the regularly 

 cupped, perfectly rounded burrow. It was obvious from the markings shown in 

 figure 85, that the shell must have been rotated by slow stages through at least 180 

 degrees in each direction in order to produce these striations radiating in all directions. 



The possibility has often been suggested that Teredo ma\- facilitate its boring 

 by the use of some secretion which has a solvent action on certain constituents of the 

 wood. It would seem that the action of such a substance, if it occurs, should spread 

 at least for a limited distance through the cells of the wood immediateh' adjacent 

 to the extremity of the burrow. The tracheids of Douglas fir are from 1 to 3 mm. in 

 length, and it is hardly c()ncei\able that a secretion applied to one end of a tracheid 

 should not spread through its entire length. Probing in the extremity of the burrow 

 with a needle, however, did not re\-eal any area of softened wood, as would be expected 

 on this hypothesis. Staining with hematoxylin, which is a specific stain for cellulose, 

 did not reveal any diminution of the cellulose content of the wood at the end of the 

 burrow. It was further attempted to compare qiiantitati\ely the composition of 

 shavings from wood immediately adjacent to burrows of Teredo with that of shavings 

 from sound portions of the same timber. An analysis of these samples did not indicate 

 that any substances had been removed by the enzymes of the borers from the wood 

 forming the wall of the burrow. 



The e\idence strongh" indicates th,it boring is performwl entirely In- mechanical 

 rasping with the shell on wood that has been to some extent softened and macerated by 

 the presence of water in the burrow. 



The R.\te of Bokixc, 



The efficienc\' with which the shell of Teredo is applied to boring is amplv illus- 

 trated by the rapidity with which untreated timbers exposed to attack are reduced to 

 mere honeycombed masses of debris. In the case of hea\'y attack by either 'Teredo 

 navalis or Bankia setacea, under conditions obtaining in San Francisco Bay. from live 

 to six months is sutticient time for untreated piling to be entirel>' destr(jyed at the 

 mud line. 



The a\erage individual rate of boring under a gi\en set of conditions is difficult 

 to determine because of the practical impossibility of knowing accuratelv the time 

 at which an>- specimen had entered the v.ootl. As attachment of larvae occurs over 

 a considerable period of time, individuals of \arying sizes and ages occur together in 

 the same timber. To include the smaller specimens in the calculations in\-olves the 

 error of the differential time element, while to consider only the largest specimens does 

 not give a fair average. Howe^■er, on the basis of the latter alternatix'e, some approxi- 

 m,ation of the rate of boring has been attempted. 



The following tables of the rate of growth of Teredo navalis and Bankia setacea 

 are based on measurements of the largest individuals found in samples of Douglas 

 fir wood exposed for known lengths of time at the localities gi\en. Onlv the largest 

 specimen occurring in each sample is listed, on the assumption that such specimens 

 are the ones which entered the wood soonest after its exposure, and hence are the ones 

 whose ages are most accurately known. The error in this method of selection lies in 

 the fact that it includes those indi\iduals which not onlv are the oldest, but which 



