360 
TRANSACTIONS OP THE PHARMACEUTICAL SOCIETY. 
though an important subject, is not at the present moment under discussion, 
their mode of action only'is what we are now considering. 
(e) Filtration in which Gravitation is aided by Hydrodynamic-force, or by an 
apparent combination of Hydrodynamic-force with Hydrostatic-pressure , or 
with Aerostatic- (Atmospheric-) pressure. —Filters have been proposed in which 
the rate of flow has been augmented by the use of a column of liquid above 
or below the medium. To understand the nature of the action in these in¬ 
struments it will be necessary to trace that action to its sources. When 
water is poured into a long glass or other tube, uniform diameter, closed at 
one end, say by a flat glass plate by means of proper cement, the natural 
gravitation of the water will exert a pressure on the bottom of the tube. 
The amount of this pressure will clearly depend upon the amount of water 
in the tube. If there is one pound of water in the tube it will be a pres¬ 
sure of one pound, if two pounds, two pounds, and so on. If the first pound 
of water form a column a foot long, then the addition of every foot to the 
column will increase the pressure by one pound. The pressure will increase 
in proportion to the simple length of the column. Such a pressure is pro¬ 
perly called “hydrostatic,” because produced by u water in a state of rest.’’ 
But now let the bottom of the tube be a filtering medium, instead of an impervious 
plate—let it be flannel or common black coat cloth; and, to keep up analogy 
with the previous experiment, let the water as it runs through the medium be 
so replaced by more as to maintain the column at one, or two, or three, or four 
feet in length. It will then be found that the amount of water which will run 
through the filtering medium in a given time when the column is a foot long is 
not doubled when the column is two feet long. The amount is certainly in¬ 
creased when the height of the column is doubled, but not to the extent it would 
be increased if the increase were proportionate to the hydrostatic pressure. 
The actual extent to which it is increased depends entirely upon the amount of 
resistance which the filtering medium offers to the passage of the liquid. I find 
that if the resistance is so great that the liquid passes more or less in the state 
of drops, then the amount obtained from a four-feet column certainly does ap¬ 
proach to nearly four times the amount obtained from a one-foot column of 
liquid. In one experiment, in which four ounces of water filtered through 
some closely packed sponge, when the column of water above the sponge was 
one foot high, fifteen ounces passed when that column was lengthened to four feet. 
This was a close approximation to an effect proportionate to the hydrostatic pres¬ 
sure inside the vessel. It is conceivable that when filtration in such a tube be¬ 
comes so slow that a liquid only passes in drops, then the pressure of the whole 
column of liquid, having, so to speak, time to accumulate, exerts itself on the 
particles of liquid near the medium with nearly its full hydrostatic power; and 
that at that moment any particle actually at the mouth of a pore of the filter¬ 
ing medium, being unequally pressed on all sides, being forced down by hydro¬ 
static pressure, and being opposed from below only by the capillary attraction of 
the pore, is forced in the direction of least resistance—in other words, shot through 
the pore with a velocity nearly proportionate to the hydrostatic pressure above, 
minus the capillary resistance of the pore. Let me not be misunderstood here. 
I mean to say, that inasmuch as a filter in an inactive state with a certain co¬ 
lumn of liquid above the medium becomes active, that is, in a state of flow 
when the length of the column is increased, it is obvious that the flow in the 
latter case is set up by hydrostatic pressure. And the rate of flow, when the 
liquid is passing in drops, is nearly in proportion to the amount of pressure 
exerted upon the filtering medium. But I do not think the rate of flow in¬ 
creases in an amount parallel with that particular portion of the total pressure 
which produces the flow through the medium. I believe it increases in an ex¬ 
tent proportionate to the square root of that particular amount of the hydro- 
