286 



so that their swimming mov^ements probably avail little more than to keep them afloat 

 as they are washed about from place to place. 



The length of this free swimming larval period is unknown. Repeated attempts 

 to rear the larvae under aquarium conditions have failed, and attempts to carry the 

 organisms through the larval period in cages covered with silk gauze and suspended 

 in the bay have likewise been unsuccessful. A week is the longest period that any of 

 the larvae have been kept alive, and during this time they showed no inclination to 

 settle on wood introduced for that purpose. Thus it appears that the free-swimming 

 period is longer than one week, and there is some slight evidence that it may be in 

 the neighborhood of a month or even more. In 1921 the earliest veliger larvae were 

 found at Crockett on June 29, w^hereas the first settlement of larvae on test timbers 

 at this locality was observed on August 8, somewhat more than five weeks later. Such 

 evidence is of extremely doubtful value, however, as the possibilities of overlooking 

 either the first veligers or the first settling larvae are practically infinite. 



The breeding season and period of larval settlement in general include the summer 

 and autumn months of the year. Along the north Atlantic coast settlement of the 

 larvae of Teredo navalis begins in June (Kindle, 1918) or the early part of July (Nelson, 

 1922). In San Francisco Bay the season does not begin as early as this except at one 

 locality. The dates of first settlement of larvae at certain localities in San Francisco 

 Bay during three successive years are given in the following table : 



TABLE No. 42 



DATES OF FIRST SETTLEMENT OF TEREDO LARV.-XE IN SAN FRANCISCO BAY 

 DURING THREE SUCCESSIVE YEARS 



With reference to the dates of first settlement at Dumbarton correction should 

 be made for the fact that test timbers at this place were examined at monthly intervals 

 only so that the earliest settlement in any year might have occurred at any time 

 during the month preceding the date given. There is some reason to believe that 

 larval settlement may occur here as early as July 1. At the other localities inspections 

 of test timbers were more frequent, and the very early stages of attack were discovered 

 on the dates given. The precise dates are not significant, how^ever. They should be 

 interpreted in a general way as indicating the month and approximate time of the 

 month when larval settlement began. 



Thus it appears from the table that settlement of teredo larvae on exposed 

 timbers at Dumbarton begins regularly about a month earlier than at Goat Island. 

 There is nothing in the slight differences in salinity at these two localities to suggest 

 a reason for this difference in dates of settlement of larvae, especially in view of the 

 fact that salinities at both places are always above the minimum salinities that appear 

 favorable to larval settlement in the upper bay. Rather should we attribute the earlier 

 breeding activity at Dumbarton to the fact that the temperature of the water in the 

 southern reaches of the bay rises more rapidly during the early spring and summer 



