whereas in both 1951 and 1952 larvae first appeared in the latter part of 

 April. Venus larvae first appeared in the week of -June 11 in 1950, but late 

 in May in 1951 and 1952. In addition, Venus spawning continued at a rela- 

 tively high level until nearly mid-September in 1950, whereas in 1951 and 

 1952 practically all of the spawning was completed by mid- July. The answer 

 appears to lie in the water temperature in 1950 as compared with 1951 and 

 1952. Figure 2 shows that in 1950 water temperature was consistently lower 

 than in either 1951 or 1952. This condition may contain an explanation of the 

 late beginning of spawning of both Mya and Venus in 1950. It could also be 

 the cause of the continued relatively intense spawning cf Venus throughout 

 the summer of 1950,, assuming that high w ater temperatures lead to the rapid 

 spawning out of a population, while relatively low temperatures, perhaps not 

 much above the minimum necessary to stimulate spawning will result in pro- 

 longed spawning at a high level. 



No attempt has been made to determine the temperature necessary to 

 stimulate spawning in either Mya or Venus . In cases where the animals 

 studied are intertidal as well as subtidal, it may well be, as Belding (1931) 

 suggested, that individuals living intertidally or in shallow water reach 

 their spawning temperature considerably earlier than those living in deeper 

 water. This point is emphasized so that the water temperatures given in this 

 paper at the time larvae were first taken each year will not be construed as 

 spawning temperatures for Mya and Venus in Narragansett Bay. However, since 

 the temperatures presented are a product of the same heat exchanges as deter- 

 mine the temperature of an intertidal flat or a shallow cove, they may s erve 

 to show the thermal trends anywhere in the area during the spawning season, 

 and explain differences in spawning patterns of the same bivalve from year to 

 year. 



Influence of early warming of exposed flats and shallow water on the 

 spawning pattern of a bivalve population may be indicated by a comparison 

 of Venus spawning in Wickford Harbor and in Greenwich Bay in 195l or 1952. 

 In Wickford Harbor in 195l and 1952, Venus spawning (figs. 6 and 8) was 

 finished for all practical purposes by mid-July. Only a relatively few 

 Venus larvae of any age were seen after this date. On the other hand, in 

 Greenwich Bay during the same period, Venus spawning (figures 9, 10, 12, 

 and 13) was much more generally distributed over the summer months. A com- 

 parison of the water-temperature records (figures 2 and 11) from Wickford 

 Harbor and Greenwich Bay shows no obvious differences that might account for 

 the difference in spawning pattern. When the average depths of water and the 

 percentages of exposed flats in the two areas are compared, a decided differ- 

 ence is apparent. Wickford Harbor has an average depth at mean low water of 

 2 to 3 feet, while Greenwich Bay averages about 10 feet. Exposed flats are 

 from one-tenth to one-fifth of the total area of Wickford Harbor at low tide 

 while exposed lfats are a negligible part of the total area of Greenwich Bay. 

 If clams in shallow water and exposed flats spawn earlier at the same temper- 

 ature than those in deeper water, the spawning early in the season will be 

 more intense in Wickford Harbor than in Greenwich Bay, and the data suggest 

 that this actually occurred. 



26 



