SPAWNING AND SETTING OF OLYMPIA OYSTERS 
467 
original activity early in the season. In other years it was not apparent that there 
was any definite renewed spawning activity. 
In his work on 0. edulis, Moebius (1883) said that as many as 20.6 percent of the 
adults bore larvae at once, and estimated from frequent observations, that at least 
44 percent of the oysters produced broods during the season. Stafford's (1914) 
results indicate a comparable proportion bearing larvae at the same time, though he 
did not estimate the proportion of the population which produced broods. The data 
obtained during the sampling described above provide a means of estimating with 
some degree of certainty the total spawning activity throughout several seasons. 
Because of the frequency of the samples, it is possible to analyze the rate of 
development of the larvae, as is described below, and to detect the relative number of 
oysters bearing newly spawned eggs. From this one may reach an estimate of the 
total number of adults which bear broods. In 1932, as shown in table 14 and figure 
15, the oysters spawned prolifically in Oyster Bay. At one time as many as 55 out 
of 100 carried broods. By referring each age group back to the date of spawning 
and then determining the total percentage of individuals spawning during the 
season it was possible to demonstrate that at least 1.5 broods per oyster were produced. 
That is, apparently all of the individuals bore one brood and at least half were gravid 
for the second time. 
In 1933, howe/ver, at the same place only about 75 percent of the individuals 
became gravid (fig. 16). During most years it appears that approximately 100 percent 
of the adults bear larvae, but only in 1932 was much greater spawning activity noted. 
Judging from all data available, it is probable that the variation in number of broods 
produced during different seasons is between about 75 and 150 per 100 adult oysters. 
In addition, the specimens would also spawn as males, as described by Coe (1931a, b). 
A source of error in such estimates is the possibility of abortions of young embryos 
which would, therefore, not be counted. It would be difficult to determine how 
frequently abortion of a brood occurs, but it is clear that it sometimes happens. 
DEVELOPMENT OF LARVAE 
Although Stafford (1914) reached the conclusion that the larvae develop normally 
for a period of 16% days within the maternal brood chamber it is probable that the 
method he emplojmd was unsatisfactory. He would periodically pry the valves of a 
gravid specimen partly open and take a sample of the larvae. Such handling of the 
specimen might readily result in a disturbance of normal function and interfere with 
larval development. A strictly biological method, therefore, would not appear to be 
adequate to solve the problems related to rate of development under natural conditions. 
It was necessary to use a system of sampling the oyster population and determin- 
ing at frequent intervals the stage of development of larvae in the various broods. 
On each of two typical grounds in Oyster Bay 100 adults were opened 3 times weekly. 
Larvae from gravid specimens were separately preserved in vials for later laboratory 
examination to determine their size or stage of development. By talcing samples at 
frequent intervals throughout the season it was possible to organize the results so 
that gravid specimens bearing broods of the same stage of development could be 
grouped and followed through the various stages. If on 1 day 10 percent of the oysters 
bore newly spawned eggs, 2 days later about the same number would be found with 
embryos of a certain stage. In subsequent samples the group would continue to recur 
until the larvae reached the size at which they are discharged. A single brood was 
found to consist of larvae of approximately the same stage, within relatively narrow 
limits. In no case were larvae of widely different stages found in a brood, and it may 
