SEA-FISHERIES LABORATORY. 931 
structed that a piece of the silk to be tested can be fixed 
over it in such a way as to prevent any water leaving, 
except through this tissue. 
The tube conducting the water to this cylinder 
opened into one side near the closed end. Another tube 
opening also into the side, but nearer the open end, 
placed it in communication with a mercury manometer, 
by means of which the pressure of water in the cylinder 
(representing the pressure under which filtration was 
taking place), could be recorded. ‘The whole cylinder 
with its connections to water supply and manometer, was 
immersed completely in water in a rectangular reservoir, 
for reasons stated above. ‘This reservoir was supported 
on a table, the top of which had the form of a trough, 
so that any water flowing over the edges of the reservoir 
was collected by it. A short tube allowed this water to 
run out of the table trough so that it could be collected 
-at any time in a flask kept under the tube, but which had 
its mouth guarded by an inverted evaporating basin 
placed over it. 
Precautions must be taken to allow of easy removal 
-of any air which generally collects in various parts of 
the apparatus during the experiment, and disturbs the 
results, and the opening closed by the gauze to be tested 
should be so constructed that the end can be taken off and 
the gauze cleaned with a sponge after each trial. 
The procedure is as follows. Everything being in 
place, the water is turned on, the reservoir being already 
filled with the same up to the edges. The water passes 
through the cleaning filter, the large flask and the regu- 
lating cock, to the brass cylinder which it leaves by way 
of the tissue to be tested, passing through this into the 
reservoir. Since this is already full, a certain quantity 
overflows equal to the volume entering. The pressure is 
Q 
