Unities, some notes on the breeding season and a statement regarding 

 its specific identity. For locations mentioned herein the reader is re- 

 ferred to the map accompanying last year's report. 



The data and graph of salinities utilized in this discussion were 

 prepared under the direction of the Sub-Committee on Salinities, 

 Charles E. Cortes, Chairman. The analyses at Crockett were made in 

 the laboratory of the California and Hawaiian Sugar Refining Com- 

 pany; at Martinez in the laboratory of the Shell Oil Company; at Avon 

 in the laboratory of the Associated Oil Company; at Bull's Head Point 

 in the laboratory of the Mountain Copper Company; at Port Costa, 

 Black Point, Greenbrae and Tiburon in the laboratory of the Wood 

 Preserving Plant of the Southern Pacific Company. 



Action ol Teredo in San Francisco Bay in 1921 



The destructive action of Teredo in San Francisco Bay in 1921 

 lias continued in unprotected timbers wherever available, as is shown 

 by the infection of experimental timbers in the bay proper, and in 

 Carquinez Strait as far up as Rowe Island Light, opposite Bay Point, 

 eleven miles above the lower entrance of the Strait and well into Suisun 

 Bay. Since there is far less unprotected piling accessible to the borer 

 this year than in the previous years, the infection and the number 

 of larvae available for new settlement .are both reduced in extent. 

 Their action in favorable localities, as for example in Carquinez Strait, 

 has been typically destructive. New untreated piling driven in April, 

 1921, in 36 feet of water at Crockett were attacked so heavily, especially 

 near the mud line, as to be broken off in December. The first settle- 

 ment of larvae at Crockett was detected on the test timbers on August 

 8. Penetration of piling to a depth of 1% inches was attained by No- 

 vember 8 and specimens 4 inches in length were found in test timbers 

 on that date. 



In 1920, piles driven at Crockett in January were attacked by 

 Teredo at some time in the spring and were penetrated to a depth of 

 two inches by May. This indicated conclusively the settlement of lar- 

 vae within that period. The interpretation was then made that these 

 were larvae of the year and not holdovers from 1919. However, in 1921 

 no settlement was detected on test timbers examined at monthly inter- 

 vals until July 20, as is shown in Table 8. 



