174 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
bottom to surface. If it is possible to obtain by the pump such a column of water, 
then the pump may very well replace the net so far as this part of the process is 
concerned. I do not say that this is not possible, but we should not assume that the 
w'ater drawn in by a pump through the submerged end of a hose, which is being 
slowly moved from top to bottom, or vice versa, is a vertical column of water. 
Before the pump can replace the Hensen net there must be sufficient evidence that 
this is so, and such evidence is not yet forthcoming. 
Having obtained the water by use of the pump, it is necessary to separate the 
plankton from it. To accomplish this, the second process into which we have analyzed 
the Hensen procedure, various means have been proposed. Frenzel, and at first 
Kofoid, 1 made use of the Hensen net to strain the water pumped. In order to 
avoid the loss of plankton due to the permeability of the net to small organisms, 
Kofoid later tried various other methods of separating the plankton from the water. 
These were the sand filter, the filter paper, the centrifuge, and the Berkefeld 
filter. By each of these methods a greater number of plankton organisms is retained 
than by the Hensen net. (Nothing is said of volumes.) In some cases as much as 98 
per cent of the total number of organisms present is retained. By none of these 
methods is it possible to obtain the plankton from a large volume of water in a short 
time, and each has besides other disadvantages which are enumerated by Kofoid. In 
the case of the Berkefeld filter, which was found to be the most efficient method, it 
was necessary to remove the catch from the surface of the filter with a “stiff brush.” 
The surface of the filter, which is composed of infusorial earth, was thereby disinte- 
grated aud the plankton contaminated by the fragments. It is to be hoped that the 
disintegration is confined to the filter. The large form of the Berkefeld filter (army 
filter) filters about 2 liters of water per minute. This is a very slow rate of filtration 
if one has to deal, as is sometimes desirable in plankton work, with a column of water 
several hundred feet long and perhaps 10 inches in diameter. 
The methods which it has been proposed to substitute for the Hensen method are 
thus seen to be deficient in two ways. For obtaining the water the pumping method 
is (so far as yet shown) defective in that the source of the water pumped is uncertain. 
It is not known that the pump can be made to deliver with accuracy the contents of 
a vertical column of water. For filtering the water the methods proposed, although 
they remove the plankton organisms more perfectly than the Hensen net, are yet 
inferior to it in that they are incapable of handling large volumes of water. Is it 
possible to so modify the Hensen method or to so combine it with other methods as 
to correct its errors and at the same time retain its good points'? Its errors are the 
variation in net coefficient, due to clogging and shrinkage, aud the permeability of 
the net for small plankton organisihs. Its advantages are that it filters a representa- 
tive vertical column of water, and that it filters rapidly very large volumes of water. 
Now, if it is possible to measure the volume of water that passes through the net at 
each haul the difficulties of clogging, shrinkage, and net coefficient at once vanish. I 
have not made any attempts in this direction, but I see no reason why a small current 
meter can not be placed within the opening of the plankton net, so as to register the 
rate of the current of water passing through the opening during each haul. If this 
rate were known the volume of water passing through the net could be calculated, 
1 Bulletin Illinois State Laboratory of Natural History, vol. v, article i. 
