Further Studies on Ionic Regulation in Marine Invertebrates

Marine invertebrates, although permeable in some degree to the ions of sea water, show a variable amount of ionic regulation in their body fluids while maintaining the same total concentration of ions in these fluids as the external medium. In a previous study (Robertson, 1949), based on the examination of twenty species belonging to five phyla, it was shown that characteristic patterns of ionic regulation existed in different groups, varying from almost complete ionic equilibrium between internal and external media in echinoderms to regulation affecting usually every ion in the decapod crustaceans. Data on other sixteen invertebrates are discussed here, and certain conclusions are reached based on both studies.

Procedure and chemical methods were the same as in the previous paper (Robertson, 1949). Ionic regulation was there defined as ‘the maintenance in a body fluid of concentrations of ions differing from those of a passive equilibrium with the external medium’. Accordingly, an analysis of a body fluid as first obtained from an animal was compared with an analysis of the same body fluid after it had been dialysed in a collodion sac against the original sea water, the latter process giving figures for physico-chemical equilibrium, in which there is formation of calcium-protein complexes and a Donnan effect with protein-rich bloods, such as those of decapod Crustacea and cephalopod Mollusca. The echinoderm, sipunculid, lamellibranchs and nudibranch gastropod all had very low protein contents, below 1 g./l., and in these cases dialysis was unnecessary. Differences from ionic equilibrium are probable only if the concentrations of ions in the original body fluids differ from those of sea water or dialysed plasma by the following percentages: Na 0·6, K 2·6, Ca 1·5, Mg 1·8, Cl 1·1, S04 1·2. In the determination of the Dorman ratios, and in some of the potassium and magnesium analyses where the concentrations in the body fluid and in sea water were very similar (e.g. in Holothuria), duplicate or triplicate estimations were made to increase the accuracy.

The chlorinity of the sea water in which the animals were kept was 19·5 %0 at Plymouth {Maia, Sepia: sea water from beyond breakwater), 22·5%0 at Naples (Holothuria, Ostrea, Mytilus galloprovincialis, Dromia, Pachygrapsus, Squilla: aquarium sea water), and 17·6-18·3%0 at Millport (the remaining animals).

Ionic exchange in holothurians may be considered to take place between the perivisceral fluid and sea water across the integument and across the epithelium of the respiratory trees. Koizumi’s (1932) analysis of Caudina chilensis indicated that in this holothurian the perivisceral fluid and surrounding sea water were practically in complete equilibrium, although potassium was always slightly higher (+ 7 %). He found also that the animal responded to changes in the values of the external ions by corresponding alterations in the internal ions, and later (Koizumi, 1935 a, b) that the isolated body wall showed complete permeability. Bialaszewicz (1933) found at Naples that Holothuria tubulosa had higher values of chloride (103·5%), potassium (no %) and calcium (109 %) than sea water (he did not estimate sodium), but his chloride figure suggests that the animals were imperfectly equilibrated with the sea water in the aquarium, and his methods of analysis for potassium and calcium were not very accurate (see Robertson & Webb, 1939).

The mean of two closely agreeing analyses of H. tubulosa are shown in Table 1, and it is evident that almost complete equilibrium obtains between the perivisceral fluid and sea water. A third specimen, apparently as healthy as the others, was found on dissection to be without a gut, thrown off no doubt on capture 4 days previously, as may happen in these animals. The analysis of its perivisceral fluid showed no regulation of magnesium, but was otherwise identical with the other analyses, thus perhaps suggesting that the gut itself was playing no major part in the slight regulation of potassium which presumably is effected by the epithelium of the integument.

An analysis of the body fluid of an unnamed species of the sipunculid Phascolosoma was made by Steinbach (1940) who was concerned chiefly with the electrolytes in the muscle. Expressed as a percentage of the values in sea water his figures for sodium, potassium, calcium and chloride come out as 86, 422, 117 and 84% respectively. It seems quite clear from the evidence given below that the animals must have been in equilibrium with sea water of a salinity about 15% lower than that given by Steinbach. The figure for potassium is open to grave doubt, as so far no marine invertebrate has shown regulation of potassium beyond 220 % (Robertson, 1949; Prosser, 1950), and that exceptional value was in an active animal (Loligo) with well-developed excretory organs concerned in ionic regulation.

The analysis of three specimens of P. vulgare in Table 1 shows regulation of all ions, but the potassium concentration in the body fluid is less than 110% of that in sea water. Magnesium is markedly decreased, and the slight rise in sodium seems to have been necessary to maintain osmotic equilibrium. Total ionic concentration comes to 1·055 g- ions/kg. water, compared to 1·065 for the sea water. To the former must be added 3 or 4 mg. ions for bicarbonate which was not determined, and this will bring the respective concentrations within 1 % of each other. Osmotic equilibrium within 1 % was also found in Bottazzi’s (1908) freezing-point determinations on the allied Sipunculus nudus. An incomplete analysis of Sipunculus by Bethe & Berger (1931) showed a reduction in magnesium to about 70% of the sea-water value. In contrast to the polychaetes Aphrodite and Arenicola, which show increases in potassium and sometimes decreases of sulphate with no regulation of the other ions (Robertson, 1949), the sipunculid Phascolosoma shows marked diminution in magnesium with a slight increase in sodium.

The decapod crustaceans are a group showing pronounced ionic regulation, the mechanism of which consists in selective excretion of ions from the blood by means of the antennary glands and a controlled uptake of ions through the gills. Owing to a high protein content, a passive equilibrium between the plasma and sea water across the gills (if it existed) would result in higher concentrations of cations and lower concentrations of anions in the plasma, in addition to a much higher calcium content because of the formation of a calcium-protein complex. Ionic regulation over and above such a possible passive equilibrium is shown by the data in Table 2.

Pachygrapsus marmoratus is peculiar in having all its ions below equilibrium values, and for its low magnesium and sulphate concentrations. The pooled plasma of twelve male specimens had a total ionic concentration of 1·163 g. ions/kg. water compared with 1·350 for sea water (1·353 including bicarbonate), equal to about 86% of sea water, a finding which may be compared with the freezing-point determinations of Schwabe (1933) A = 2·11 and 2·33° C. respectively, the blood being equivalent in concentration to approximately 91 % sea water. Hypotonicity of other members of the grapsoid group when living in sea water is well established (P. crassipes—Baumberger & Olmsted, 1928; Jones, 1941; Leptograpsus and Sesarma—Edmonds, 1935; Eriocheir—Conklin & Krogh, 1938).

Portunus depurator and P. púber belong, like Pachygrapsus and Carcinus, to the tribe Brachyrhyncha. Analysis of the plasma from five male specimens in each case revealed a regulation similar to that of Carcinus maenas (Webb, 1940), although the concentrations of magnesium and sulphate are somewhat higher. Margaria (1931) compared the vapour pressure of the blood of these two species with sea water, finding both fluids isosmotic within 2%. This has been confirmed from the present chemical data. The total concentration of the plasma of Portunus depurator was 1·070 g. ions compared with a sea-water figure of 1·091, but an additional 5 or 6 mg. ions must be added to the first figure for bicarbonate. P. púber plasma was slightly hyperosmotic, 1·105 g. ions compared to 1·082 g. ions.

Dromia vulgaris is peculiar among decapods in showing a low sulphate concentration (53%) coupled with a high magnesium (99%) and reduced calcium (84%). At first sight a sodium concentration of only 97 % of the value after dialysis is incompatible with osmotic equilibrium. However, the total ionic concentration in the plasma is 1·336 g. ions/kg. water, compared with a figure of 1-354 for the surrounding sea water, and this slightly lower sodium concentration can be anticipated from the reduced sulphate.’ In a mixture of electrolytes corresponding to sea water, a reduction of the sulphate concentration to a half and replacement by chloride will lower the total concentration of ions by 0·9%, increase the chloride to 103 % and decrease the sodium to 97%, while maintaining complete equilibrium, all these changes arising from the poor osmotic properties of sodium sulphate (Robertson, 1949). The figures for Dromia refer to the mean of eleven crabs, six males and five females. Little difference was found between the sexes, except in potassium. The separate figures are (males first) Na 97, 96; K 113, 126; Ca 82, 86; Mg 98, 100; Cl 104, 103; SO₄ 54, 52.

The ionic regulation shown by the two spider crabs Hyas araneus and Maia squinado is not dissimilar, although sulphate is markedly reduced in Maia. They maintain relatively high magnesium concentrations compared with the species of Brachyrhyncha. Hyas plasma, which came from six male specimens, appears to be almost 2% above equilibrium, 1·102 g. ions/kg. water compared with 1-082 for the sea water in which it was living.

The hermit crabs Eupagurus bernhardus (six males) and E. prideauxi (twelve females) are unique among decapods in accumulating considerable quantities of sulphate, and correlated with this have lower chloride concentrations to maintain the balance of anions. Perhaps part of the inorganic sulphate is of endogenous origin, but this point has not been investigated. Both species had high protein contents, the highest so far found in intermoult decapods. Osmotic equilibrium within 1 % is indicated by the g. ions/kg. water, 1·094 in the plasma of both species compared with sea waters of 1·104 and 1·095 in the cases of E. bernhardus and E. prideauxi respectively.

Analysis of the pooled plasma from four female specimens of the stomatopod Squilla mantis showed regulation of the same type as is usually found in decapods, namely, accumulation of cations except magnesium which is markedly reduced, and some diminution of sulphate. Compared with a concentration of 1·359 in the sea water of Naples aquarium the plasma had 1·380 g. ions/kg. water.

In Table 3 are set out comparative analyses of the plasma and the secretion of the antennary glands in a large male Maia. Although the secretion is isosmotic with the plasma, selective excretion by the glands of every ion except potassium is shown by the data. While the higher levels of magnesium and sulphate in the secretion are correlated with plasma values much below those of equilibrium, and a lower level of chloride with a value slightly above equilibrium, there is no such correlation in the case of calcium. The calcium excretion is tending to reduce the blood calcium, since at 99 % of the plasma level it is 9 % above the ultrafilterable calcium. However, the plasma calcium of this specimen is about 16% higher than the dialysis value.

Potassium of the blood is above that of sea water or of dialysed blood, and the antennary gland secretion is tending to maintain the high value since its potassium is at the level of an ultrafiltrate of the plasma. A level of sodium ions in the secretion higher than in an ultrafiltrate of the plasma would tend to lower the sodium of the plasma, which, however, is in approximate dialysis equilibrium with the sodium of the sea water. This relatively high sodium of the secretion would seem to be necessary to balance the high level of divalent sulphate ions. It can be calculated that if sulphate had not been selectively excreted, that is, had been 104% instead of 214% of the plasma value, the sodium figure would have been about 2·8% below its value in Table 3 to maintain osmotic equilibrium with the plasma.

Selective excretion is only one factor in ionic regulation. Equally essential is the controlled uptake of water and ions which occurs simultaneously with the formation of antennary gland secretion. In the Maia specimens of Tables 2 and 3 absorption of potassium, calcium and chloride must take place against concentration gradients, although magnesium and sulphate may be able to enter the blood in accordance with the diffusion gradient. Sodium ions and water must also be absorbed, probably actively, completing the final constituents of a fluid, isosmotic with the plasma, which replaces the isosmotic secretion of the antennary glands. The data in Table 3 amplify those of Bialaszewicz (1932) for the same species, although there are a few discrepancies between the two sets of results. On a volume basis he found in two cases potassium in the secretion to be 17·9 and 17·0% below that of an ultrafiltrate of the blood, a more usual finding than the present non-selective excretion (cf. Robertson, 1949). Calcium was 22 and 44%, and magnesium 39 and 51% above the values in ultrafiltrates, while one value for sulphate was 45% higher. Bialaszewicz claims, however, that a proportion of certain of the elements is bound to colloids, 3-8% of the potassium, 16·4% of the calcium and 12·8% of the magnesium, only chloride and sulphate being completely diffusible across a collodion membrane. He is probably in error here with respect to potassium and magnesium, since no binding of these ions has been found by Webb (1940) or Robertson (1939, 1949) in crustacean blood, although a Donnan ratio of 1·01−1·03 depending on the protein content has been found for all the cations except calcium. In a Maia specimen with 41 g. protein/1. I have found the ratio magnesium plasma/magnesium ultrafiltrate to be 1·065, giving a Donnan ratio of 1·03, and the ratio for plasma dialysed against sea water to be 1·051, the agreement between ultrafiltration and dialysis being fairly good. The 12·8% of bound magnesium found by Bialaszewicz is probably based on an error in the interpretation of the magnesium analysis. He determined magnesium in the filtrate from calcium estimations apparently as magnesium ammonium phosphate, and in the ashed plasma the copper from the haemocyanin would come down as phosphate and be included as ‘magnesium’. This error would not arise in the protein-free ultrafiltrate. Robertson & Webb (1939) have taken copper into account in their hydroxyquinoline method for magnesium used in this paper.

Ionic regulation of blood. The figures in Table 4 reveal only slight regulation in the three lamellibranchs. Although all have considerably higher concentrations of potassium in their blood, Ostrea and Mytilus edulis are virtually in equilibrium with the external medium with respect to the other ions. The figures are from analyses of pooled plasma from ten and twelve specimens respectively. Accumulation of sulphate ions shown by the Mediterranean M. galloprovincialis is a rare feature in marine invertebrates. In analyses of nearly forty invertebrates, this has been found only in the two species of Eupagurus and in Hyas araneus. The plasma from twenty specimens of the bivalve had a sulphate concentration 125 % of that in the sea water of Naples aquarium. Ten individuals from the same batch were kept for 4 days in thoroughly aerated water, and the pooled plasma had a value of 115%. These high values were checked and confirmed by microgravimetric determinations of the sulphate as barium sulphate, in addition to the usual volumetric barium iodate procedure. While the meaning of the accumulation of sulphate ions is obscure, the consequences are slight alterations in sodium and chloride ions, the latter falling significantly owing to the presence of excess sulphate anions, and the sodium ions increasing slightly to maintain osmotic equilibrium in a solution containing more divalent sulphate ions than sea water.

Archidoris, a common opisthobranch, shows more ionic regulation than the bivalves, especially in its considerable accumulation of calcium (132%) and magnesium (107%). As in the other opisthobranch Pleurobranchus analysed previously (Robertson, 1949), the low protein content of Archidoris points to the absence of haemocyanin in this group, as already deduced by Webb (1937) from copper estimations. McCance & Masters (1937) have given figures for the ‘mucus’ secreted by Archidoris from which the visceral mass had been removed. Their figures appear to show considerably more ionic regulation in the formation of this fluid, but the ‘mucus’ would seem to have been a mixture of blood and mucus.

The peculiarities of ionic regulation in the plasma of Sepia are those already discovered in Eledone and Loligo (Robertson, 1949), a striking accumulation of potassium and a lowered sodium and sulphate, with a protein concentration exceeding 100 g./l. Separate analyses of three individuals, a male and two females, were made, the ranges of regulation being Na 92-94 %, K 193-223 %, Ca 84-97 %, Mg 97-100%, Cl 105-106%, SO₄ 17-29%. Of interest are the correlations between chloride, sulphate and sodium ions in the three cephalopods, a decreasing sulphate in the series Eledone, Loligo and Sepia being associated with increasing chloride and decreasing sodium. Granted a marked decrease in sulphate as a result of active regulation and maintenance of approximate osmotic equilibrium between the plasma and sea water, chloride ions must rise to balance the cations, while the cations themselves would decrease slightly. It can be calculated that in a synthetic sea water in which the sulphate is reduced to a fifth and the potassium doubled the chloride would be 104·7% and the sodium 93·8 % of the values in an isosmotic sea water of normal composition. These values are approached in the analysis of Sepia plasma given in Table 4. With reduced sulphate but normal potassium, the sodium figure would be higher—95·9 %.

It is probable that the excretory organs play a part in the ionic regulation of the plasma in cephalopods (Robertson, 1949). The fluid from the renal sac of the male Sepia had a protein concentration of only about 1 g./l. compared with 109 g./l. in the plasma. Moreover, it showed wide differences from the plasma or a plasma ultrafiltrate in all ions except chloride (Table 5). These differences are tending to lower the sulphate of the blood and raise the remaining potassium, calcium, magnesium and chloride ions, but only in the cases of potassium, chloride and sulphate are these tendencies in line with the composition of the blood. The surprising finding that the sodium of the fluid was only 79 % of the plasma sodium pointed to the presence of some other cation in order to balance the anions and make up the total concentration of ions to about that of the plasma. In a second Sepia the sodium of the renal sac fluid was 88 % of the plasma concentration, in a third 92 % ; the latter had received an injection of sucrose for other purposes.

The missing cation proved to be ammonium which was measured by a microdiffusion method (Conway, 1947). In the first specimen (Table 5) the amount was 2·64 mg. NH4/g. water, 146 m.equiv./kg. water, which added to the sum of the other cations gave a concentration of 613 m.equiv. compared with 609 m.equiv. for chloride and sulphate. Electro-neutrality was also satisfied in the third Sepia which had 1·06 mg. NH4/g. water, 58·7 m.equiv./kg. water, the total cations being 622 and anions 619 m.equiv. Figures for Sepia given by Delaunay (1931) are 1·12 and 1·19 mg. NH4/ml. in the ‘urine’.

In Eledone, however, the sodium concentration in the renal sac fluid was on the average 2 % higher than in the plasma, suggesting that the amount of ammonium in the fluid was negligible (Robertson, 1949). Delaunay (1931) gives two figures, 29 and 4 m.equiv./l. for Octopus vulgaris. The discrepancy between the Sepia and Eledone analyses may be connected with the different conditions of the animals when analysed. The Sepia were kept only 16-24 hr. after capture before analysis at Plymouth, and their high ammonium excretion may be considered to reflect recent feeding and breakdown of protein. Eledone specimens at Millport, on the other hand, were kept several days in the tanks without food, and their nitrogenous excretion had probably fallen to a low level.

An attempt was made to estimate the turnover in water and ions of a male Sepia. Fluid lost from the renal sacs must be balanced by uptake of water and ions from sea water, through the gills or via the gut. Isosmotic sucrose was injected into the anterior vena cava in front of the siphon, and 2 hr. later all the fluid in the renal sacs and samples of blood were collected. The sucrose in the two fluids was determined as the difference between total reducing substances before and after hydrolysis with an invertase preparation, Somogyi’s (1945) copper method being used for measurement of reducing power.

Weight of Sepia, 895 g.

Sucrose injected, 5 ml. of solution containing 332 g./kg. water, equivalent to 275 g./l. (A = 1·99° C.).

Renal fluid collected, 7·0 ml. (containing 6·86 g. water).

Glucose-fructose equivalent of sucrose: plasma, 5-i9mg./g. water; renal fluid,

2·37 mg./g. water (= 45·6% plasma value).

Total glucose-fructose of 7 ml. renal fluid = 2·37 × 6·86= 16·26 mg.

Volume of plasma water containing this amount = 16·26/5·19 = 3·133 ml.

This is equivalent to a volume of ‘urine’ of 3-133/0-98 = 3-20 ml. (factor 0-98 since i ml. urine contains 0-98 ml. water).

The accuracy of this assessment of fluid turnover in Sepia, 4·3 % of the body weight per day, depends on the correctness of a number of assumptions, namely, that sucrose is excreted by filtration only, that injection of 5 ml. fluid does not stimulate filtration, and that no sucrose-containing fluid escapes through the ureters before collection at the end of a 2 hr. period. The latter assumption especially is rather hazardous, and the figure of 4·3% may perhaps be regarded as a minimal one.

From this experiment an estimate can also be made of the total volume of extracellular fluid, taking sucrose as a substance which will not diffuse significantly into cells. The calculation is made by dividing the glucose-fructose equivalent of the 1·375 g sucrose injected by the plasma equivalent.

Volume of extracellular ⁠.(the factor of 1·05 depends on the fact that 1 g. sucrose gives 1·05 g. glucose-fructose on hydrolysis).

Percentage extracellular fluid = 278/895 × 100 = 31·1%. Correction for the facts that the 14 ml. vitreous humour from the eyes and the 7 ml. fluid from the renal sacs had respectively concentrations 2 and 46 % of the glucose-fructose level of the plasma increases this percentage to 33·0%. Unfortunately, it has been found that sucrose appears to enter muscle cells to some extent in the crustacean Nephrops, and this may well be the case in Sepia, making this volume of 33·0% an overestimate. In mammals sucrose is considered one of the best indicators of extracellular space (Wilde, 1945 ; Kruhoffer, 1946), but it appears to penetrate slowly amphibian muscle cells (Krogh & Lindberg, 1944).

Excretion of a fluid amounting to 4 % of the body weight per day implies a turnover of water equivalent to the volume of the extracellular fluid every 8 days. From what has been said about possible errors, 8 days is probably an overestimate.

The many structural similarities between the cephalopod and the vertebrate eye have long been regarded as an outstanding example of convergence in evolution. Certain non-sensory aspects of the cephalopod eye are investigated here. In both types of eye a fluid is present in front of the lens, the aqueous humour; behind it and in front of the retina in the vertebrate eye is the vitreous body, a gel, and in the cephalopod the liquid vitreous humour. Ophthalmic arteries supply the retina, ciliary body and iris. In mammals considerable attention has been given to the mechanism of formation of the aqueous humour, particularly the problem of how far it is an ultrafiltrate or dialysate of the plasma in the capillaries of the ciliary body. Varying conclusions have been reached, depending often on which ions or non-electrolytes have been considered. Apart from the chloride analyses of Derrien (1938), who found the concentrations of this ion in plasma and vitreous fluid to be the same, no data on the composition of the eye-fluids seem to exist for cephalopods. Analyses were therefore made of the plasma, vitreous and aqueous humours, the sea water with which the animals were in equilibrium, and samples of plasma dialysed against sea water.

Seven plasma-vitreous fluid comparisons are summarized in Table 6 in which are given the mean figures with mean deviations for three Sepia officinalis and three Loligo forbesi, together with a single analysis of Eledone cirrosa. A striking feature is the wide difference between the two fluids in practically every ion. Magnesium shows a remarkable decrease in the vitreous humour in the three genera, falling even to 10% of the level in the plasma in Sepia. Compensating increase of sodium ions seems to adjust the cation-anion balance and maintain the total ionic concentration which is within 1 % of that of the plasma. Chloride falls definitely in Sepia and Eledone. Most of the other ions in the vitreous humour are lower than in the plasma, except calcium in Sepia, potassium in Loligo, and potassium, calcium and sulphate in Eledone. In both Sepia and Loligo the mean deviations from the average figures are small.

Cation-anion balance, as determined by summing the principal ions, is satisfactory (Table 6). The deficiency of anions in the plasma, 22-28 m.equiv., is probably made up chiefly by protein, the concentration of which in g./kg. water was 122 ± 4 in Sepia, 173 ± 6 in Loligo, and 117 in Eledone. Heat-coagulable protein is practically absent from the clear vitreous humour, a value of only 0·24 g./kg. water being obtained in Sepia. The contribution of bicarbonate and phosphate to the anions of marine invertebrate body fluids is small. Several estimations of these were made, bicarbonate by Conway’s (1947) method, and inorganic phosphate according to Sumner (1944). In Eledone plasma bicarbonate was only 4 m.equiv., and the pH as determined by the glass-electrode 7·10 (15o C.). This was blood collected under oil from the dorsal aorta ; the pH of vitreous humour collected from the same specimen was 6·70. Inorganic phosphorus comparisons in plasma and vitreous fluid were Loligo—0·094 and 0·057 mg. P/ml., Sepia—0·053 0·035 mg. P/ml. Such figures represent 1-3 mg. ions and 2-5 m.equiv. H2PO4_ + HPO42-.

Taking into consideration the data in Tables 6 and 7, it is quite evident that the vitreous fluid is far from being a dialysate of the plasma in these three cephalopods. Apart from the fact that it is virtually a protein-free fluid of approximately similar osmotic pressure, it differs widely from a dialysate in respect of each of the principal ions and of inorganic phosphate. The equilibrium here between the plasma and vitreous fluid would seem to be a steady state in which all the ions are actively maintained, presumably by processes of secretion and absorption by the cells of the retina.

The aqueous humour in front of the lens exists in small quantities, and it has been examined only in Sepia. In this animal a small pore in the cornea allows communication between this fluid and the external environment (Tompsett, 1939). That the aqueous fluid is not sea water is at once apparent from the analyses of the eye-fluids from a Naples specimen shown in Table 6. Neither is it a dialysate of the plasma, but in most respects it is intermediate in composition between the vitreous fluid and sea water. Its composition cannot be explained simply as a mixture in certain proportions of a plasma ultrafiltrate and sea water; magnesium in such a case would be at a higher level. Diffusion between the aqueous and vitreous fluids is a possibility, but judging from the respective compositions of the two fluids must be very restricted. Two hypotheses may be suggested:

The aqueous fluid is a plasma filtrate modified by the secretion of extra Na+, Ca²⁺ and ions, and reabsorption of K+, Mg²⁺ and Clions by the epithelia lining the anterior chamber.

The aqueous fluid is chiefly sea water from which some Mg²⁺ and ions have been absorbed by the epithelia, and into which have diffused from the blood small quantities of other ions in accordance with the concentration gradient.

It might perhaps be argued against (1) that it is unlikely that cellular activity of the epithelia and retina in adjacent parts of one organ would produce fluids of such different composition as the aqueous and vitreous humours. The second hypothesis implies a steady state : sea water entering through the corneal pore would have to be removed continuously to prevent the concentrations of ions other than Mg²⁺ and from reaching levels characteristic of a dialysate of the plasma. Further work is necessary before the origin of the aqueous humour can be definitely established..

Certain deductions can be made about the forces concerned in maintaining the water balance of the vitreous humour. The colloid osmotic pressure due to the haemocyanin of the blood, and hydrostatic pressure may be important if the blood and vitreous humour have osmotic pressures similar to those of a Donnan equilibrium, although such an equilibrium in respect of each ion is not found. To prevent fluid being withdrawn by the plasma through the retina, the hydrostatic pressure in the retinal capillaries must either equal or exceed the colloid osmotic pressure. If the retinal cells are able, by transporting ions, to raise the concentration of osmotically active particles in the vitreous fluid slightly above that of the plasma, water would tend to pass across the retina from the capillaries, keeping up the turgor in the interior of the eye.

The data in Table 6 are not sufficiently exact to decide whether the vitreous humour is slightly hyperosmotic to the plasma. Such slight differences as exist are within the error of the summation method of determining total ionic concentration, which is about 1 % (3 × S.D.).

The further data on ionic regulation given in this paper confirm and amplify the previous findings of Robertson (1949). Animals of the echinoderm and lamelli-branch groups, Holothuria and the bivalves Mytilus and Ostrea, show little ionic regulation in their coelomic fluid or plasma, apart from the accumulation of potassium. Magnesium remains within 3 or 4% of the equilibrium value, while calcium sometimes exceeds this by a few per cent. Phascolosoma differs from the polychaetes in having the magnesium concentration of the coelomic fluid reduced to about 70%, and a corresponding increase of the sodium concentration to 104% is necessary to maintain the balance of cations and osmotic equilibrium. Calcium and magnesium are accumulated to a considerable extent in Archidoris, a feature distinguishing it from the lamellibranchs, but protein is still very low owing to the absence of haemocyanin in this gastropod.

High concentrations of protein and regulation of every ion characterizes Sepia and all the Crustacea examined. With respect to Mg²⁺ and ions, the eight crustaceans isosmotic with sea water show all possible combinations: high Mg²⁺ and ⁠, e.g. Hyas-, high Mg²⁺ and low —Dromia; reduced Mg²⁺ and ⁠, e.g. Squilla ; accumulation of with low Mg2+—Eupagurus. Whatever were the combinations, the concentration of Na+ ions was apparently altered in such a way as to maintain electroneutrality and total ionic concentration. The usual finding was a rise in Na+ ions to compensate for loss of Mg2+. Where ions were reduced and Mg²⁺ maintained (Dromia), Clions increased but Na+ ions fell to 97% of the dialysis value. This finding, at first surprising, is nevertheless in harmony with the maintenance of osmotic equilibrium, since Na+ ions paired with CH ions are more ‘active’ than when paired with ions.

The excretion of ammonium ions by the renal organs of Sepia in amounts sometimes as great as 146 m.equiv./kg. water or 24% of the total cation-equivalents presents another problem in osmotic regulation. It is met by a reduction in the other cations excreted, so that the total concentration of the ions in the fluid from the renal sacs is approximately equal to that of the plasma. In the specimen with 146 m.equiv. or mg. ions NH4, these concentrations were within f8% of each other. When NH4+ ions are excreted in quantity adjustment of osmotic pressure by diminished excretion of other cations falls mainly on the Na+ ion, since in mg. ions it constitutes about 84% of the total plasma cations.

The rate at which fluid is excreted by the renal organs in cephalopods would illustrate the importance of this excretion in regard to ionic regulation. In the literature the only data seem to be those of von Fürth (1900), who found the daily urine of a large Octopus of 4 kg. to be 80 ml., representing an excretion of 2% of the body weight per day. His method consisted in ligaturing the ureters, and this figure may be a minimal value as filtration might be expected to be reduced if pressure were to build up in the renal sacs. The figure of 4·3 % obtained in this paper for Sepia is somewhat uncertain, but may be of the right order of magnitude. This would amount to the excretion of a volume of fluid equal to that of the extracellular fluid (33 % of the body weight) within 8 days. A comparable calculation for the crab Car anus maenas with 35 % extracellular fluid and an output of 5 % of the body weight per day is 7 days (Webb, 1940).

The peculiarities in ionic composition of the vitreous humour of the eyes of Eledone, Loligo and Sepia demand the expenditure of energy in effecting and sustaining the marked differences in this fluid from a dialysate of the plasma. Every ion is maintained at a level different from what it would be if there were physicochemical equilibrium with the plasma across the retina. Magnesium concentrations in the vitreous fluid of 10-20% of those of the plasma on the other side of the retina are especially noteworthy, but what bearing this has on the sensory perception of this layer is unknown. A relative impermeability of the blood-vitreous fluid barrier is necessary for the maintenance of such differences of concentration, and this was found to be the case for sucrose injected into the plasma. Two hours later the concentration of sucrose in the vitreous humour was only 2% of that in the plasma.

Detailed comparison with the mammalian eye is not possible owing to the inadequacy of the chemical data for any one mammalian species (e.g. Davson, 1949; Duke-Elder & Goldsmith, 1951). It would seem, however, that concentrations of ions such as sodium, potassium and chloride in the aqueous humour and vitreous gel are only slightly different from those of a true dialysate of the plasma. There are indications that magnesium of the vitreous body may be considerably lower than the plasma magnesium, at least in the horse (Duke-Elder, 1929). The vitreous fluid of cephalopod eyes is far from being a true dialysate of the plasma, and the marked differences are presumably due to secretion and absorption of ions by the retina. In Sepia the aqueous humour has a composition quite unlike the vitreous humour or a plasma dialysate but more like sea water; its precise origin has not been established.

In Table 8 is given the relative ionic composition of the plasma or coelomic fluid of the animals which have been analysed, based on a chloride value of 100; three ionic ratios of equivalents have also been calculated. Several points may be emphasized: the general resemblance to sea water of the coelomic fluid of Holothuria and the plasma of the bivalves; the high potassium and low sulphate of Sepia-, the increase of sodium where magnesium is reduced in Phascolosoma and most of the crustaceans, with corresponding changes in the ionic ratios; the high sulphate values of Mytilus galloprovincialis and the two species of Eupagurus.

In a series of six crustaceans, it was found previously (Robertson, 1949) that the more active ones had low concentrations of magnesium in the blood. Further data have enabled Fig. 1 to be drawn, illustrating the values of magnesium in sea water and the plasma of sixteen crustaceans. Of these sixteen any one familiar with them would pick out without hesitation the two spider-crabs Maia and Hyas, the anomuran Lithodes, and Dromia as being the least active and those which respond most slowly to mechanical stimulation. It is just these four genera that have the highest magnesium values, 84-101 % of that in sea water. The two hermit-crabs are anomurans like Lithodes but more active, and they have correspondingly lower concentrations of magnesium. The remaining crabs (with one apparent exception), the lobsters and Squilla are much more active, and all have magnesium values less than half that in sea water. Cancer might be considered anomalous as in aquaria it usually shows little activity; but when displaced from a preferred comer, it can run quite rapidly back.

When it is recalled that excess of magnesium ions in the form of solutions of magnesium chloride or sulphate are used to narcotize marine animals (see e.g. Pantin, 1946), there seem grounds for assuming that reduction in magnesium and increased activity are causally related. From this standpoint Dromia, Lithodes and the spider-crabs might be considered as living in a semi-narcotized state. Evidence from the study of the isolated walking legs of Carcinus indicates that perfusion with a fluid containing 1·5-2 times the blood concentration of magnesium depresses neuromuscular transmission (Katz, 1936), and perfusing with a fluid containing only 5-20% of the blood concentration enhances the submaximal muscular response (Boardman & Collier, 1946). Findings that the mechanical response to nerve stimulation varies inversely with the magnesium concentration in the perfusing fluid have also been reported in three other decapods, Maia, Panulirus and Cambaras (Waterman, 1941).

However, other ions as well as magnesium influence activity, and the balance between certain ions is often important. Bethe (1929) found that the magnesium concentration of the blood of Carcinus increased to a value over three times the normal when the external magnesium was increased by about the same factor, resulting in the lessening of muscular tone and impairment of the normal reflexes. Exactly similar results were found when the calcium concentration of the blood was reduced; but crabs became very excitable when the plasma value of calcium was increased by keeping them in calcium-enriched sea water. Thus the balance between calcium and magnesium is probably important; the ratio of the equivalents of Ca/Mg ranges from 0·19-0·31 in the Lithodes-Dromia-spider-crab group to 0·39-2·0 in the remaining decapods and Squilla (Table 8; Table 9, Robertson, 1949). It may be recalled that anaesthesia produced by injection of Mg²⁺ ions in mammals is counteracted by injection of Ca²⁺ ions (Heilbrunn, 1947, p. 460).

Wide variations in organization and activity exist among members of the Mollusca, a phylum in which magnesium is maintained at a uniformly high level. Calcium in the blood does not vary very much, although it is usually slightly higher than the value in sea water. The most variable cation is potassium. Moderate increases in potassium ions have a stimulatory action on the neuromuscular system (e.g. Wells, 1928; Ross & Pantin, 1940). Bethe (1927) found that increase in potassium in the external medium had similar effects to increase in calcium on medusae, Phoronis worms and various crustaceans, these ions acting synergically in augmenting rhythmical movements. It is therefore suggested that the pre-eminence in muscular activity of members of the Cephalopoda compared with the Lamelli-branchia or Gastropoda, although based primarily on structural characters, may be at least enhanced by the very high potassium concentrations maintained in the blood by members of this class.

Undoubtedly factors other than ionic composition of the blood may influence activity, such as hormones. Florkin (1949) relates the torpor of Cancer and the nimbleness of Carcinus to differences in the blood-sugar content, high in the former and low in the latter (Roche & Dumazert, 1935 ; Florkin, 1937), but no supporting evidence is given for assuming a direct effect of the level of blood sugar on activity. Indeed, other authors (Gray & Hall, 1930) have found an opposite correlation in marine fishes, the most active having high blood sugars.

Certain general conclusions can be drawn from the present data and those of Robertson (1949), which consist of analyses of the mesogloea of a coelenterate and of the plasma or coelomic fluid of thirty-four marine invertebrates belonging to the Echinodermata, Annelida, Sipunculoidea, Mollusca and Arthropoda. All these animals, with the exception of the grapsoid crab Pachygrapsus, are in osmotic equilibrium with sea water, within 1-2%. To a varying degree ionic regulation exists in all the animals examined, ranging from a slight accumulation of potassium in the Echinodermata to regulation of every ion in the Cephalopoda and most of the Crustacea.

Ionic regulation is slight in the more simply organized, relatively inactive coelenterates, echinoderms, polychaetes, lamellibranchs and gastropods. All these, excepting the prosobranch gastropods, have very low protein contents in their body fluids (or mesogloea in Aurelia), usually below 1 g./l. Pronounced regulation is shown by the more active, highly organized Crustacea and Cephalopoda, all of which contain the respiratory pigment haemocyanin ; high concentrations of protein are characteristic, 29-80 g./l. in the decapods and 105-150 g./l. in the cephalopods.

Where there is ionic regulation of the blood, the equilibrium which exists across gills and other permeable membranes is a steady state in which the concentrations of ions are maintained actively. Replacement of these membranes by one of collodion results in a physico-chemical equilibrium in which Donnan forces come into play. The Donnan ratio for sodium, potassium, magnesium, chloride and sulphate does not exceed 1·03 in the decapod Crustacea and the cephalopod Mollusca. In these animals none of the cations is bound to protein except calcium, of which 10-20% exists as a calcium-protein complex.

Active regulation exists also in the fluids produced by the excretory organs of decapods and cephalopods, and in the eye-fluids of the latter.

The mechanism of ionic regulation of the plasma in decapods and cephalopods involves the continuous selective excretion of ions in the excretory fluids and controlled uptake of ions by the permeable surfaces.

The concentration of magnesium in the blood of decapod crustaceans is related to activity; those animals with high levels are slow-moving and inactive, whereas those with low levels are capable of quick movement and are generally more active. Cephalopods have high levels of potassium ions which may contribute to their powers of active movement.

Variations of ionic regulation in different specimens of a species are small in the few cases investigated. In general, potassium is the most variable ion ; sodium and chloride are the least variable, as is to be expected, since they form such a large proportion of the total concentration of ions which is maintained at the same level as the external medium in all the animals except one.

I am indebted to the Directors of the Stazione Zoológica, Naples, the Marine Laboratory, Plymouth, and the Marine Station, Millport, for the facilities which they so kindly granted me during my visits. I am grateful for permission to occupy the British Association Table at Naples. Some of the apparatus used in this work was purchased by means of a Royal Society grant from the Parliamentary Grant-in-aid for Scientific Investigations. To the Carnegie Trustees I am especially indebted for the Research Fellowship which enabled me to pursue this work.