Physiological Effects of a Hypotonic Environment: I. The Action of Hypotonic Salines on Isolated Rhythmic Preparations from Polychaete Worms (Arenicola Marina, Nereis Diversicolor, Perinereis Cultrifera)

The writers were studying the action of ions on an isolated rhythmic preparation from A. marina, and, for the reasons stated above, thought it worth while to make some additional experiments on the behaviour of the preparations in hypotonic, but balanced, solutions. Comparative experiments were also undertaken on preparations from two other species: Nereis diversicolor, because of its power of osmotic regulation, and Perinereis cultrifera, a species closely related to the last, but much less able to survive under brackish conditions. The results are described below, and discussed as they arise, with reference to the relations between the different species and their environments.

The “isolated extrovert” of Arenicola marina has already been fully described (Wells, 1937; Wells & Ledingham, 1940). It consists of the proboscis with a certain amount of oesophageal tissue attached, and gives a characteristic behaviour pattern, consisting of alternating periods of vigorous rhythmic activity and rest.

We found it possible to make similar isolated extroverts—i.e. consisting of corresponding parts—from Nereis diversicolor and Perinereis cultrifera. In no case did we include the circum-oral nerve ring, the brain or any part of the ventral nerve cord in an isolated extrovert. The preparations consisted of proboscis and oesophagus only, and contained the stomatogastric nervous system. Extroverts from Nereis and Perinereis showed prolonged activity, but they were more delicate than that of Arenicola—the preparation from Nereis diversicolor is particularly fine and easily injured by careless handling. We therefore made other experiments, in the case of Nereis and Perinereis, on longitudinal strips of body wall cut from the anterior third of the worms. These strips were ventral and included the nerve cord (but no part of the nerve ring, as they began some segments behind the mouth). They were active, robust and very satisfactory for experimental purposes.

Movements were recorded by means of light isotonic levers—lighter for the extroverts than for the body-wall strips, whose activity is somewhat improved by a very moderate degree of stretching.

All the preparations used remain active for many hours when suspended in sea water. We therefore assumed that sea water was a normal saline. Other fluids were made by diluting sea water with a thoroughly aerated solution of NaHCO₃M/400 in distilled water. This ensured constancy of pH at about the sea-water value. Concentrations are given throughout as percentage of sea water in the mixtures.

Most of the experiments, including all those incorporated in Figs. 4–6, were done in London at temperatures ranging from 14 to 20° C. A few, which were done during a heat wave at temperatures over 20° C., gave unsatisfactory results and were discarded. The experiments on a sudden return from hypotonic to normal fluid were done partly in London and partly after our evacuation to Bangor, where the temperatures were lower (down to 7 ° C.); this did not make any essential difference to the results described below.

The reactions of all the preparations used, to dilution of the bathing medium, are essentially alike, irrespective of species and of whether they are innervated by nerve cord or by stomatogastric system. This general similarity is interesting because, even within a single species, the tissues of extrovert and body wall show striking differences in their reactions to drugs, such as adrenaline and acetylcholine (Wells, 1937; Wu, 1939).

If a preparation is bathed by sea water, which is then suddenly replaced by a hypotonic mixture, its reaction consists of the following three phases (Fig. 1):

• (1) Excitation. The immediate response to the change is a brief period of vigorous rhythmic action, with tonic contraction.

• (2) Inhibition. After the excitation phase, the rhythm becomes inhibited. The depth and duration of this inhibition vary with the new salinity, and also with species. In low salinities, activity ceases altogether. In 50 % sea water, Perinereis tissues show complete stoppage for a time (Fig. 1), while those of the other two species show only irregularity and loss of vigour. Typically, the preparations of all species relax fully as they become inhibited, but with very low salinities they may remain in a state of partial contracture for many hours.

• (3) Accommodation. If the new salinity is not below a “lower limit” characteristic of the species, spontaneous activity gradually reappears, sometimes after several hours’ quiescence (Figs. 1, 2, 3).

Evidently, two distinct factors are affecting the tissue. The sudden change of salinity has evoked phases (1) and (2), which we term “shock effects”. The new salinity may also have permanent effects which will be seen in phase (3) after the shock effects have passed off. Either a permanent effect or a shock effect might be of biological significance.

To meet this situation we worked according to the following plan. A number of experiments (constant exposures) were carried out as just described, i.e. the preparations were first allowed to settle down in sea water, then a hypotonic fluid was suddenly applied instead. The behaviour was then recorded for many hours in the new solution. In other experiments (drifts) the preparations were first mounted in sea water, and then buffered distilled water was slowly run in, in such a way that the salinity fell smoothly and steadily along an exponential curve. Fast and slow drifts were tried, the main object of these experiments being to find out the highest rate of dilution to which the preparations could be exposed without the appearance of shock effects. The details of the methods used are described in an Appendix (p. 350).

The chief results of the experiments are represented diagrammatically in Figs. 4–6. Each horizontal line summarizes the results of our constant exposures to that dilution. The initial brief excitation is omitted from the diagram, which therefore only shows phases (2) and (3) as listed above. It will be noted that the general form of the diagrams is the same for all of the three preparations shown, though they differ quantitatively, e.g. in the absolute magnitude of the lower limit. In all, the length of the inhibition phase increases greatly as that limit is approached.

The drift experiments appear on the diagrams as exponential curves. In the slowest drifts, the preparations are active down to a salinity roughly equal to the lower limit found in the constant exposure experiments. This indicates that the shock effects of sudden dilution are entirely temporary and reversible. If they included any permanent change in the tissue, the constant exposures and slow drifts would give different lower limits. It may be objected that if yet slower drifts had been done, the preparations might have remained active at even lower salinities. In the case of Perinereis cultrifera this is probably true. Unhappily, we could not test this possibility experimentally because, with the slowest drifts here recorded, the duration of the experiment is about as long as the isolated preparations will stand. The general form of the salinity diagrams strongly suggests that the lower limits reached by the two methods of experimentation are in fact identical.

In the more rapid drifts, as can be seen in the diagrams, inhibition appears above the true lower limit. It is easy to show, by stopping the drift when this inhibition appears and recording the gradual reappearance of activity as the preparation accommodates itself to the new salinity, that we are simply concerned here with a reappearance of phase (2) of the reaction to sudden dilution. In very fast drifts phase (1) also appears, as a brief period of hyperactivity when the drift begins (Fig. 3).

The foregoing remarks describe the general behaviour common to all the preparations used. We now turn to certain special problems—first, the lower limits for the three different species, in relation to their ecological distribution, and then the possible biological significance of the effects of too rapid change.

The experiments on Arenicola were mostly done in London on material from Plymouth. A few points were completed at Bangor, using worms collected locally.

The lower limit for the extrovert of this species seems to lie near 15%. In constant exposures to 20 % rhythmic activity, of the characteristic intermittent type normally shown by this preparation, invariably reappeared. In 10 % no spontaneous activity was seen. Four extroverts were exposed to 15 % for 22 hr., with the following results : one gave no activity at all; one showed very small, irregular contractions some hours after the drop, and after 16 hr. exposure began to give well-defined outbursts of normal frequency (12 outbursts per hour in sea water, 10 in 15 %) but of minute amplitude; two gave very slight, sluggish tone waves but no true rhythm. In the slowest drifts, the existence of a lower limit in the range 15–20 % was confirmed (Fig. 4). We noted, in the constant exposure experiments, that even after 22 hr. in 10%, the extrovert shows temporary excitement on returning to sea water and a slight contracture can be elicited by K excess. After 22 hr. in 5 %, however, they seem to be dead.

At first sight the existence of a lower limit at 15–20% seems to agree with the known facts about the distribution of the species. It occurs down to about 20–25 % sea water; Krogh (1939, p. 47) quotes it as “living in nature at concentrations between ocean water and 8‰”. Its body fluids are isotonic with the environment at all salinities (Schlieper, 1929). There is, however, a serious difficulty. At salinities much higher than the lower limit our records show a considerable loss of amplitude, whose exact extent varies considerably from preparation to preparation. This occurs both in constant exposures and in slow drifts. The amplitude in is usually about a half, or even less, of its original value in sea water, while at 20–25 % it has fallen to a quarter or a fifth (Fig. 2). It seems that the contractile mechanism, or perhaps some part of the conducting system, fails before the excitor mechanism. In slow drifts, the contractions get smaller and smaller, but the characteristic intermittent behaviour pattern appears with normal timing, right down to the lowest salinity at which activity can be detected. Similarly, in constant exposures, when a preparation is emerging from the inhibition phase, the contractions first appear with normal timing but with minute amplitude, which gradually increases until it stays at the level determined by the new salinity.

The result, then, of gradually diminishing the salinity to which an Arenicola extrovert is exposed, is a steady falling away of amplitude, which finally leads to the complete disappearance of activity. It is impossible to say at what point the preparation becomes unfit for normal functional activity. The “useful limit” of the extrovert is certainly well above the lower limit below which no spontaneous movements, however small, can be seen. Our results seem therefore to conflict with the known distribution of the species, and we are faced by two alternatives. Either the reduced amplitude is due in some way to our experimental methods, and would not occur to a significant extent if a corresponding dilution of the body fluids of an intact worm took place, or there is a physiological difference between the worms used by us and those occurring in situations at the brackish end of the distribution range.

The Nereis body-wall strip gives vigorous records, in which it is generally possible to detect periodic “activity outbursts”, like those seen more clearly in Perinereis, but more or less masked, in this case, by a background of practically continuous activity. The extrovert shows a rather irregular and variable rhythm, and activity outbursts are usually evident in this case too.

The tissues of Nereis diversicolor are far more tolerant of low salinities than are those of either of the other two species used in this work. The precise values seem, however, to vary to some extent, either with season or with locality, for we got rather different results in the following two sets of experiments :

• (1) Most of our work was done in London in June and July on Plymouth worms. Both body walls and extroverts were studied, and the results got with the former are summarized in Fig. 5. The lower limit seems to lie near 10 %. In constant exposures, the following results were obtained with the body-wall strips: in 15 %, resumption of activity after inhibition lasting 4–5 hr.; in 10%, four preparations became vigorously active after 7–12 hr. inhibition, while four only became slightly active; in 5 %, five preparations out of six showed no activity, while the sixth gave a few weak, slow contractions at long intervals, beginning $1512$ hr. after the salinity drop. The results got with the extrovert agreed with those of the body wall.

• (2) We also made some experiments at Bangor in April, on worms from a very brackish creek at Malltraeth, Anglesey. The animals were kept in sea water for some days before use, to make the conditions of the experiments as similar as possible to those of the others. Only body walls were used. The preparations were more resistant than those just described. In 5 % sea water, all of three strips became quite vigorously active, although with rather reduced amplitude, after inhibition lasting from 10 to 13 hr. In $212$% sea water no activity was seen. The duration of the shock effects was also somewhat shorter, in animals of this group, than in those of Fig. 5.

The general conclusion, that the tissues of N. diversicolor are exceptionally resistant to low salinities, is further supported by the fact that the preparations from this species show no falling off in amplitude down to about 10 %.

As is well known, N. diversicolor is found over a very wide salinity range. It is usually conspicuous in the fauna of brackish waters and penetrates far into the Baltic (Heinen, 1911); it occurs at all intertidal levels (Thamdrup, 1935); it can also live in highly concentrated saline waters (Hesse et al. 1937; Fauvel, 1923). According to Schlieper (1929) it can be kept alive for 14 days in fresh water. This euryhalinity has been developed in two ways : firstly, the worm is able to maintain a considerable degree of hypertonicity even in very dilute media (Schlieper, 1929; Beadle, 1937; Zenkewitch, 1938), and secondly, its tissues have acquired an exceptional degree of resistance to hypotonic bathing fluids. Indeed, at first sight the ability of its neuromuscular mechanisms to function down to 5–10 % of sea water would seem to be unnecessary, since even in water of Δ = −0·04° it can maintain an internal osmotic pressure of Δ = −0·5° (Schlieper, 1929), and we seem to be confronted with another instance of the phenomenon of “duplication of mechanism” to which Barcroft (1934) has drawn attention. Possibly the high degree of resistance of the tissues is useful if the external surfaces of the animal are injured, or if it is exposed to high temperatures—a factor which has been shown greatly to reduce the effectiveness of osmoregulation in various crustacean species (Widman, 1935; Otto, 1937) and of volume regulation in N. diversicolor (Beadle, private communication).

The Perinereis body wall typically gave a rhythmic pattern in which more or less well-defined activity outbursts were conspicuous—a fact whose possible biological significance has been discussed elsewhere (Wells, 1939). The extrovert gave records like those of the Nereis extrovert, except that the rhythmic outbursts were usually more pronounced.

The tissues of Perinereis are even more sensitive to dilution than those of Arenicola, but in this case, as in Nereis, we have found slight differences between two batches of animals. Both were sent from Cullercoats. The first was studied in London in July, the second in Bangor in April.

Fig. 6 was drawn from the results of the first batch. Both body wall and extrovert showed a lower limit between 20 and 25 %. A curious feature of this batch was the extraordinarily long inhibition period shown after downward change. In the case of the body-wall strips, even 50 % produced complete quiescence for over an hour, while of two strips exposed to 25 %, one was inhibited for $1412$ hr. and the other for $2012$⁠, before accommodation resulted in the resumption of activity. The drift seems to have been too fast for the tissues of this batch; it was done on two body-wall strips and two extroverts; the former became inactive at 30 and 23%, and the latter at 33 and 25 %. Probably this is another expression of their great sensitivity to shock, and a slower drift would have brought them all to 20–25 % before activity ceased.

In the second batch only body walls were used. They became active in 20 but not in 15 %. This group showed more resistance to dilution and less marked shock effects than the other. Both, however, were more sensitive than the extrovert of Arenicola marina.

Perinereis cultrifera has a very limited power of penetrating into water of low salinity. It generally lives fairly far down the beach. In the north-eastern Atlantic, however, it penetrates to some extent into estuaries (Hesse et al. 1937), and it occurs at Sevastopol at a salinity of 18‰ (Zenkewitch, 1938). It is found in the Kattegat but not in the Baltic (Heinen, 1911).

The body fluid of our worms was approximately isotonic with the external medium. This statement is based on measurements kindly made for us by Dr N. K. Panikkar, using A. V. Hill’s thermocouple technique. Worms from our London batch were left overnight in 25 % sea water. The following values were then found, expressed as percentage of NaCl in a solution isotonic with the fluid investigated: external medium, 0· 834; body fluids of three worms, 0·848,0· 857, 0·874 respectively. There was therefore a very slight degree of hypertonicity.

To judge by our records, the activity of the preparations, even of the most sensitive batch, is not seriously interfered with down to about 25 %. Even assuming complete isotonicity of internal and external media, this would allow the worms to live in fairly dilute waters, including Sevastopol and Kattegat water. The story is, however, more complicated than these data suggest. According to Zenkewitch (1938), Sevastopol worms can maintain a markedly hypertonic internal environment over a limited range of external salinities; in 25 % sea water his animals showed this ability to a much greater extent than did ours, as measured by Panikkar. There seems, therefore, to be a physiological difference between the two groups of animals. Although the Cullercoats’ worms could probably live in Sevastopol water if they were transferred to it slowly enough, the Sevastopol worms seem to have developed special adaptations to their environment.

Under estuarine conditions, the salinity of the environment fluctuates, and we must now enquire whether the worms are likely to encounter salinity changes in their natural habitats, rapid and great enough to produce shock effects in their tissues, like those seen in in vitro experiments.

Any salinity change in the environment will of course be “damped” before it acts on the tissues, since the internal osmotic pressure will only follow the external after a certain time-lag. Fortunately, L. C. Beadle (1937) has published measurements of the rate at which the internal osmotic pressure of Nereis diversicolor falls when the whole worm is suddenly transferred from 100 to 25 % sea water at 15° C. His measurements are included as circles in Fig. 5. Each point represents the vapour pressure of the body fluid of a single worm, using A. V. Hill’s thermocouple method. Mr Beadle has also kindly allowed us to incorporate, in Fig. 4, unpublished measurements made by him in the same way with Arenicola marina. In this case also, the worms were transferred from 100 to 25 % sea water but the temperature was 21° C. It will be seen that the curve falls more steeply in the latter species, which lacks an osmoregulator mechanism.

To find out what these salinity-time curves mean to the tissues of the worms, we “aimed” drift experiments through Beadle’s points. This was made difficult by the fact that the points do not lie on an exponential curve; they are too steep at the beginning. In the case of Nereis diversicolor, we decided that a close enough approximation could be got by running 40 % sea water into 100 % at an appropriate rate. The resulting exponential lies fairly near the points (thin broken line in Fig. 5). In the case of Arenicola marina, we started with the preparations in 100 % sea water and ran in 28 %, using two Mariotte bottles and capillaries which delivered at approximately equal rates. One was used during the whole drift and the other for the first half hour only, to steepen the first part of the curve. Even with this method our curve was less steep at the beginning than Beadle’s; but the agreement is fairly good (thin broken line in Fig. 4). The drift was stopped after 6 hr., at which time the salinity had fallen to 29%, and the record was continued with the preparations still at that strength.

When a whole A. marina is placed from pure into dilute sea water it becomes very restless, and makes frequent gulping and burrowing movements. After a little while it becomes more passive. This may help to account for the rapid water uptake at the beginning of the immersion period, shown by the great steepness of the first part of Beadle’s curve.

If Arenicola extroverts are drifted along the Arenicola dilution curve (i.e. the fine broken line of Fig. 4), the disturbance of function is at no stage very severe. There is initial excitement, resulting in continuous activity for some minutes, and from about 55% the normal pattern is somewhat disturbed; but there is no paralysis at any stage. On stopping the drift at 29%, the normal behaviour pattern was very rapidly resumed. Clearly, when whole worms are suddenly transferred from 100 to 25 % sea water, their internal osmotic pressure drops rapidly enough to disturb the normal functional pattern to some extent, but not rapidly enough to cause inhibition. It should be noted that in this discussion we are limiting ourselves to the actual effects of osmotic pressure, and leaving aside any consideration of turgor or other mechanical effects of the volume of fluid taken up; in the case of A. marina, these effects have been described by Reid (1929 a).

If Nereis body-wall strips are subjected to the same drift, i.e. the fine broken line of Fig. 4, they show rather more effect. The amplitude drops to about a quarter of its original value, and then recovers when the drift is stopped. The picture is very different if the Nereis dilution curve is followed. Two body-wall strips, and two extroverts, were drifted along the fine broken line of Fig. 5. They all remained vigorously active throughout the experiment, and at no stage was there the slightest sign of disturbance. This shows that when whole worms are suddenly transferred from too to 25 % sea water, their internal osmotic pressure falls too slowly to have any effect on their neuromuscular mechanisms. The result of drifting Nereis preparations along the Arenicola curve shows that Nereis has its osmoregulator mechanism to thank for its immunity against osmotic shock effects.

The osmotic change in sudden transfer from 100 to 25% sea water is probably more violent than any that the worms encounter under natural conditions. Measurements of salinity variations in the interstitial water of the mud in which Nereis and Arenicola burrow have been published by Bruce (1928), Reid (1929 b, 1932) and Thamdrup (1935). At the depth to which the worms burrow the changes are not great. It should, however, be remembered that the water inside the burrow is not identical in composition with the interstitial water of the mud. Both Nereis and Arenicola ventilate their burrows vigorously (van Dam, 1937; Lindroth, 1938), and the existence of a yellow layer of oxidized sand in the wall of the burrow shows that the water inside differs, at least as regards oxygen content, from that outside. We do not know what happens inside a worm burrow when heavy rain falls at low tide. However, the results of our experiments based on Beadle’s measurements make it almost incredible that the salinity changes are severe enough to produce osmotic shock in the tissues of these two species.

The tissues of Perinereis are more sensitive to sudden dilution than those of Arenicola or Nereis’, unhappily, data on the “damping” effect are not available for Perinereis, so we are unable to extend our analysis to this case.

Hitherto, we have been concerned exclusively with the effects of downward salinity change. A number of experiments were also made with upward changes, using the Arenicola extrovert and the Nereis diversicolor body wall. The results were peculiar, for in the case of downward changes all the different preparations gave essentially the same type of response; with upward changes on the other hand, the responses seem to differ in kind. With the Nereis body wall, we find that a sudden return to sea water evokes a cycle of reactions very like those seen after sudden downward change (Fig. 7). As one would expect, these shock effects can be avoided by slowly drifting upwards from the hypotonic to the normal fluid. The Arenicola extrovert, on the other hand, shows temporary excitement, whose duration is short as regards the rhythmic mechanism and somewhat longer as regards tone (Fig. 7); it does not show the shock inhibition, which is such a conspicuous feature of the Nereis records. This difference was consistently found in Plymouth and Bangor specimens of both species.

1. The reactions of isolated rhythmic preparations from Arenicola marina, Nereis diver sic olor and Perinereis cultrifera to hypotonic salines are described.

2. The preparations used were (1) the “isolated extrovert” of all three species, (2) ventral longitudinal body-wall strips of the two last named. All these preparations are essentially alike in their reactions to dilution of the bathing medium.

3. On abruptly changing from sea water to a hypotonic fluid, responses of the following general type are seen: first, brief excitement; then a phase of more or less complete inhibition; finally, provided the hypotonic fluid is not below a lower salinity limit characteristic of the preparation, gradual return of activity as the preparation accommodates itself to the new medium. The first two phases are shock effects of sudden dilution. The inhibition phase may last for many hours.

4. Preparations were exposed to salinities which fell gradually at various speeds. From the results of these experiments it is inferred that shock effects of rapid change are unlikely to be evoked under natural conditions, at least in Arenicola and Nereis.

5. The lower salinity limits for spontaneous activity in the tissues of the various species are: Perinereis cultrifera, 20–25% sea water; Arenicola marina, 15–20%; Nereis diversicolor, 5–10%. These results are discussed with reference to the ability to live in brackish water.

6. On suddenly returning from a hypotonic fluid to normal the responses vary. There may be relatively slight excitation (Arenicola marina extrovert) or a cycle of excitation—inhibition—accommodation like that evoked by a sudden downward change (Nereis diversicolor body wall).

The work was begun at University College, London, and completed after our evacuation to the University College of North Wales, Bangor. We are glad of this opportunity of expressing our thanks to Prof. Brambell and his staff for the hospitable welcome with which we were received. Our thanks are also due to Dr L. C. Beadle for kindly allowing us to use his unpublished measurements in Fig. 2, to Dr N. K. Panikkar for determining the vapour pressure of Perinereis cultrifera body fluid, and to Miss Clementina Gordon for assistance with some of the earlier experiments.

As we have used the same methods in the present work and in connexion with other problems, we publish the details as an Appendix to facilitate subsequent reference. Our experiments were of two types, which we term “constant exposures” and “drifts”.

In this case the changes of solution are made as suddenly as possible, but between changes the preparations are exposed to constant environments.

The device shown on the left of Fig. 8 has proved very convenient for this purpose. The preparation (dotted) is suspended inside tube A, which has an open cup-shaped top and a sloping side-arm B, along which surplus fluid can run away. Aeration is done through capillary tube C, blown to the lower end of A. Inflow tube D is constricted where it fits into A, and the dimensions of the two tubes are such that a centimetre of pressure tubing, run on to D, acts at this point as a rubber stopper. Above, D ends in a hook to which the preparation is attached. Just below this hook, a hole in the side of D allows new fluid to enter. By running in a generous amount of the new fluid, very rapid and complete changes can be made without interrupting the record. Thread L runs to the recording lever.

For constant exposure experiments of long duration, a vessel of larger capacity is desirable, so that the bathing fluid need not be frequently renewed. In this case it has been our practice to mount preparations in the cylindrical jars used for drift experiments (see below), and to change the fluid simply by siphoning off the old as completely as possible and then pouring in the new through a tube running to the bottom of the jar. An excess of the new fluid was always run in, so that some of it escaped over the rim of the jar and washed away any residue of the old, and the completeness of the change was always checked by titration of some appropriate constituent of the fluids. This method has the drawback that the record is somewhat defaced at the moment of change.

In experiments of this type the preparation is exposed to a fluid whose composition “drifts” slowly and steadily along an accurately predictable curve, whose steepness can be varied at will.

Cylindrical jar E (Fig. 8) stands in an outer vessel F, and contains the preparation attached to glass hook G. Thread L runs to the recording lever. (We generally used jars of capacity 385 c.c., in which two or three preparations could be mounted and recorded simultaneously.) At the beginning of the experiment, E is filled to overflowing with solution I (generally sea water in our case, which is a suitable “normal” fluid for polychaete preparations). When the preparation has settled down, the “drift” begins. A second solution is run in from a Mariotte bottle through a fine capillary H, i.e. at a constant, slow rate. The Mariotte bottle should be connected to H with stout pressure tubing and the flow rate should be regulated by adjusting its height (a screw clamp on the tubing leads to irregularities). The contents of E are kept continually mixed by air entering through J, and, as fast as the second solution flows in, a mixture of the two fluids escapes over the rim of E into F.

In other words, x will fall along an exponential curve. Note that vt is the volume of the second solution used at time t. For determining this quantity, the Mariotte bottle should be provided with a volume scale. From vt it is easy to derive the composition of the mixture in E at any moment.

Our routine is to read the volume of solution II used at fairly frequent intervals, to make sure that variations in flow rate do not occur, and to plot the composition of the fluid in E against time from the readings. At the end of the experiment, x is checked by means of an appropriate titration of a sample of the fluid from E—in the present case, by titrating for chloride with silver nitrate. The method has proved to be convenient, accurate and reliable in a great number of experiments, with a wide range of different drift rates.