329 



the minimum salinity over a limited space of time determines the distribution of 

 Limnoria. This is of course the case only when the a\erage does not remain below 

 6.5, which we have found to be lethal in 24 hours. 



Referring again to the distrii)ution in San Francisco Bay, it will be observed 

 from a study of the salinity graphs, figures 9,? and 94, that at Black Point, a locality 

 a few miles above the uppermost limit cif the occurrence of Limnoria in .San Pablo 

 Bay, the average salinity approximates 11 parts per 1000, with occasional drops to 

 5 parts per 1000 or less. These figures should be weighted somewhat for the fact 

 that Black Point is at the mouth of IVtaluma Creek, and hence is perhaps not rep- 

 resentative of the conditions in general along the west shore of -San Pablo Bay. 



Sumner, ct al. (1914) reports average salinities in the lower portion of San Pablo 

 Bay to be between 24 and 25 parts per 1000. His figures are based, however, on mid- 

 channel samples, in years of less than normal rainfall. It is probable that the in- 

 shore waters of San Pablo Bay during years of normal rainfall and ri\er discharge 

 are for a considerable period below 16 parts per 1000, and, especialh- in the ni-ighborhood 

 of the mouths of small streams, they must fretiuently drop below the rapidly lethal 

 salinity of 6.5 parts per 1000. It seems unlikeK', therefore, that Limnoria will establish 

 itself as an organism of economic importance in San Pablo Bay. 



Breeding H.\hi ts ok Limnoria in S.\n Fr.\ncisco B.w 



Limnoria is dioecious, but the sexes do not show \ery decided secondary sexual 

 characters. The most noticeable external difference between the male and female is 

 in size, the female being the larger of the two. 



When the eggs are extruded from the bo(l\ through the oxiducts they pass into 

 the brood cavity, the lamellae of which become cemented together. The eggs undergo 

 development in this cavity. They are bathed by a fiuid which is probably secreted by 

 the brood lamellae. When the extruded eggs fill up this chamber, the enlarged brood 

 pouch renders the detection of gra\id females eas>'. Ventral views of Limnoria with 

 eggs in the brood pouch are shown in figs. 124 and 125. The number of eggs carried 

 by a gra\-id female varies from 1 to 17; 17 is exceptional; 15 is fairly frecjuent. The 

 average for twenty-one gravid females examined b>- us was 9.5. Coker (1923) in his 

 work on the breeding habits of Limnoria at Beaufort, N. C, found the average in 

 May to be between 4.2 and 6.6. Twelve was the maximum number obser\-ed at 

 Beaufort. 



In San Francisco Bay it has been found that graxid females occur during the 

 entire vear. There appears to be no season in which breeding entirely ceases, even 

 though winter temperatures in the l>ay may fall as low as 8° C. Seasonal x'ariation 

 in the number of lar\ae in the brood pouch has not been studied by us. 



At Beaufort it has been found that there is a definite breeding season and that 

 it is correlated with the temperature. The data of Coker indicate that 14° C. is a 

 critical temperature above which breeding occurs, and below which it ceases. The 

 seasonal variation in temperature produces therefore an annual breeding season at 

 that locality. It is to be noted, however, that the range in temperature at Beaufort 

 is more than twice that in San Francisco Bay. According to Coker the range is 20.3° C. 

 at Beaufort. In San Francisco Bay the range of temperature varies considerably 

 w'ith the locality; the greatest range is at Carquinez Strait, where it is 12.65° C; while 

 the minimum range is at Golden Gate where it is 4.92° C. The conditions at both 

 these localities are extreme and the annual range in the greater part of the bay is 

 approximately 8.5° C. 



Although the temperature in San Francisco Bay is much of the time below 14° C, 



