with a bung pierced by two tubes. One is long and runs down to the bottom of the flask, and the other ends flush with the inside of the bung and is bent into an S-shape outside. A U-shaped piece of glass rod fits into a length of rubber tubing attached to each tube. When the bottle is at the required depth, a messenger is sent down to release a strong spring which pulls the glass rod out of the two tubes. Water runs down into the flask through the long tube, driving air out of the other tube until the flask is completely filled. A bubble of air remains in the bent tube so that no mixture can take place between the sample in the flask and the surrounding water during haulage of the whole apparatus to the surface. In practice, a number of flasks, each with its stopper and tubes, are sterilized in the laboratory and then a series of samples for bacteriological examination can be taken at different depths or at different places during the same outing.
Water may be obtained from any depth by lowering a tube and sucking. In the early days of the Freshwater Biological Association, when lack of money placed a premium on ingenuity, Mortimer used a bicycle pump with the washer reversed to obtain samples. If two bottles are connected in series, with the larger nearer the pump, the smaller and the tube will have been sufficiently washed by the time the larger is full. Water can also be raised by a stream of air bubbles emitted from a small tube inside, and extending almost to the lower end of, a larger one.
Plankton is commonly caught by means of a conical net made of material woven in such a way that the holes retain their size. A mesh of 60 meshes to the inch is generally used for animals, one of 180 meshes to the inch for algae. The efficiency of a net falls as the catch blocks the pores and for quantitative work the amount of water that has passed through the mouth must be measured by means of a propeller attached to a recorder. Many methods of catching plankton have been tried, particularly at the station at Pallanza, and the quest continues. One difficulty is that some of the animals swim away from an object they see coming through the water, or away from the pull of a current caused by suction into a pipe.
The easiest medium to sample is the mud on the bottom of a lake, though each sample must be subjected to a tedious process of sieving before the animals can be isolated. Often a simple tube will secure enough animals. If they are scarce, a larger sample may be obtained with a Birge-Ekman grab, which is a metal box open at the bottom and provided with two hinged lids at the top. Two jaws to close the bottom are held along the sides against the pull of strong springs. Going down, the apparatus passes through the water with little disturbance. This is important, for if there is obstruction the apparatus will not pass through the water, but push it aside, and it will also push aside the top layers of the mud if these are fine and fluid. The lids fall when the box sinks into the mud and comes to rest. A messenger trips the bridle that holds the jaws up and the springs then pull them together to close the bottom of the box.
Stones and vegetation are less easy to sample quantitatively. Several workers have found that the number of animals caught in a given time or in a given number of sweeps of a net indicates, sometimes with unexpected accuracy, relative numbers in different places. Numbers per unit area can be calculated if samples with a quantitative sampler are taken in the same part of the lake at the same time. The Danish workers have used a square box open top and bottom to sample stony substrata near lake margins. It is placed over the bottom, stones are removed, and the water inside is baled out and poured through a net. This method cannot be used in running water because the box deflects the current downwards and causes it to scour the area that is to be sampled.
In Windermere H. P. Moon used a square frame on which he could pile stones to represent an area of natural substratum before lowering it onto the bed of the lake and leaving it until it had been colonized. The frame is one-third or one-half of a square metre in area and underneath it is covered with fine gauze to prevent the loss of animals while the frame is being hauled up. Stout wire-netting beneath the gauze adds additional support for the stones.
The Surber sampler is used by some workers to sample the stony substratum of streams and rivers. It consists of two frames, generally about one tenth of a square metre in area. These fold into one plane for transport and open at right angles for use. The horizontal one is placed on the bottom and the vertical one supports a net. Stones are then removed from the bottom inside the horizontal frame, and brushed in the mouth of the net to dislodge animals clinging to them, after which the remaining small stones, gravel and debris are stirred with a stick until it is believed that all living material has been swept into the net. We have not found this a satisfactory instrument because the current is often so swift that when one stone is picked up the stones above it shift to fill the gap. If the current is slow many good swimmers probably swim out of the net, if they are ever carried into it. More satisfactory, though not by any means free of error, is a shovel of some kind which can be pushed into the substratum for a known distance. Designs have ranged from a shovel with high sides with a net at the back, to a cutting edge connected to the handle by two strips which also support the frame of the net. A strong coarse net arrests the stones and a long tapering fine one any animals that have let go. If the stones are tipped into a solution of high specific gravity, calcium chloride or magnesium sulphate are suitable, the animals float to the top.
Weeds in rivers trail downstream and may be severed with shears and caught in a large bag. Another method is to hold a box with sharp edges a known distance above a lid and then bring the two together enclosing and severing the weed in a known volume. Weeds in still water rise vertically, and a device that cuts each leaf or stem as it meets it is preferable to one that pushes them downwards and does not cut them until they are pressed against the bottom. One such instrument consists of two tubes, about 8 cm. across, fitting one within the other. A boss on the inner passes through a slit in the outer and holds it in position, allowing a small amount of rotation to and fro. As the tubes are lowered into a weed-bed, the outer tube is rotated and the vegetation is severed between the sharp teeth which have been cut in the lower end of both tubes. They pass across each other like the teeth of a haycutter.
Incidentally parallel samples with this instrument and a net have shown the latter to be unexpectedly selective. It collects an unduly high proportion of species that tend to flee and an unduly low proportion of those which, like leeches, tend to cling to the substratum.
Larvae of Chironomids and many Trichoptera cannot yet be named, and in order to find out what species are present it is necessary to trap the emerging adults. In still water a box open at the bottom may be floated in a frame. The top should be of some transparent plastic material to keep the rain out, but at least one side should be of gauze to prevent condensation. Dr. J. H. Mundie has devised various modifications for use in both still and running water. In a lake he used conical traps into the top of which a screw-top jar could be screwed. Entrance to it is through a cone which prevents the animals falling back into the water. The whole apparatus can be submerged, an advantage in a lake to which the general public has access. For use in streams he built a heavy trap that could be anchored to the bottom. Triangular in both plan and elevation it offered minimum resistance to the current, which tended to press it downwards. Three legs kept it raised off the bottom and the catch entered a screw-top jar as in the other model.
Much decomposition takes place in the top few centimetres of the mud, and substances diffuse from it into the water. A study of these processes, important to the general economy of the lake, requires a sample disturbed as little as possible. The Birge-Ekman grab does not bring up such a sample and for this purpose the Jenkin surface-mud-sampler was invented. It consists of a large glass tube about 6 cm. in diameter, held by a band about its middle on to a metal frame, standing on four spreading legs. The sampler sinks into the soft mud when lowered to the bottom, but without disturbing it, and then, a messenger sent down the wire having released a catch, two pairs of arms travel forward to place a cap on either end of the glass tube. The speed at which these caps move into position has to be very slow in order not to disturb the mud and water in the tube, and this is effected by means of a pressure chamber of the same kind as that used for preventing doors from slamming. When closed, the glass tube contains a sample of the top few inches of deposit together with the water above, and the caps at either end are held tightly in place by springs. At this point the whole apparatus is hauled
