previous sub-section
Chemistry of Sea Water
next sub-section

Elements Present in Sea Water

Thus far, only the major constituents of sea water have been considered. In table 36 are entered all elements that are known to occur in sea water as dissolved solids, except hydrogen and oxygen. They are not given as ions in this case but as the amounts of the individual elements which occur in water of chlorinity 19.00 ‰. The elements are arranged in the order of their abundance. In the first column they are reported as milligrams per kilogram, and in the second as milligram-atoms per liter. For convenience, the 1940 atomic weights and their reciprocals have been included. These constants are necessary when converting weight units to gram-atom units, and vice versa. The values for the major elements correspond to those given in previous tables and, in general, are valid for surface water. For many of the elements ranges in concentration have been indicated. No doubt ranges should be shown for others, but the lack of sufficient observations or uncertainty as to the reliability of reported data leaves these problems unsettled. For many of the elements that are present in very low concentrations there are only one or two determinations available, and in some cases only indirect estimates have been made. Hence, in these cases the indicated values can represent only the order of magnitude of the quantities present. Omitting the six most abundant elements, only carbon (CO2 components), silicon,

nitrogen, and phosphorus compounds have been studied with sufficient completeness to provide a fairly good idea of their distribution. Less complete studies have been made on the variations in the amounts of boron, iodine, iron, manganese, copper, gold, and radium. Cadmium, chromium, cobalt, and tin have been found in the ash of marine organisms, and hence it is implied that they occur in sea water, although so far they have not been shown directly.

Forty-four elements are listed in table 36, and if we add hydrogen, oxygen, and the inert gases neon, helium, and argon, we obtain a total of forty-nine elements that are known to occur in sea water. Further investigations will undoubtedly demonstrate the presence of others. Certain problems of the origin and concentration of the dissolved solids relative to their concentration in the earth's crust will be discussed later.

The following brief discussion is limited to those elements that either occur in relatively large amounts or whose distribution has been shown to be affected by biological activity. For elements in the latter group additional data are given in chapter VII. In table 36 references are given for those elements not discussed in the text. A comprehensive discussion is given by Thompson and Robinson (1932), and other results are reported by Goldschmidt (1937) and Wattenberg (1938). The elements are considered in the order in which they appear in the table.

Chlorine, present as chloride ion, is the most abundant ion and makes up about 55 per cent by weight of the dissolved material. It is rarely measured except in combination with other halides in the chlorinity determination. The bromide and iodide are then computed as if they were chloride. It should be kept in mind that the ratio of the chlorine equivalent of the halides to the chlorinity is 1.00045 (p. 52). The chlorinity is of the greatest importance, not only as the basis of density computations, but also as the standard to which those substances present in major amounts are referred.

Sodium is the most abundant cation in sea water, but it is rarely determined directly, owing to the technical difficulties involved in the determination of the alkali metals. The average ratio to chlorinity, 0.5556, as obtained by Robinson and Knapman (1941) agrees exactly with the value that Lyman and Fleming (1940) calculated by difference. It is somewhat higher than the average of 0.5509 given by Thompson and Robinson (1932), but is in fair agreement with the ratio 0.5549 obtained by Webb (1939) by direct analysis. The sodium: chlorinity ratio may be modified near river mouths.

Magnesium content of sea water has been investigated rather carefully, particularly by Thompson and Wright (1930). The magnesium is usually determined by a special modification of the magnesium-ammonium-phosphate method. The ratio of magnesium to chlorinity is very uniform.

ELEMENTS PRESENT IN SOLUTION IN SEA WATER (Dissolved gases not included)
Element mg/kg Cl = 19.00 ‰ mg-atoms/L Cl = 19.00 ‰ Atomic weight (1940) 1/atomic weight Authority
Chlorine 18980 548.30 35.457 0.02820
Sodium 10561 470.15 22.997 0.04348
Magnesium 1272 53.57 24.32 0.04112
Sulphur 884 28.24 32.06 0.03119
Calcium 400 10.24 40.08 0.02495
Potassium 380 9.96 39.096 0.02558
Bromine 65 0.83 79.916 0.01251
Carbon 28 2.34 12.01 0.08326
Strontium 13 0.15 87.63 0.01141
Boron 4.6 0.43 10.82 0.09242
Silicon 0.02 –4.0 0.0007 –0.14 28.06 0.03564
Fluorine 1.4 0.07 19.00 0.05263
Nitrogen (comp.) 0.01 –0.7 0.001 –0.05 14.008 0.07139
Aluminum 0.5 0.02 26.97 0.03708
Rubidium 0.2 0.002 85.48 0.01170
Lithium 0.1 0.014 6.940 0.14409
Phosphorus 0.001–0.10 0.00003–0.003 30.98 0.03228
Barium 0.05 0.0004 137.36 0.00728
Iodine 0.05 0.0004 126.92 0.00788
Arsenic 0.01–0.02 0.00015–0.0003 74.91 0.01335
Iron 0.002–0.02 0.00003–0.0003 55.85 0.01791
Manganese 0.001–0.01 0.00002–0.0002 54.93 0.01820
Copper 0.001–0.01 0.00002–0.0002 63.57 0.01573
Zinc 0.005 0.00008 65.38 0.01530 Atkins (1936)
Lead 0.004 0.00002 207.21 0.00483 Boury (1938)
Selenium 0.004 0.00005 78.96 0.01266 Goldschmidt and Strock (1935)


0.002 0.00002 132.91 0.00752 Wattenberg (1938)
Uranium 0.0015 0.00001 238.07 0.00420 Föyn et al (1939)
Molybdenum 0.0005 0.000005 95.95 0.01042 Ernst and Hoermann (1936)
Thorium <0.0005 <0.000002 232.12 0.00431 Föyn et al (1939)
Cerium 0.0004 0.000003 140.13 0.00714 Goldschmidt (1937)
Silver 0.0003 0.000003 107.880 0.00927 Haber (1928)
Vanadium 0.0003 0.000006 50.95 0.01963 Ernst and Hoermann (1936)
Lanthanum 0.0003 0.000002 138.92 0.00720 Goldschmidt (1937)
Yttrium 0.0003 0.000003 88.92 0.01125 Goldschmidt (1937)
Nickel 0.0001 0.000002 58.69 0.01704 Ernst and Hoermann (1936)
Scandium 0.00004 0.0000009 45.10 0.02217 Goldschmidt (1937)
Mercury 0.00003 0.0000001 200.61 0.00498 Goldschmidt (1937)
Gold 0.000006 0.00000002 197.2 0.00507 Haber (1928)
Radium 0.2 – 3 × 10−10 0.8 – 12 × 10−13 226.05 0.00442 Evans, Kip, and Moberg (1938)
Cadmium Fox and Ramage (1931)
Chromium Webb (1937)
Cobalt Thompson and Robinson (1932)
Tin Thompson and Robinson (1932)


Sulphur is present in sea water as sulphate ion, and is in this form usually determined by precipitation as barium sulphate. An extensive study of the sulphate distribution has been made by Thompson, Johnston, and Wirth (1931). Under stagnant conditions occurring in certain isolated basins, and in and near bottom sediments, a part of the sulphate may be converted to sulphide ion. Considerable quantities of sulphide occur in the Black Sea and in certain Norwegian fjords, and its presence has been reported in many localities. The sulphate: chlorinity ratio may also be modified by dilution with river water, which is generally relatively high in sulphate. Processes of freezing and melting may possibly affect the relative concentration (p. 216).

Calcium is present in much smaller quantities than either sodium or magnesium, but its distribution in the ocean has been studied much more thoroughly, mainly because calcium is a major constituent of many skeletal remains found in marine sediments. By deposition of such remains calcium is permanently removed from the water, but this removal does not necessarily imply that the calcium concentration is decreasing, because a large supply is maintained by the river waters flowing into the sea. Detectable differences in the calcium: chlorinity ratio have been observed. In the Baltic, Gripenberg (1937a) has shown that the type of river water which has diluted the sea water can be determined from that ratio. Furthermore, Moberg and Revelle (1937) have demonstrated the existence of vertical differences in the calcium: chlorinity ratio which they attribute to the removal of calcium in the surface layers through biological activity. Interest in the concentration of calcium has also centered around the question of the solubility of calcium carbonate in sea water and the factors that control precipitation and solution. In certain areas calcium carbonate is apparently precipitated inorganically, and in other regions it apparently passes into solution. In addition to these problems, knowledge of the calcium concentration is important in an understanding of the carbon dioxide system in the sea, which will be discussed later. The quantity of calcium is usually determined by precipitation as the oxalate under carefully controlled conditions and subsequent titration with potassium permanganate. One such method has been described by Kirk and Moberg (1933).

Webb has pointed out that in this method for the estimation of calcium the strontium will be carried down, and hence the calcium figure will be too high by the equivalent amount of strontium. As the ratio calcium: strontium is apparently constant, Webb suggests that the “calcium” shall be taken to mean the calcium after the strontium and barium have been replaced by calcium. Since the barium is negligible in this case, the values of “calcium” will be given directly by volumetric methods, but when the quantities are determined by weighing, corrections

must be applied (Webb, 1938). Values cited in this discussion are for calcium alone and have been obtained by correcting the analytical data for the presence of strontium. The “calcium” Cl-ratio as defined by Webb and corresponding to the values of calcium and strontium in table 35 is 0.0214.

Potassium is the fourth most abundant cation and is present in amounts of only a few per cent of that of sodium. The potassium is rarely determined directly, but apparently it bears a very constant relationship to the chlorinity (Thompson and Robinson, 1932). However, the content of potassium may be modified by biological agencies, since some organisms, particularly the large algae, concentrate potassium to a marked degree. The ratio of the potassium to chlorinity may also be modified by dilution with river water. The potassium may react with the colloidal and clay particles brought to the sea by rivers and run-off, and consequently this agency may influence the ratio. Certain minerals formed on the sea bottom, such as glauconite, contain potassium.

Bromine shows a very constant ratio to the chlorinity and is apparently all present as bromide ion.

Discussion of the concentration of carbon in sea water is complicated by the fact that it occurs not only in the form of carbonic acid and its salts but also in appreciable amounts as a constituent of organic material, either living or dead. The detrital organic material may be either particulate or in solution. The solubility of carbon dioxide depends upon the temperature and salinity of the water, and exchange of carbon dioxide with the atmosphere takes place at the surface. Photosynthesis in the surface layers reduces the amount of carbon dioxide in the water, and respiration increases the concentration. Consequently, the quantities of carbon present as either free carbon dioxide, bicarbonate, or carbonate will show a considerable range. These problems will be discussed in the sections dealing with the carbon dioxide system in the sea. The quantity of carbon given in table 36 was calculated on the assumption that only bicarbonate ions were present. The organic carbon, which is probably of the order of 2 to 3 mg/L (0.15 to 0.25 mg-atoms/L), was not included. The methods by which the different carbon dioxide components and the particulate and dissolved organic carbon may be determined are discussed later.

Strontium has not been investigated in detail, as it is extremely difficult to determine quantitatively. In determinations of calcium by means of the oxalate precipitation, the strontium is carried down with the calcium, and consequently the ratio of calcium: chlorinity usually reported for sea water represents the calcium plus strontium reported as calcium. Strontium is a constituent of the calcareous skeletons of certain organisms.


Boron occurs in sea water in a surprisingly high concentration and bears a constant relationship to the chlorinity. Apparently it is present as undissociated boric acid. There has been considerable uncertainty as to the form in which boron occurs, but the method of determination is standardized against boric acid and the values can at least be expressed as equivalent to a certain concentration of boric acid. The determination of boric acid in sea water is based on titration with very dilute sodium hydroxide in the presence of mannitol. Methods have been described by Harding and Moberg (1934) and by Igelsrud, Thompson, and Zwicker (1938). The amount of boron present in sea water is of interest in the carbonate equilibria and in this connection will be discussed later. Boron is concentrated by certain marine organisms.

Silicon has been studied extensively because it is utilized by diatoms and other silica-secreting organisms. According to a tabulation by Thompson and Robinson (1932), the silicate-silicon varies by more than one hundredfold—namely, from 0.0007 to 0.11 mg-atoms/L (0.02 to 3.0 mg/L). Clowes (1938) found values slightly exceeding .14 mg-atoms/L (4.0 mg/L) in the deep waters of the Antarctic. Surface samples are usually low, owing to the development of silica-secreting organisms, but a progressive increase in silicate takes place with depth, which is ascribed to the dissolving of soluble silicates. However, there is always the possibility that the water contains silicon in some compound present in colloidal form. River water contains a high content of silicon, both in solution and as colloidal particles. Diatom and radiolarian oozes contain the siliceous remains of organisms that have developed near the surface and settled to the bottom after their death. Although siliceous deposits of organic origin cover large areas, most of the siliceous skeletal remains dissolve after the death of the organisms. Silicon present as soluble silicate is determined colorimetrically. The method has been described by Thompson and Houlton (1933) and by Wattenberg (1937). Because of the rapidity with which water samples are contaminated by silicate that dissolves from the glass, the analyses should be made soon after the water samples are collected. Waxed containers are sometimes recommended, and it is always desirable to use “aged” bottles that have been thoroughly leached with sea water. Tourky and Bangham (1936) tested the reaction between the molybdate reagent and colloidal silica and found that the color development was not proportional to the amount of silicon present. Treatment of the colloidal silica with alkali prior to analysis yielded correct values. Experiments with sea water indicated that colloidal silica may pass into true solution on ageing.

Fluorine is present in oceanic sea water in concentrations slightly above 1 mg/L. It is present as fluoride and, according to the work of Thompson and Taylor (1933), bears a constant ratio to the chlorinity.


The method of determination is described by these authors. Little is known concerning the role of fluorine in the sea.

Nitrogen occurs in sea water both in compounds of various kinds and as free dissolved nitrogen gas. As it is an essential constituent of living matter, nitrogen is found in organic compounds both in organisms and in particulate and dissolved organic material in amounts between 0.1 and 10.0 μg-atoms/L (p. 254). In addition, it is present as nitrate, nitrite, and ammonia. In routine observations only the inorganic nitrogen compounds are determined. Nitrate- and nitrite-nitrogen are determined colorimetrically, and the ammonia either colorimetrically (Robinson and Wirth, 1934) or by micro-titration after distillation (Krogh, 1934).

The nitrate method originally described by Harvey (1926) is given by Wattenberg (1937). Rakestraw (1936) and Wattenberg describe the procedure for the determination of nitrite. Since the inorganic nitrogen compounds are subject to change after the water samples have been collected, analyses must be run within a few hours. Even the addition of preservatives may not prevent changes in the NH3 and NO2, indicating that purely chemical transformations may be involved. Ammonia tends to disappear in storage, and nitrite sometimes decreases, but at other times shows an increase. The nitrate, which is more abundant, does not show such relatively large changes.

Because of their relatively low concentrations and their utilization by organisms, the inorganic nitrogen compounds show a wide range in values:

The distribution of nitrate in the oceans has been and is studied a great deal, as it may limit the production of phytoplankton when it is reduced to minimal quantities in the surface layers. Nitrate-nitrogen usually shows a subsurface maximum at a depth of several hundred meters. Nitrite nitrogen has a peculiar distribution and is generally found in a rather thin stratum in or above the thermocline. Lessi is known concerning the distribution of ammonia, as it is not so readly measured as the other inorganic compounds of nitrogen, but it is apparently rather uniform throughout the water column.

Nitrogen compounds are carried to the sea by rivers and by precipitation. The greater part of these are supposed to have been fixed by electrical discharges in the atmosphere. Possibly a certain amount of the fixed nitrogen in the sea is liberated as free nitrogen and returned to the atmosphere. Bottom sediments contain a small percentage of organic nitrogen in resistant organic detritus, and a part of this is

permanently lost from the water, as it is found in all types of sediments, both recent and fossil. As the carbon: nitrogen ratio in organic material is relatively constant, the organic nitrogen is frequently used as a measure of the amount of organic matter in marine sediments and also in the water. The distribution of nitrogen compounds and the nitrogen cycle in the sea are discussed in chapters VII and XVIII.

Aluminum is present in sea water in very small amounts. The colloidal clay particles which are carried to the sea contain a large percentage of aluminum, and hence analyses of water samples collected near shore may show the presence of aluminum, but it is not necessarily all in solution. The value given in table 36 is the average quantity reported by Haendler and Thompson (1939). Their values range between 0.006 and 0.065 mg-atoms/L (0.16 and 1.8 mg/L) with an average of 0.02 mg-atoms/L (0.54 mg/L).

Although earlier workers (Thompson and Robinson, 1932) were unable to detect rubidium in sea water, Goldschmidt (1937) has reported about 0.002 mg-atoms/L (0.2 mg/L).

Lithium content of sea water has been investigated by Thomas and Thompson (1933), who found 0.014 mg-atoms/L (0.1 mg/L).

Phosphorus, which is present in sea water as phosphate ions, is another of the essential constituents of living organisms, and its distribution in the sea is markedly affected by organic agencies. In addition to the nitrogen and silicon compounds, phosphate-phosphorus has been considered as one of the substances that may limit production of plant life. The inorganic phosphorus concentration varies from virtually zero at the surface, under certain conditions, to approximately 0.003 mg-atoms/L (0.090 mg/L) at subsurface levels when values are corrected for salt error. There is frequently a subsurface maximum similar to that in the distribution of nitrate-nitrogen. Phosphorus removed from the surface layers by phytoplankton is largely returned to solution on the death and decomposition of the organisms. It is supplied by rivers, and some is removed from the sea, as a small quantity is present in most marine sediments. In certain shallow areas, phosphatic concretions are found that contains a rather high concentration of phosphorus. The mode of origin of these concretions is not yet known. It has been suggested that in many regions the water is supersaturated in respect to tricalcium phosphate which, therefore, may be deposited inorganically (Dietz, Emery, and Shepard, 1942).

Phosphate phosphorous is determined colorimetrically. The method has been described by Robinson and Wirth (1935) and Wattenberg (1937). Cooper (1938a) has discussed the magnitude of the salt error. Phosphate analyses are frequently carried out as routine observations, and our knowledge of the distribution of phosphate in the ocean is fairly comprehensive. The rather scant knowledge we have concerning the

amount of phosphorus present as particulate or dissolved organic pbosphorus will be discussed in connection with the phosphorus cycle in the sea (chapter VII).

The amount of barium in sea water has been reported by Goldschmidt (1937) as 0.0004 mg-atoms/L (0.05 mg/L). This is lower than the values reported by Thompson and Robinson (1932). Barium occurs in marine organisms and it is a constituent of most marine sediments. In certain localities the deposits contain large amounts of barium sulphate in the form of concretions and nodules. The mode of formation of these structures is not yet understood.

The distribution and concentration of iodine in the sea has received a great deal of attention because of its important role in the physiology of man and terrestrial animals. Marine products are an important source of iodine-rich foods, The form in which iodine occurs in sea water is not yet clearly understood, but at least part of it is present as iodide and iodate. It is concentrated to a marked degree by marine plants, and for many years sea weeds have been used as a commercial source of iodine. The distribution and determination of iodine in sea water and marine organisms have been discussed by Closs (1931) and Reith (1930).

Arsenic content of sea water has been investigated by Rakestraw and Lutz (1933), who report values ranging from 0.15 to 0.3 μg-atoms/L (9 to 22 μg/L). This wide range is attributed to the fact that organisms may utilize arsenic in place of phosphorus. It is known to be a constituent of the tissues of many marine forms. The exact form in which arsenic occurs in sea water is not yet known.

Iron is an essential constituent of plants and has been considered as one of the substances that may limit the amount of plant production in the sea. Investigations show that at least part of the iron is not present in true solution, as it can be removed by ultrafiltration. Cooper (1937b) has pointed out that the amount of iron in true solution as ferric or ferrous salts is probably less than 2 μg/L, whereas the total iron present is generally about ten times this quantity. The amount present in the plankton may be as much as 16 per cent of the total iron of the water. Harvey (1937) considers that diatoms are able to adsorb and utilize colloidal iron. Iron is brought to the sea in relatively large quantities in the colloidal clay particles, and consequently considerable amounts of iron are found in the marine sediments. In many instances the iron content of the sediments is even higher than should be expected, indicating addition of iron through physical, chemical, or organic agencies. In inshore areas near the source of supply the total iron content of the water is sometimes much higher than that found in the open ocean. Methods for the determination of iron in sea water in its various forms have been described by Thompson and Bremner (1935a and b), Cooper (1935), and Rakestraw, Mahnke, and Beach (1936).


Manganese is apparently subject to concentration by marine organisms. Thompson and Wilson (1935) have reported values between 0.02 and 0.2 mg-atoms/L (1 and 10 mg/L). The value cited by Goldschmidt (1937) is 4 mg/L. Interest in manganese has been aroused by the occurrence of manganese nodules which are widely distributed in certain types of marine sediments, particularly in the Pacific Ocean.

The quantity of copper present in sea water probably lies between 0.02 and 0.2 μg-atoms/L (1 to 10 μg/L) (Marks, 1938, Wattenberg, 1938). Copper is an essential constituent of many marine organisms and it is also considered a factor in the life history of oysters, as a relatively high copper content of the water is apparently necessary for proper development of the larvae.

Much interest is attached to the content of radioactive elements in sea water, because deep-sea sediments are high in radium, compared to igneous rocks, and it is considered that the enrichment must be due to precipitation from the water of radium or its precursors. The radium content of sea water has been studied by many investigators, using various techniques, but it is only recently that methods have been sufficiently refined to yield trustworthy results. Studies by Evans, Kip, and Moberg (1938) and by Pettersson and Rona (Föyn et al, 1939) show that the radium content, measured by the radon emanation technique, varies between about 0.2 and 3.0 × 10−13 ‰ in sea water of salinity approximately 35 ‰. The low values are found in the surface layers, and it is suggested that organisms are responsible for a selective removal of this element. Both groups of workers found that organisms concentrate the radium about one hundredfold in their soft tissues. Calcareous structures show an increase in the radium:calcium ratio over that in the water. The maximum value listed above—namely 3.0 × 10−13 ‰— was found in water in contact with the sediments (Evans, Kip, and Moberg, 1938), and generally the radium content of the deeper waters is about 1 × 10−13 ‰.

Pettersson and co-workers (Föyn et al, 1939) have emphasized the importance of searching for the radioactive precursors of radium, as this element has the relatively short half-life period of only 1690 years. Of these elements uranium and ionium are probably the most; important, but thus far only uranium has been examined. Karlik (Föyn et al 1939) has analyzed a number of samples from various parts of the oceans and obtained for oceanic water a mean value of 1.5 × 10−6 ‰. Surface waters have a somewhat lower content than those from greater depths, but Karlik does not consider that the data are sufficiently adequate to show any differential removal. Studies of the dilute waters of the Baltic Sea showed that the uranium content was a function of the salinity.

Föyn and Rona (Föyn et al) have sought for thorium in sea water, but have been unable to detect it by the most refined methods. By examining

very large samples they have fixed an upper limit of 0.5 × 10−6 ‰ for this element. Older and apparently less accurate methods yielded considerably higher values.

The radium content of marine sediments and the theories concerning the deposition of radium and its precursors are discussed in the chapter on marine sedimentation.

previous sub-section
Chemistry of Sea Water
next sub-section