232 



also have grown most rapidly. This error is to some extent compensated for by the 

 consideration that in estimating the ages of the various individuals measured a con- 

 siderable margin has been allowed, the "presumed date of attack" being set at the 

 earliest time at which it is believed that larval settlement might have occurred. 



A majority of the figures given represent measurements of specimens occurring in 

 test boards immersed for stated periods of time during the breeding season. In such 

 cases the presumed date of attack is arbitrarily set at two weeks after the date of 

 immersion, as it has been observed that test boards of seasoned lumber are rarely or 

 never attacked until after they have been in the water for this length of time. Some of 

 the figures tabulated (those marked with an asterisk) are measurements of specimens 

 occurring in test boards immersed prior to the commencement of the breeding season, 

 or in piling exposed for other than experimental purposes. In these instances the pre- 

 sumed date of attack is set at two weeks prior to the earliest date at which larval 

 settlement is known to have occurred in the year in question. The presumed age is 

 therefore the maximum that is considered at all possible. Even so, the rate of boring 



TABLE No. 28 

 RATE OF BORING OF TEREDO NAVALIS IN SAN FRANCISCO BAY 



