3. Water-Bottles.

All the water-bottles used during the first cruise, in 1900, were provided with releasing propellers. During the subsequent cruises, messengers have always been
used, and not propellers

The Great Nansen Insulated Water-Bottle

Having had much experience with the Pettersson Insulated Water-Bottle, during the Fram Expedition 1893-96, Nansen had come to the conclusion that by improvements on the principle of this instrument, it might be possible to construct an insulated water-bottle which could be used for taking trustworthy temperatures, even at great depths. The result was the socalled Great Nansen Insulated Water-Bottle (made by Mr. ANDERSEN of Christiania ). Its chief advantages were:
(1) the use of the above-mentioned Nansen Deep-Sea Thermometers, fixed with perfect water-tightness in its lid.
(2) An unusually high capacity of insulation.
(3) absolute water-tightness when closed, as the lid was pressed down by excentric levers, multiplying the force of   the lead 6 times.
The instrument has been already mentioned in Nansen's preliminary report [1901, p. 131, Pl. 2, Figs. 1-4]. It was insulated by concentric water-jackets, and by a number of evacuated glass-tubes. In the inner part there were seven concentric water-chambers, the outer-most cylinder of these having a diameter of 12.5 cm. Outside this cylinder there was a much larger concentric chamber filled with about 80 evacuated glass tubes arranged in two layers (Fig. 5, B) a thin plate of ebonite being inserted between the layers, in order to diminish the circulation of the water among the tubes. In the lid and the bottom, there were 7 plates of indiarubber, placed horizontally, alternating with intervals of insulating water-layers.

Fig.5.The Great Nansen Insulated Water-Bottle. (Scale of A and B 1/10 of natural size. C,D , and E   1/5 of natural size. (Remark. May be out of scale!))

The water-sample in the central chamber of the bottle was thus protected against conduction of heat from its surroundings, at each end by 6 horizontal water-layers and 7 plates of india-rubber (having altogether a thickness of about 7 cm.), and on the sides by 7 thin concentric water-layers, and one thicker concentric water-layer containing eva­cuated glass tubes, having altogether a thickness of 6 cm. The insulation was consequently very complete. The arrangement (see Fig. (6) for releasing the lid by a propeller when the instrument was set, was the same as that mentioned in the "Oceanography of the North Polar Basin " ( NANSEN , 1902, p. 138). By means of a stopper (Fig. 6, s),which could be lengthened or shortened by a screw (b), the propeller could be set to release the lid of the water-bottle at any desired distance within the first 15 metres of its passage upwards. The propeller always worked without a hitch; it never failed to release when desired, and the possibility of releasing unintentionally was precluded. When set, the cylinders were suspended under the lid by two long rods, one on each side, sliding in two slots (Fig. 5. B, s, s) at the upper end of the external brass cylinder. These rods had balls at their lower end, which would catch in the slots and raise the cylinders when the lid was raised.
When the instrument was set the excentric levers (l, l) were kept in position by two hooks (Fig. 5, C, h, h), having each a pin (c) that fitted into a hole (E, c") in the lever. When the lid was released and dropped on to the cylinders in situ, these two hooks (h, h) were pushed aside and opened, in passing the notch k (see Fig. 5, D), thus releasing the excentric levers. These, being heavily weighted by the lead attached to them, instantly dropped, and in so doing, their inner, short ends were caught under the said notch (Fig. 5. E, k), and pressed the lid down upon the cylinders, and these again upon the bottom, with a force amounting to six times that of the weight of the lead. Thus the water-bottle would be locked with absolute water­tightness, and every possibility of the lid being again lifted, or the slightest chance of water entering during the hauling up, was precluded. This was an important improvement upon most of the earlier water­bottles.
The Deep-Sea Thermometer was fixed so as to be water-tight, by a screw and a ring of india rubber, to the brass lid, and passed, by means of small round holes which it fitted excactly , through the 7 india -rubber plates of the lid.
By a central stop-cock, the water could be tapped from the central chamber, and by two other stop-cocks from the outer concentric chambers. By opening a screw in the lid, air could be admitted to allow the water to run out through the stop-cocks. The total quantity of water enclosed in the bottle was about 5 or 6 litres.

Fig. 6. Releasing Arrangement


The Great Pettersson - Nansen Insulated Water-Bottle.

In coöperation with Professor OTTO PETTERSSON, Nansen constructed for the cruise in 1900 another large insulated water-bottle, which was made by L. M . Ericsson & Comp., in Stockholm. Many of the improvements of the instrument just described had also been introduced into this instrument. It was arranged for fixed deep-sea thermometers in a similar manner, and was insulated by a great number of concentric water-jackets, separated partly by brass tubes, and partly by tubes of ebonite and celluloid ; but it had no evacuated glass-tubes. The bottle would hold nearly the same quantity of water as the former bottle. Its insulation was very good, but it had no arrangement with excentric levers for pressing the lid down and locking it, and was therefore not so absolutely trustworthy as regards water-tightness as the former instrument. Since then, however, Nansen has also had this improvement introduced in the Pettersson-Nansen Water-Bottles made by L. M. ERICSSON. In 1900 the instrument was released by a propeller of ERICSSON 'S construction which, however, was so arranged that when the instrument was let down through the water, the propeller would screw itself out of the screwthread and run freely. It then sometimes happened that when the bottle was hauled up, the propeller would not at once catch the screwthread again, and thus the lid of the bottle was not released; the bottle was sometimes hauled up as much as 1000 or 2000 metres before the lid was released, and thus the water-sample obtained might be from a stratum entirely different from that expected. This fault was afterwards remedied, generally by using a messenger instead of a propeller. Dr. V. WAL FRID EKMAN [1905, pp. 13-15, Pl. I] has described the smaller type of this improved water-bottle, now generally used, and we will not describe it here. Our water-bottles of 1900, were larger and had many more concentric water-jackets than that small instrument and had therefore a better insulation. From 1900 to 1903, the large water-bottle has always been used for depths greater than 600 metres. The small pattern of the improved Pettersson-Nansen Water-Bottles was used during most of the cruises from 1901 to 1903, for depths down to 600 metres .

The Small Pettersson Insulated Water-Bottle.

In 1900, we had two instruments of this older pattern, described by PETTERSSON [1894]. One of them was altered so that a Nansen Deep­Sea Thermometer could be fixed in the lid. This was an improvement, which greatly increased the accuracy of the temperature-determination and also saved much time; and with this arrangement the bottle insulated sufficiently well to be used down to 300 and 400, or even 500 metres. These water-bottles were originally provided with propellers, which did not work with sufficient certainty. Before the cruises during the winter of 1900-1901 they were therefore altered so as to admit of using a messenger.

The Nansen Stop-Cock Water-Bottles.

In 1900 Nansen also had several water-bottles made for attaching to the sounding line at intermediate depths, when deep soundings were taken. One of these water-bottles was made of two brass tubes about 1 metre long and 2 cm. internal diameter (Fig. 7). These tubes had stop-cocks at both ends similar to those used in the Buchanan Stop-Cock Water-Bottle of the Challenger Expedition. (Cf. Challenger Report, Narative , vol. 1, Part 1, p. 113 .) The arms of the four stop-cocks are attached by hinges to a rod, placed between the two tubes in such a manner that the tubes are movable up and down, round the hinges formed by the arms of the stop-cocks. At the upper end of the rod is an arrangement for releasing by a propeller, of the same pattern as that of the Nansen Insulated Water-Bottle (see above Fig. 6).

Fig. 7. The Nansen Stop-Cock Water­ Bottle (Scale 1 / 10 of natural size (Remark. May be out of scale!))

When the tubes were raised, and attached by their upper ends to the two hooks that are set by the releasing propeller, the stop-cocks are open, and when the instrument is let down, the water may run freely through the tubes, being forced into them through the conical mouth-pieces at their lower ends. As the diameter of the apertures of the open stop-cocks, was the same as those of the tubes, the water in the latter, even without these mouth-pieces, would be changed with almost equal rapidity as the instrument passes through the water, it being only impeded by the friction against the sides of the tubes.
When the instrument has been hauled upwards a certain distance, which can be regulated as desired, the propeller will release a hook (on the principle mentioned above, see Fig. 6); the tubes become free, and will drop, by their own weight, on the hinges on both sides of the rod, to their lower position, thus closing the stop-cocks. By means of two small ratchets or stop-springs, the tubes are now prevented from again being raised. At the lower end of each tube there is a small stop-cock for taking the water-sample, and at the upper end is a small screw with a hole for letting in air when opened. In this screw there is also a small safety-valve consisting of a tiny hole, over which an india-rubber bladder is tied at the head of the screw, to make it water-tight. When the water expands by being hauled up from great depths, the superfluous water can escape through the tiny hole into this bladder.
By means of jam-nuts at the upper and lover ends of the central rod, the instrument can easily be attached anywhere to the sounding-line.
The instrument worked very well, and closed perfectly; but by some mistake the tubes had not been tinned or nickel-plated inside. The brass was therefore soon corroded by the sea-water, and copper­salts were formed on the inner surface of the tubes. It thus happened that some of the water-samples were so much contaminated by copper­salts that the determinations of their specific gravity showed considerable errors, and the samples were useless for that kind of determination; but the amount of chlorine they contained, as determined by titration, did not appear to have appreciably altered.
Another water-bottle on the same principle was also constructed in 1900. It had only one tube with stop-cocks, the tube on the other side of the central rod being replaced by a reversing apparatus for reversing thermometers. The tube of this water-bottle was widest in the middle, and tapered off towards the stop-cocks at either end; the diameter of the apertures of the tube was thus much smaller than that of the central portion. Though of no great length, this tube would hold more than half a litre of water; and by means of a conical mouth-piece at its lower end, the water was made to flow through it with sufficient rapidity during the descent, without running the risk of water from one water-stratum being dragged down in it into another. We think that this kind of water-bottle is to be recommended for future use, as it is very handy, perfectly water-tight, and can be made at a comparatively small cost.

The Ekman Reversing Water-Bottle.

This water-bottle [cf. EKMAN , 1905, pp. 27-28] was used for greater depths during the cruises of 1903 and 1904. The first instruments made, however, were too delicate; after being used for some time, the brass rods which press the lids towards both ends of the cylinder and close the water-bottle, became bent and therefore did not work sufficiently well. For this reason the instruments had to be frequently tested and repaired. As they are now made, they work very well and are very easily handled.

Back to Contents             Continue to Chapter III. 4 The Preservation of the Water-Samples.