TRANSACTIONS OF THE PHARMACEUTICAL SOCIETY. 
367 
less hydrodynamic than it should be, owing to the resistance of the medium, a 
resistance that is more evident with short columns than with long. 
To. continue these experiments with other filtering media would be of little 
additional interest or value. JSTo two different filtering media offer equal re¬ 
sistance, nor do any two pieces of the same medium give exactly equal results; 
indeed, any given piece of a medium alters to some extent in porosity even 
during the performance of a few experiments : add to this the fact that in filtra¬ 
tion as practised by analysts and pharmaceutists the medium is constantly receiving 
accessions of deposited matter until it chokes and ceases to be a filter any longer, 
and it will be evident that such experiments would be of no practical value for 
reference, and would only tell us what those I have described teach us already. 
By way of connecting the experiments on flannel and cloth with the next 
series in which columns of liquid of increasing length up to forty-two feet were 
used, I may state that the rate of flow through a medium made of a particular 
kind of coke was taken. The coke was that known as the moulded carbon filter 
of Messrs. Atkins and Son, of Fleet Street. I believe it is manufactured by 
igniting, under pressure, a mixture of charcoal or coke and pitch. This ma¬ 
terial I found to be more constant in its operation during a series of experiments 
than those already alluded to. The form known as the “siphon filter” was 
first used. It is simply a block or disk of the carbon, in which a hole is bored 
from any point in the circumference to the centre, and into which hole a short 
glass tube is inserted. An india-rubber tube of any convenient length is fixed 
on the glass tube, and now, on aspirating, filtered water can be drawn into the 
mouth from any source, or a tumbler filled by arranging. the tube as a siphon, 
that is, by lowering the block to the bottom of a jug of water and letting the 
tube hang over the spout of the jug and dip into a tumbler placed below the 
level of the jug. In the experiments the level of the water was of course kept 
constant. The orifice of outflow being raised or lowered, the following was the 
rate of flow in two minutes at different distances from the surface of the 
water :—The rate through the tube without the carbon is also given for com¬ 
parison, and to show that the whole resistance to the flow must not be ascribed 
to the carbon. 
With carbon. Without carbon. 
At 4 inches ... 3 ounces ... 9 ounces. 
3 ,, . . . o ,, ... 16 
12 „ . . . 6f „ ... 22 
16 „ . . . 8 | „ . . . 27 
Here, again, it is evident that the flow is chiefly hydrodynamic. 
In order to ascertain the extent to which the rate of flow is increased, when 
the column of liquid beneath the filtering medium is prolonged to great dis¬ 
tances (30 to 40 feet), some special precautions were necessary. To ensure 
constancy of action and strength of material, as well as to obtain the greatest 
practical proportion of aerostatic pressure, a carbon block was chosen of closer 
texture and larger size. 
The flow from the block itself, without any column, was not a continuous 
current, but simply a succession of drops. The block fairly represented, there¬ 
fore, the state of a practical (tincture, etc.) filter in a clogged condition, a 
point at which an operator becomes anxious, for economical reasons, to in¬ 
crease the flow. The nature of the action of such a clogged filter cannot be ascer¬ 
tained by direct experiment upon it, because the medium would be getting 
thicker and denser during even one series of experiments. But the nature 
of the action of a carbon block filter, which is giving the same amount of re¬ 
sistance to the passage of liquid that the clogged filter gives, must fairly repre¬ 
sent the nature of the action of the clogged filter itself. It w r as a cylinder, 
six inches long and the same in diameter (see the accompanying figure), and 
hence afforded a filtering surface of more than 150 square inches. The flow 
