BULLETIN OF THE UNITED STATES FISH COMMISSION. 21 



1 and 4, is laid, upon which a layer of fine, clean sand, S, is spread even- 

 ly from one end of the trunk to the other. The end board e, extending 

 half way up at the outer end of the box, runs up past the level of the 

 wire and cloth to confine the sand at that extremity, as shown in Fig. 2, 

 while the sand is confined by the board i at the other end of the trunk 

 next the pond, as shown in Fig. 3. The wire cloth and bars b h con- 

 stitute the support for the sand as it lies upon the gunny cloth, which 

 is supported in turn by the wire cloth or screen W. This is essentially 

 the construction of the filtering apparatus in which the layer of sand, 

 S, is at all times accessible, so that it can be removed if it becomes clogged 

 with ooze carried in by successive tides under the gate G, Figs. 1, 2, 

 and 4. This layer of sand can also be increased or diminished in thick- 

 ness so as to strain the inflowing and outflowing water more or less ef- 

 fectually, as may be desired, or in order to more or less efl'ectually pre- 

 vent the escape of any eggs or embryos of oysters which may be develop- 

 ing within the pond and wafted to and fro by the ebbing and flowing cur- 

 rents which are carried in and out of the pond through the diaphragm 

 by tidal action. The gunny cloth C, Fig. 4, may possibly be replaced by, 

 first, a layer of coarse gravel, then a layer of finer gravel superimposed 

 upon that, which would prevent the fine sand from sifting through the 

 sujiporting wire screen W- Gravel would be more durable than gunny 

 cloth or sacking, which, like all other textile fabrics, will rot if immersed 

 in salt water for a few weeks. In practice, however, a mode of getting 

 over all such difficulties would soon be devised ; a coarse sacking to be 

 used for the i)urpose might be saturated with a drying oil or with tar dilu- 

 ted with oil of turpentine, which when dry would act as a preservative 

 of the material, but not cause it to become impervious. 



In the old style of diaphragm used in the experiment at Stockton, it 

 was difficult to renew or clean the sand, inasmuch as the apparatus con- 

 sisted essentially of a box open at the top, and as wide and as high as 

 the trench connecting the pond with the open water. Its depth was 

 three feet, its width two feet, and its total thickness about four inches. 

 The sides forming its greatest depth and width were jjerforated with' 

 numerous auger holes. On the inside, this narrow, deep box, of the 

 above dimensions, was lined with gunny sacking and the intervening 

 space filled with fine sand. This diaphragm was placed vertically in the 

 trench, and it will be readily understood that the filtering surface was 

 limited by the depth of the water in the ditch, while its free action was 

 also to some extent impeded by the small amount of ingress and egress 

 offered to the ebbing and flowing tide, in passing in and out of the pond 

 through the auger holes, in the sides of the box, on either side of the ver- 

 tical stratum of sand. It will also be readily understood that it would be 

 impossible to remove the sand from the box to repair or renew the filter 

 without destroying its effectiveness for the time being. 



The diaphragm, of which I am about to describe the working, obviates 

 all of these objections, while it is possible to augment the extent of the 



