SPAWNING AND SETTING OF OLYMPIA OYSTERS 
475 
The graph (fig. 21) suggests that the effect of angle of surface on the number of 
spat caught is purely mechanical and not due to any definitely biological reactions of 
the larvae. Hori (1933) stated that swimming larvae commonly are in an inverted 
position, with the velum uppermost. This has been observed also by the writer and 
is well illustrated in Prytherch’s (1934) work on larvae of Ostrea virginica. The velum 
projects through the valves as a flattened, ciliated swimming organ, while, the heavier 
shell hangs downward. The foot with which the larvae must adhere is beside the 
velum and projects more or less upward, although it is extensible in all directions. 
It is most likely that the swimming larva, as it comes into contact with a surface 
from below, is able to hold on with the foot, while on coming down upon a surface it 
is the hinge portion of the shell that touches. In this manner, as the angle of the 
surface departs more and more from the under horizontal there is constantly less 
chance of the foot touching. This interpretation, in effect, is that- the observed 
results are due to accidental contact of the foot with the surface as the larvae is swim 
ming and being washed about by tidal currents. Prytherch’s descriptions of setting, 
as directly observed with the microscope, suggest that larvae of 0. virginica may react 
differently in this respect. 
If the above-described interpretation is correct it would be expected that in 
places where the water is highly turbulent the larvae would frequently be turned 
over so that they might also catch on upper surfaces. It has frequently been observed 
on oyster grounds near Olympia, in places where the water flows over dikes, that the 
rocks and shells close to the dikes bear spat also on upper surfaces. 
METHOD OF DETERMINING FREQUENCY OF SETTING 
It is of considerable importance to know the duration of the setting season and 
the times when maxima are reached. For this purpose it was necessary to plant 
cultch periodically during the entire season. The system finally adopted, after the 
first year, was to plant a wire bag of shells on each experimental ground and allow 
it to remain for 7 days. It was then removed and brought to the laboratory. As one 
bag was taken from the ground a new one was planted and allowed to remain for the 
following 7 days. It was found necessary, however, to carry on at each place two 
such 7-day series so that one overlapped the other. In one series, for example, the 
bags would be in the water from Monday until the following Monday; in the other 
series, from Friday until the next Friday. A clean lot of shells was therefore put into 
the water every 3 or 4 days. 
After being brought from the grounds the shells were allowed to dry and counts 
made of the number of spat caught. Bags were made of 1-inch mesh galvanized wire 
netting and were about 30 inches long by about 8 inches in diameter. Each held 
something over tliree-quarters of a bushel of Japanese oyster shells. These shells 
were preferable because of their large size, generally 4 to 6 inches long, and the white 
color of the inside surfaces. In the bags the shells were held at all possible angles, 
eliminating any error that might be traceable to the angle of the surfaces. 
Counts were made only of the spat on the inside surfaces, because of their color 
and smoothness and because the outside surfaces are too rough, and often lamellate, 
so that all spat are not readily seen. This is not difficult to understand when it is 
realized that the shells were in the water only 7 days and the oldest spat a millimeter 
or less in diameter. Two bags of shells were left in the water for a month and a 
half to allow the spat to grow to a large enough size to permit accurate counts of the 
number on inside and outside surfaces. The results are summarized in table 20. In 
one case 33 percent were on inside surfaces, and in the other, 36.1 percent. The 
