1. Several species of marine invertebrates, and an elasmobranch, have been kept in diluted media. The increase of body weight so caused was compared with the resulting dilution of the body fluids.

  2. The bounding membrane of the invertebrates was permeable to salts when the animals were immersed in diluted sea water.

  3. The bounding membrane of the elasmobranch was semipermeable, i.e. permeable to water but not to solute. There is a close quantitative agreement between the osmotic swelling observed and the diminution of the osmotic pressure of the blood.

L. Frédéricq (1885) found that the content of soluble salts in the body fluid of several marine invertebrates is approximately equal to that of the surrounding sea water, and suggested that salts are able to penetrate the gill membrane like a dialysis membrane. This theory was tested by several investigators, either by measuring the amount of osmotic swelling, or by determining (by chemical analysis) the actual transfer of substances through the membrane.

Margaria (1931) and the author have confirmed that when a typical marine invertebrate, or an elasmobranch, is transferred to a diluted medium an equalisation of osmotic pressure between the internal and the external media takes place. If this equalisation is effected by dilution of the internal medium by the surplus water entering through the bounding membrane, the weight of the animal as a whole should increase. If, on the other hand, dilution takes place by the disappearance of salts and other dissolved substances from the internal medium, the weight should remain unchanged.

Quinton (1900) observed a progressive increase, or decrease, in the weight of Aplysia, Sipunculus and Arenicola when kept in diluted or concentrated media. Bottazzi and Enriques (1901) suggested that this change of weight might be due to the variable amount of water contained in the gastro-intestinal tube. Osmotic swelling was noticed by Schücking (1902) in Sipunculus, by Garrey (1905) in Asterias, Nereis and Chaetopterus, and by Dakin (1908) in Cancer pagurus, Hyas araneus, Doris tuberculata, and Arenicola marina. Recently some measurements were made by Schlieper (1929), who confirmed that the increase of body weight in diluted media was transitory in Cancer pagurus, reaching a maximum in 3–10 hours, as well as in Nereis diversicolor, beginning to diminish on the second day. Bethe (1930) found the weight increase in Aplysia also to be transitory, reaching a maximum on the second day, but failed to observe an increase of body weight in Carcinus maenas.

Experimental

The following experiments were performed in Plymouth in September, 1930. The weights of animals (Portunus puber, P. depurator, Carcinus maenas, Maia squinado and Cancer pagurus, kept in diluted media, were measured. Moisture adhering to the skin was absorbed by a dry duster, in which the animal was wrapped and placed on the balance. Weighing was to 0·1 gm. The greatest possible care was taken to shake out the water contained in the gill cavities.

The results for the crustaceans used agreed with those of Schlieper (1929), showing a transitory increase of weight directly after the animals were transferred to diluted media1.

The increase never exceeded 4 per cent, of the initial weight. The equalisation of osmotic pressure, however, was almost complete in less than 24 hours, as was shown by direct measurement (Fig. 1,,A). The osmotic pressure was determined by the vapour pressure method (Hill, 1930,a), (Margaria, 1930). Now if it were supposed that the equalisation took place by osmosis of external water, it would be possible to calculate the amount of water necessary to dilute the body fluid to the degree shown in curve A. For this purpose the water content of the animals was determined and the following values were found (Table II).

Table I.

Changes of weight of animals in dilute media.

Changes of weight of animals in dilute media.
Changes of weight of animals in dilute media.
Table II.
graphic
graphic
Fig. 1.

Portunus puber. A, osmotic pressure observed (ordinate on left) ; B, water transfer calculated by assuming semipermeability (ordinate on right); C, actual increase in body weight observed (ordinate on right).

Fig. 1.

Portunus puber. A, osmotic pressure observed (ordinate on left) ; B, water transfer calculated by assuming semipermeability (ordinate on right); C, actual increase in body weight observed (ordinate on right).

If it be assumed, from the analogy of mammalian blood (Hill, 1930 6), that 97 per cent, of the water in crustacean haemolymph is in the “free” state, being able to participate in osmotic dilution, approximately 70 per cent, of the body weight of P. puber is free water. The amount of water necessary to increase the dilution n times, i.e. to decrease the osmotic pressure to 1/n of the normal value, is therefore (n − 1) × 70 per cent, of body weight. The difference between the actual (curve C) and the calculated (curve B) increase of body weight is so large as to make the assumption of osmosis, or the existence of a true semipermeable membrane, impossible. It would have to be assumed that water entered to the extent of one-third of the body weight in 5 hours, and that urine was excreted at the inconceivable rate of 21 per cent, of the body weight in the first hour. For other crustaceans there are similar determinations, reported by Margaria (1931), of the time relations of the equalisation of osmotic pressure. It is certain, therefore, that in these animals the equalisation occurs mainly by the disappearance of salts and other dissolved substances from the internal medium, not by the transfer of water.

The presence, however, of an osmotic swelling, though in a very slight degree, suggests that the membrane is slightly semipermeable in the sense postulated by Schücking (1902), that is to say, it is permeable to water and in a less degree to salts. The permeability of the bounding membrane thus investigated is really that in an abnormal condition, the animal being immersed in a medium more dilute than its natural environment. Whether and how far the normal properties of the membrane are thereby affected is unknown, although the behaviour of the animals under investigation was apparently normal (Care, maenas, Portunus depurator and Cancer pagurus). This applies also to Portunuspuber, except on the last day of their survival, when some of the group showed signs of weakness.

The salts that disappear from the circulation could be accounted for in two ways : (i) by diffusion through the bounding membrane, or (ii) by being fixed and deposited in the body, either to be stored or excreted. The demobilisation, however, of salts is not likely, although Bethe (1928) demonstrated the mobilisation of Ca from a hypothetical Ca reservoir in Carcinus maenas, and Collip (1920) also found the mobilisation of Ca from the shell of molluscs, but not from the carapace of crustaceans, into the haemolymph. It is possible that, in the earlier period of immersion in diluted media, salts and other dissolved substances are excreted, together with the excess of water. The greater part, however, of the equalisation of osmotic pressure between the internal and the external media is probably effected by diffusion of salts through the bounding membrane, leading to the almost complete disappearance of the concentration gradient. In other words, the bounding membrane of marine crustaceans is permeable not only to water but also to salts. The same conclusion has been reached by Quinton (1900) on Carcinus maenas, and by Bethe (1928, 1930) on Carcinus maenas and Aplysia, by means of chemical analysis.

L. Frédéricq (1904) observed that the osmotic pressure of the blood of Scyllium kept in diluted sea water decreased to the same level as the external medium. The new equilibrium, in his opinion, was established by “transport of water from the external medium into the blood.” In animals which had been kept in a diluted medium containing i-o per cent. NaNO3 no nitrate was found in the blood. Further support to his view is the remarkable difference between blood and surrounding medium in the content both of salts and of urea. The urea content of blood, according to v. Schröder (1890), is 2·5–3 Per cent. and that of soluble salts is 1·6–2·3 per cent. Thus the blood of elasmobranchs and the surrounding sea water are in equilibrium in osmotic pressure but not in concentration of individual solutes. The bounding membranes of the dogfish appear to be semipermeable, keeping the solutes in the internal and external media from diffusing towards each other and becoming equalised in concentration. If this is so, and if the semipermeability is maintained even in diluted media, a considerable amount of water may be expected to pass by osmosis through the membrane.

Dogfish (Scyllium) were kept in various dilutions of sea water and weighed at intervals. A considerable increase of weight (Fig. 2) was invariably observed, together with marked diffuse oedema, especially on the abdominal surface. The duration of immersion necessary to increase the body weight by 10 per cent, is shown below (Table III) (interpolated from the curve of Fig. 2 and similar ones not shown).

Table III.
graphic
graphic
Fig. 2.

Time-weight curves of dogfishes kept in three-quarters sea water. No. 2, re-transferred into sea water at the time marked with an arrow, showed a quick decrease of weight toward the original.

Fig. 2.

Time-weight curves of dogfishes kept in three-quarters sea water. No. 2, re-transferred into sea water at the time marked with an arrow, showed a quick decrease of weight toward the original.

Details of the case of dogfish No. 7, which survived in excellent condition for 108 hours until killed, are given below (Table IV), and in Fig. 3.

Table IV.
graphic
graphic
Fig. 3.

Time-weight curve of a dogfish which survived in four-fifths sea water in excellent condition.

Fig. 3.

Time-weight curve of a dogfish which survived in four-fifths sea water in excellent condition.

The time-weight curve rose rapidly at first and then less rapidly, reaching a maximum in 30 hours, after which it decreased gradually, 1·1 gm. on the first day and then 1·5 gm. per day. This slow decrease in weight may be due to loss of water (or solutes) by excretion of urine, partially compensating the gain by osmosis.

If it be supposed that the bounding membrane is semipermeable, it is possible to calculate the amount of water necessary to dilute the internal medium to the same osmotic pressure as the external. The water content of a dogfish, weighing 283·5 gm., was found to be 68·8 per cent, of the total weight. Assume that 97 per cent, of this water is in the “free “state. Some lower homologues of fatty substances must have escaped in the procedure of measuring the water content (described above). The amount of free water therefore is taken as 65 per cent, of the body weight. At the end of 108 hours the blood of dogfish No. 7 was found to have an osmotic pressure equal to 0·861 of that of sea water. The water, therefore, that had been transferred into the body was equal to . of the body weight. 1·075 here is the osmotic pressure, expressed as a fraction of that of sea water, of the initial external medium, in which the dogfish had been kept for several days preceding the experiment, so that it corresponds to the initial value of the osmotic pressure of the blood. The increase of weight observed was 15·3 Per cent. Considering the complicated nature of the organism, the agreement between calculated and observed is close. The kidneys of dogfishes are not able to regulate the water output enough to diminish appreciably the state of hydropsy.

In cases which developed more or less serious pathological symptoms, the increase of body weight was large. This is possibly due to abnormal metabolites aiding in concentrating the internal medium. Even in cases in which no external symptoms other than diffuse oedema were noticeable, the unusual environmental condition must have interfered in some way with the tissue metabolism. At the end, when the dogfishes were killed and the osmotic pressure of their blood was measured, it was always somewhat higher than the value calculated from the swelling, except in No. 7, which survived in highly excellent condition for 412 days. The swelling, in fact, generally continued in these cases.

A test of semipermeability can be made in another way, i.e. by comparing in the early stages the degree of swelling with the change of the osmotic pressure of the blood. The osmotic pressure expected from the degree of swelling, under the assumption of semipermeability, is calculated by the above formula. Taking the mean of three similar experiments (No. 4–N0. 6) in three-quarters tank water, the following results are obtained (Table VII).

Table V.
graphic
graphic
Table VI.
graphic
graphic
Table VII.
graphic
graphic

Taking the mean of seven experiments (Nos. 7-13) in four-fifths tank water, the results are (Table VIII):

Table VIII.
graphic
graphic

These calculated values of the osmotic pressure agree closely with those actually determined by Margaria (1931).

It is my pleasant duty to express my sincere gratitude to Prof. A. V. Hill, who suggested these experiments, and to Dr R. Margaria and Mr J. L. Parkinson for their co-operation.

Bethe
,
A.
(
1928
).
Pflügers Arch
.
221
,
344
.
Bethe
,
A.
(
1930
).
Journ. Gen. Physiol
.
13
,
437
.
Bottazzi
,
F.
and
Enriques
,
P.
(
1901
).
Arch. (Anat.) Physiol. Supp
. p.
109
.
Collie
,
J. B.
(
1920
).
Journ. Biol. Chem
.
45
,
23
.
Dakin
,
W. J.
(
1908
).
Biochem. Journ
.
3
,
473
.
FrùDùRicq
,
L.
(
1885
).
Arch. Zool. exp. gén
.
3
,
xxxiv
.
FrùDùRicq
,
L.
(
1904
).
Arch. Biol
.
20
,
709
.
Garrey
,
W. E.
(
1905
).
Biol. Bull
.
8
,
257
.
Hill
,
A. V.
(
1930a
).
Proc. Roy. Soc. A
,
127
,
g
.
Hill
,
A. V.
(
1930b
).
Proc. Roy. Soc. B
,
106
,
477
.
Margaria
,
R.
(
1930
).
Journ. Physiol
.
70
,
417
.
Margaria
,
R.
(
1931
).
Proc. Roy. Soc. B
,
107
,
606
.
Quinton
,
R.
(
1900
).
C. R. Acad. Sci. Paris
,
131
,
905
and 952
Schlieper
,
C.
(
1929
).
Z. vergl. Physiol
.
9
,
478
.
V. Schroder
,
W.
(
1890
).
Hoppe-Seylers Z
.
14
,
576
.
SchüCking
,
A.
(
1902
).
Arch. (Anat.) Physiol
, p.
533
.
1

All dilutions were of tank water with tap water: the animals had been for some time in tank water (which is slightly more concentrated than sea water) before use : Plymouth tap water may be taken for the present purpose as equivalent to distilled water. Osmotic pressures are expressed as a fraction of that of Plymouth sea water.