212 
BULIvETiN OF THE BUREAU OF FISHERIES. 
water, as will be shown below, no oxygen except the very small amount absorbed by 
their shells. 
It appeared early in the work that constancy of results was exceedingly difficult to 
obtain. This was due to a variety of causes, chief of which was the great variability in 
the openness of the shells, for not only could the oyster or clam entirely or at least 
obviously close so that the experiment had to be terminated, but it could partially and 
unnoticeably close or indeed fail to open wide even from the beginning of the experiment. 
For the oyster, at least, the author succeeded in demonstrating this by graphic records. 
One shell of an oyster was connected by tying a string to the projections with a lightly 
balanced lever recording on a slow kymograph. As soon as the oyster in a water bath at 
a temperature between i8° and 20° C. had opened somewhat, the kymograph was started 
and the temperature of the bath raised at the rate of i ° C. in about eight minutes. The 
oysters remained fairly well open with brief periods of partial closure as the temperature 
increased to about 22° C. At that point there invariably appeared in the three individ- 
uals observed periods of maximum openness lasting as long as no disturbing factor inter- 
vened. Sufficient stimulus for partial closure, however, was likely to occur frequently. 
A light tap on the table or water bath, a heavy step in the room, the slamming of a door 
in a neighboring part of the building, or, indeed, any slight jar was surely registered by 
some movement of the shell. As the temperature increased up to about 26° or 27° C., 
the effect of these stimuli was much less marked. The oysters then maintained their 
maximum openness very persistently. Between 27° and 30° a tendency to very slow 
and incomplete opening after closure was noticeable, indeed no maximum openness was 
seen. At about 30° or 31° C., the oysters closed tightly, even if no mechanical stimulus 
was given. 
As the k)rmograph method served to detect movements of the shell not noticeable 
to unaided observation and also slight openings of the shell not otherwise visible, these 
experiments were an aid to planning and interpreting measurements of oxygen utiliza- 
tion. They showed that below 19° C. and above 26° C. observations on the opened 
oyster were impossible, that temperature must be maintained constant throughout the 
experiment, and mechanical disturbances must be avoided as far as possible. A further 
source of difficulty had also to be overcome. When the oyster excreted it closed vio- 
lently, to drive the fecal matter out of the shell. If the position of the animal rendered 
complete excretion difficult, closures were frequently repeated, and sometimes the oyster 
shut up tightly. It was necessary, therefore, to lay the oyster in the desiccator tipped 
so that the more concave side of the shell, where excretion occurs, would be lower than 
the other. 
In spite of all precautions, however, perfectly consistent results could not always 
be obtained. Under the same or comparable conditions of temperature, oxygen content 
of the water, and physical conditions, the same individual would sometimes in different 
experiments give results disagreeing beyond the limits of the calculated, probable, experi- 
mental error. Various observations make it seem likely that the nutritive condition 
of the individual could account for some, at least, of these discrepancies. Thus, after 
