346 
BULLETIN OF THE BUREAU OF FISHERIES 
fattening. It appears possible that salinity may be one of the important factors influ- 
encing the feeding of oysters. 
The gills of oysters, by means of cilia, pump a stream of water from which food 
particles are filtered and passed along definite grooves to the labial palps, which convey 
them to the mouth (Nelson, 1923b). The rate at which water is pumped was used by 
Galtsoff (1928) as a criterion of the rate of feeding under different conditions of 
temperature. Nelson (1921, 1923a) kept records of the opening and closing of oysters 
immersed in the open bay along with those of temperature, salinity, turbidity, etc. 
He concluded that the oyster feeds most rapidly on the flood tide even when the 
density is approximately the same as during ebb tide. He concluded also that for 0. 
virginica a density of 1.008 (10.42 parts per mille) is the lowest at which feeding will 
go on, and that the minimum density required varies in proportion to that of the 
water in which the oysters have been grown. His method, however, was to study the 
opening or closing of oysters under different conditions, judging an open specimen as 
actively feeding. But, as will be shown below, an oyster may be open without feeding, 
though clearly when it is closed feeding is impossible. 
The mechanism of physiological adaptation to changes in salinity is not the 
question primarily discussed in this paper. This study deals with the pumping 
activity of the gills as influenced by salinity, although it is realized that final ex- 
planation of the results is a matter of cellular physiology. 
MATERIAL AND METHODS 
Because of their relatively large size, specimens of Ostrea gigas are most favorable 
material for experimental study. Tins species is imported from Japan as small 
seeds and grown in various waters of the Pacific coast. In portions of Puget Sound 
propagation occurs, and the resulting well-shaped, large oysters were used in these 
tests. 
The method employed has already been described (Hopkins, 1933). A simple 
lever, resting with the least possible weight upon the upper (right) valve of the 
specimen, served to record on a kymograph paper the position and movements of 
the valves, as affected by activity of the adductor muscle. A paper cone, thoroughly 
waterproofed with a solution of celluloid in acetone, attached to a system of levers 
made of lightweight straw, was placed so that all water pumped by the gills struck 
the cone. The recording tip of the straw was made of a sliver of cellophane, and 
this made its record on the kymograph paper directly below that of the lever 
recording shell movements. Fixed levers continously recorded the zero positions— 
the closed position of the shell and the level of cessation of pumping by the gills. 
Since the salinity of the water had to be maintained at a constant level, it was 
necessary to employ a method for continuous aeration and circulation. For this 
purpose small centrifugal pumps, made of celluloid (Hopkins, 1934) were used. 
Two of these pumps were used to circulate and aerate the water, without causing any 
strong currents which might disturb either the oyster or the mechanism for recording 
the relative rate at which the gills pump. An extra aquarium was interposed between 
the experimental chamber and the pumps for the control of temperature. In this 
aquarium either an electric-light bulb or an immersion heater was so controlled that 
the temperature in the experimental chamber was maintained during all tests at 
between 17° and 19° C. This temperature level was chosen because it had been 
previously determined (Hopkins, 1933, 1935) to be most favorable for stability of 
