around the pipe and fastened over the baffle opening. The discharge end 

 of the flmne was fitted with a movable tailgate to control the water depth. 

 The slope of the flume, adjustable by means of an automobile jack, remained 

 at 0.8'? throughout the experiment. 



The sediment comprising the miniature flat was collected from the 

 upper 3 inches of a tidal flat in Hales Cove, Plum Island Sound, Mass. 

 The uppermost layer in Hales Cove consisted of silt about one-half inch in 

 thickness. An attempt was made to duplicate the Hales Cove flat by placing 

 the mud in the recess while full of water and allowing the particles to 

 settle overnight. As shown in the analysis below, the proportions of silt 

 and fine sand were high, which, combined with considerable organic detritus, 

 gave the sediment cohesiveness . The amount of organic material was not 

 measured in this particular case, but samples collected from the same area 

 contained 2,9^ organic material hj weight. 



Particle size Classification 



(microns) (Sverdrup, et al. 191x2.) 



under 50 silt 



50-ino very fine sand 



inO-25o fine sand 



250-UnO medium sand 



over iiOO medium to coarse 



sand and larger 



Three current velocities were empirically selected, based on the 

 effects of the lowest and the highest velocities on trial groups of clams, 

 and one which appeared to be midwajr between^ The lowest was barely able to 

 set in motion a few clams of a trial group, and the highest moved them all 

 quite easily. The three velocities used were calculated to be 13, l8, and 

 25 cm. per second at the center of the flat (50 cm/sec « about 1 knot) , 

 Ov/ing to the flume's slope and the resulting variation in water depth, 

 velocities were slightly higher at the upstream end than at the downstream 

 end. Hjulstrora (1939) states that water generally flows in a turbulent way 

 under conditions found in nature.. This is doubtless true of the three cur- 

 rent velocities selected, although turbulence was made visible only in 

 the 25 cm. per second current when fluorescein dye was injected into the 

 water for this purpose. 



Mean current velocities were calculated by dividing the discharge by 

 the cross-sectional area of the water in the flume (Menard and Boucot 

 195lj Krumbein 19U2) . Admittedly, this method does not determine the 

 velocity at the soil-water interface where friction retards a layer of 

 water less than 1 millimeter in thickness (Menard and Boucot 1951) , but 

 it seems safe to assume that the clams extended above this layer into the 

 turbulent flow above o The only exceptions might be a few of the smallest 

 clams (2mm. long) which may have come to rest in minute depressions in the 

 flat. In the present experiment, the mean current velocity is used. 



