OCEANIC SALT DEPOSITS By F. C . F’HILLIPS M.A. Ph.D. (GEORGE HERDMAN PROFESSOR OF GEOLOUY UNIVERSITY OF LIVERPOOL) THE rocks classified by the geologist as sedimentary are formed by the deposition from transporting agents of the products of disintegration and chemical decomposition of previously-existing rock masses. Material carried in suspension by rivers is laid down to form clastic sediments- sandstones clays etc.-but many of the products of chemical weathering pass into solution and are carried down int,o lakes or eventually to the sea. Most of these dissolved product>s are only recovered when laid down as chemical deposit’s as a consequence of tlic evaporation of the aqueous solvent. Of such sediments produced by evaporation the oceanic salt deposits or mariue evaporites are geologically and chemically by far the most important.Though the exact circunistnnces have been and to some extent still are a matter for disputation it has long been realiscd by geologists that at several periods during the course of past geological ages large bodies of sea-water have been evaporated sufficiently far to cause at least partial crystallisation of the dissolved salts. The waters of the present-day oceans the hydrosphere in volume about 3 x 108 cubic miles have been found to display a remarkable degree of constancy in the relative proportions of the more important ions present though the absolute concentration may vary to some extent. Early analytical work by Forchhammer and others was superseded by the deter- minations made by W. Dittmar on 77 samples of ocean waters collected during the “ Challenger ” expedition ; it is a tribute to Dittmar’s work that later investigations have effected few important modifications of the values which he gave.The complexity of the solution and the difficulty of separation of certain related substances present awkward analytical problems. Even the total salt content cannot be accurately determined by direct evaporation for it is difficult to drive off all traces of moisture without loss of other constituents. For these reasons oceanographers use two defined quantities in the discussion of the chemistry of sea-water. The “ chlorinity ” is determined by precipitation of the halogens with a silver salt and is essentially the chlorine equivalent it being assumed that the bromine and iodine have been replaced by chlorine.(For more precise discussion chlorinity has been redefined 2 in terms of the weight of silver precipitated and hence is independent of any changes in accepted values of atomic weights.) The “ salinity ” is also a defined quantity slightly less than the total weight of dissolved solids and can be calculated from 1 Reports of “ Challenger ” Expedition Physics and Chemistry 1884 1 1-251. a J. Jacobsen and M. Knudsen Assoc. d’oceanog. phys. Union geodes. geophya. intern. 1940 Publ. Sci. No. 7. 91 92 QUARTERLY REVIEWS the chlorinity or determined from a measurement of the density. Both chlorinity and salinity are customarily expressed as g. per kg. of sea-water using the symbol o/oo (per mille). of the proportions of the major constituents of sea-water for a chlorinity of 190/,, is given below A recent tabulation Na+ Mg++ Ca+ + Kf Sr+ + I Ion.j c1 = 19~oo~/o,. 1 %. 11 Ion. j c1 = 19~ooo/oo. I %. 10.556 1.272 0.400 0.380 0.013 I I 30.6 1 3.69 1-16 1.10 0.04 c1- HCO,- Br - F- so,= ' 0.026 Total 34.482 I /1 The waters of the open oceans thus contain about 34 parts per thousand of dissolved salts of which about 55% by weight is chlorine and 31% sodium. The salinity varies only slightly except near land masses where it may be greatly reduced by the influx of river-water. We shall have occasion later to discuss further whether the salinity of sea-water has remained sensibly constant for long geological periods ; it is interesting to note here that the contributions received a t the present time from river-waters are of quite a different composition from the salt content of the oceans Ca++ constitut- ing about 20% and CO,= about 35% but much of this material is soon abstracted by animals or plants and involved in a biological cycle.One of the earliest attempts to study the order of separation of salts on the evaporation of sea-water was made by J. U~iglio,~ who nearly one hundred years ago carried out a long series of experiments by evaporating samples of water taken from the Mediterranean. He was able to demon- strate a general succession calcium carbonate and sulphate being followed by sodium chloride and the sulphates and chlorides of magnesium and potassium. The problem however was much too complex for such a direct analytical approach to be successful and the scientific study of the crystallisation of oceanic salts really began with the classic studies of J.H. van't Hoff and his associates. These studies were directed towards obtain- ing a closer understanding of the conditions of formation of the salt deposits of the Stassfurt region a t that time the most important potash-producing area in the world. Instead of experimenting as did Usiglio with actual sea-water van't Hoe undertook a systematic study of the solubility relationships of all the salts in question. By working initially under atmospheric pressure and at a single defined temperature (25") solubilities were determined first in pure water and then in the presence of other salts full precautions always being taken to maintain saturation by the presence H. U. Sverdrup M. W. Johnson and R. H. Fleming " The Oceans," New York Ann. C'him. Phys. 1849 27 92-107 172-191.1942 p. 166. PHILLIPS OCEANIU SALT DEPOSITS 93 of the solids as “ Bodensalze,” and to ensure the attainment of equilibrium -the slowness of attainment of equilibrium in some instances was one of the reasons why the early experiments of Usiglio couId never have been completely successful. A second series of investigations was carried out at a temperature of 83” rtnd other particular temperatures corresponding to the appearance or disappearance of individual compounds were deter- mined. “ It is hard to say how far such researches could have been carried by chemists unacquainted with the phase rule. They would have had no guide to the apparently chaotic results obtained on evaporating mixed salt solutions. The meager results obtained before the problem was taken up by van’t HOE show that little progress could have been made by the older methods of experimentation in which the results rtnd the guiding principles of modern physical chemistry had no place.” A long series of separate contributions were conveniently summarised in extended accounts dealing first with the chlorides and sulphates of sodium potassium and magnesium and secondly with the calcium salts and boron compounds.Further work by later investigators notably J. d’Ans,6 has established in detail the conditions of equilibrium in the systems involving the chlorides and sulphates of sodium potassium magnesium and calcium over a tempera- ture range of 0-120”. The necessary preliminary work on the simpler systems whilst having no direct application t o the complex problem of the natural deposits has proved of immense importance in providing the data on which to base the subsequent extraction and refinement of the vayious salts.The following table lists the more important salts which may occur in marine evaporites Chlorides and sulphates of sodiuni potassium and magnesium Halite (rock-salt) . . NaCl Sylvine . . KC1 Bischofite . . MgC12,6H20 Carnallite . . KCl,MgCI,,6H2O Thenardite . . Na2S0 Mirabilite (Glaubersalz) . . Na2S04 10H20 Glaserite (aphthitalite) . . (K,Na),SO Kieserite . . MgSO,,H,O Hexahydrite . . ruZgS0,,6Hz0 Reichardtite (epsomite Bittersalz) . MgS0,,7H,O Vanthoffite . . 3Na,SO,,MgSO Astrakanite (bloedito) . . NazS0,,MgS04,4H,0 Loewite . . 2Na2S0 , 2MgSO ,,6H,O Langbeinite . . K2S0,,2MgS0 Leonite . . K,S0,,MgS0,,4H20 Schoenite (picromerite) .. K2S0,,MgS0,,6H,0 Kainite . . KC1,MgS04,3H,0 Qe W. A. Gale Ind. Eng. Chem. 1938 80 867. 5 J. H. vm’t Hoff “ Die Bildung ozeanischer Salzablagerungen,” I and 11 Bruns- wick 1905 1909 ; “ Untersuchungen uber die Bildungsverhiiltnisse ozeanhher Salz- ablagerungen,” Leipzig 19 12. 6 “ Die Losungt3gleichgewichte der Systeme der Salze ozeanischer Salzablagerungen,” Berlin 1933. 94 QUARTERLY REVIEWS Calcium salts Anhydrite . . CaS04 Gypsum. - . CaS0,,2H,O Glauberite . . CaSO,,Na,SO Syngenite - . CaSO4,K2SO4,H,O Polyhalite . . 2CaS0,,MgS04,K,S0,,2H,0 Tachhydrite . . CaC1,,2MgCl2,12H2O Boron compounds Boracite . . 5MgO7MgC1,,7B,O Pinnoite . . Mg0,B20,,3Hz0 Ascharite . . 2MgO,B,O,,H,O Sulphoborite . . 2MgS04,4HMgB0 7H,O Lueneburgite . 3MgO,B,O ,,P206 8H20 Iron salts Rhmeitc . . NaC1,3KCl,FeC12 Douglasite .. 2KC1,FeC12,2H,0 Ccrtairi of these names are applied by systematic mineralogists only to massive material of the kind usually found in marine evaporites other names (some of which are given above as alternatives) being applied to well-developed crystals but we shall find it convenient to continue to use those names which are well established in the literature of oceanic salt deposits. It is a fortunate circumstance that amongst so many possible solid phases there is only a very limited degree of isomorphous replace- ment ; glaserite shows a limited substitution of potassium by sodium and very small amounts of sodium may be taken up by leonite and of potassium by astrakanite. Upon evaporation of sea-water the first salt to separate is calcium carbonate. The surface layers of ocean waters may in fact be considerably supersaturated with calcium carbonate,’ and in some regions of shallow water .such as the Great Bahama Bank where the shallowness reduces circulation precipitation of carbonate is already taking place a t the present time.As evaporation proceeds it is probable that dolomite may be directly precipitated though it must be admitted that the conditions of formation of dolomite are not yet fully understood and it is often not possible to dcterniine how far a particular dolomite may have been produced by t’he later alteration of a limestone by saline waters rich in magnesium. The actual quantities involved are small-from a 1000-m. depth of water a limestone only a few cm. thick would be formed if there were no considerablc degree of supersaturation.The conditions determining the separation of gypsum or of anhydrite were studied by van’t Hoff but it has recently been pointed out 8 that his determinations were based on the incorrect assumption that gypsum dissociates directly to form anhydrite. Actually the hemihydrate CaSO,,+H,O is always formed first and gypsum and anhydrite only coexist at a four-phase equilibrium point. In saturated aqueous solutions of gypsum and anhydrite this point lies at * E. Posnjak Amer. J . Sci. 1938 35 A 247-272. As evaporation proceeds calcium sulphate appears. H. Wattenberg Portschr. Min. 1936 20 192. PHILLIPS 00EANIO SALT DEPOSITS 95 a temperature of 42'. Investigating the efFect on this transition temperature of ealt solutions of approximately the composition of sea-water E.Posnjak 9 showed that a t a temperature of 30" gypsum will begin to separate when the salinity has been increaaed by evaporation to 3.35 times the normal value and that nearly one-half of the total amount of calcium sulphate present will be depoaited as gypsum before the concentration is reached at which anhydrite becomes stable. As evaporation is continued further halite eventually separates when the water content has been reduced to less than one-tenth of the original and anhydrite and halite continue to separate together until the field of stability of polyhalite is reached. The degree of evaporation neceasary to precipitate gypsum or anhydrite and halite has been attained at various periods in past geological ages and important deposits of rock-salt are worked in many parts of the world.In this country the aalt industry of Cheshire and Durham is based on occurrences in the Permian and Triassic rocks and the associated gypsum and anhydrite are also extensively worked but practically no potassium salts have been found though polyhalite has recently been recorded. lo Only when the evaporating body of sea-water has been reduced to 1.57% of the original volume do the salts of magnesium and potassium begin to appear. Such a high degree of evaporation has only infrequently been reached and the Permian deposits of Germany still remain the best-known example of natural potash resources. The earliest workings were around Stassfurt after which the deposits are still popularly known but subsequent exploration involving the sinking of over two hundred shafts and some thousands of boreholes revealed the wide extent of the deposits both northwards towards Hanover and south of the Harz on both flanks of the Thuringer Wald.The Werra district between Eisenach and Fulda has recently become the most important producer. National requirements for potash during two world wars induced extensive exploration in other countries. In the United States of America isolated records of potassium salts had been known from the Permian Salt Basin of the Texas-New Mexico area since 1912 and after an intensive programme of exploration the &st shaft was sunk in 1929 and production begun in 1931." In the U.S.S.R. deposits closely resembling those of Germany were discovered in 1916 in the region around Solikamsk in the province of Perm and less important deposits are known also in Alsace Poland and Spain.The detailed study of the further course of crystallisation in the complete system (Na,K,Mg,Ca),(Cl,SO,) presents considerable practical difficulties inherent in the diagrammatic representation of a system which with water has six independent components. The concentration of calcium at this stage is so low that it may conveniently be leff out of account for the present but we still have to deal with a quinary system. Since further crystallisation occurs always in the presence of excess of sodium chloride which forms no binary or ternary compounds with any of the other salts a diagrammatic @ Ibid. 1940 238 659-668. lo C. E. Tilley Min. Mag. 1943 H lvii. 11 J. W. Turrentine " Potaeh in North America," Reinhold 1943 p. 27. 96 QUARTERLY REVIEWS representation can be effected in terms of MgCl, Na,SO, and KCI and van't Hoff constructed various types of isothermal model.He also made extensive use in his discussions of plane " paragenetic diagrams " which showed the combinations of salts in equilibrium with halite and saturated solutions a t various temperatures but possessed no quantitative significance. In devising further simplifications E. Janecke has been particularly active. Van't Hoff expressed his results in terms of the solubility of the various salts in a constant amount of water but Janecke proposed an alter- native method with a variable water content expressing the amount required to produce a saturated'solution. The water content is thus treated separ- ately from the composition of the salt mixtures so that any inaccuracies in determination of the water content are not carried over t o statements about the salt contents.The device also leads t o a simple graphical pro- cedure. Expressing the composition in the form mH,O,sMg,yK,,( 100 - z - y)SO,,sNa, since only neutral salts are in question all possible compositions can be plotted on a triangular diagram in terms of K, Mg and SO, by using the values of x and y and an isothermal representation in a plane diagram is thus achieved. I n such a diagram the paths of crystallisation can be traced as in a simple ternary system the usual relationships of congruent and incongruent fields still hold and quantitative information can be derived by application of the centre-of-gravity principle. For a given temperature a further quantity such as the water content m or the associated amount of sodium chloride can be plotted as ordinate whilst if the various isothermal diagrams are set above each other the equilibrium conditions throughout a range of temperatu-es can be conveniently displayed in a triangular prism.Certain features o?. the courses of crystallisation can be studied also in projections on a face of the prism. Fig. 1 illustrates such a prism constructed by Janecke l 3 in the form of a wire model. Within the prism between 0" and 120° there are 33 invariant points at which four salts (in addition to halite) co-exist in equilibrium with solution. *4mongst these three important types may be distinguished. At the first of the type S + S + S + S4 + Solution a mixture of three salts gives way on rise of temperature to the appearance of a new solid phase.As examples may be quoted the reactions Mirabilite + Reicliardtittt f - Schoenite + Astrakanite + Solution marking the appearance of astrakanite a t 4-5" and Mirabilite + Astrakanite + Glaserite + Thenardite + Solution corresponding to the formation of thenardite above 13.5" l2 See '' Handbuch der Mineralchemie," C. Doelter and H. Leitmeier 1929 IV 2 l 3 2. Elektrochent. 1934 40 741. 86-91 1253-1258 (a convenient summary of numerous earlier papers). PHILLTPS OCEANIC SALT DEPOSITS FIG. 1 Solubility of oceanic salts between 0" and 120' (Junecke). Many of the invariant points are of the type S + S + S + S4 + Solution Carnallite + Kainite t Kieserite f Sylvine + Solution indicating a change of paragenesis. is one example of this kind taking place at a temperature of 72".The important reaction 97 98 QUARTERLY REMEWS A third type 4 + S + S + S + Solution marks the disappearance of a particular solid phase a t higher temperatures as with schoenite at 26" and kainite a t 83". Fig. 2 shows the temperature ranges of formation between 0" and lOO" of the various salts in the presence of sodium chloride. It will be seen that whilst certain salts can crystallise throughout this range others have a much more limited field the highly hydrated salts giving place t o less hydrated or t o anhydrous compounds a t higher tem- peratures. Van't Hoff distinguished three stages ; below 37" schoenite reichardtite and hexahydrite disappear between 37" and 55" loewite langbeinite and vanthoffite appear whilst above 55" astrakanite leonite and kainite in succession cease t o form.It should perhaps be emphasised that Fig 2 refers only t o temperatures of formation of the various salts FIG. 2 Temperature ranges of formation of oceanic salts. within the system under consideration ; the salts themselves are stable under suitable conditions over much wider ranges and langbeinite for example will remain unchanged indefinitely in a dry atmosphere at room temperatures. If we turn next to the course of crystallisation of a solution of the composition of normal sea-water the possibilities are much restricted. For isothermal evaporation at 25" the point marking this composition falls just within the field of aatrakanite (Fig. 3) which is therefore the first salt to crystallise. The path of crystallisation following a straight line directly away from the point marking the composition of this salt soon reaches the boundary of the reichardtite field.Reichardtite begins to separate and the further course depends upon whether the previously- PRILLIPS OCEANIC SALT DEPOSITS 99 separated astrakaaite remains in contact with the solution. If it does it will be resorbed as the reicherdtite continues to crystallise ; under natural conditions however it is likely that the salts already separated will become crusted over and thus be protected from the action of the solution. In either event the path of crystallisation eventually passes across the field of reichardtite to its boundary with that of kainite. From here the path lies along the boundary with the two salts crystallising together. There follow in succession the pairs hexahydrite-kainite kainite-kieserite kieserite-crtrnallite and finally the three salts kieserite-carnallite-bischofite Hexahgdrite Reichwdtite FIG.3 Stability $fields of oceanic salte at 25'. separating together until the evaporation is complete; in this mixture of salts the bischofite is of coume greatly predominant in amount since the composition is now so close to the Mg vertex. Throughout this later part of the crystallisration halite is still separating together with small further amounts of calcium salts. The fields of stability of these latter can be delineated in a similar Mg K, SO triangle and in many of the published figures the two isothermal diagrams are superposed (me for example Fig. 1). The actual amount of calcium present is so small that its effect on the course of crystallisation is negligible and the path of crystallisation outlined above can be applied directly to the double scheme.100 QUARTERLY REVIEWS The polyhalite which is stable when the crystallisation of magnesium salts begins gives way finally to anhydrite during the course of separation of kainite. We can thus construct a theoretical profile of the salt succession to be expected from this course of crystallisation at 25" Sylvine with kieserite and I halite Kieserite carnallite bischofite Kieserite carnallite Kieserite kainite Hexa.hydrite kainite Reichardtite kainite Reichardtite Astrakanite Polyhalite Anhydri te Gypsum Carbonates Halite with kieserite and Halite with kieserite and carnalli te sylvine h.2 a ___- Halite with langbeinite Halite with loewite Halite with vanthoffite Bischofite zone Carnallite zone Kainite zone magnesium sulphate zone Polyhalite zone Anhydrite zone Gypsum zone Basal limestone and dolomite I A comparison of this theoretical profile with the successions in natural deposits reveals in most areas a general correspondence a t least up to a certain stage.Marine limestones or dolomites passing up into anhydrite and halite with or without polyhalite are developed for example in Germany in the Texas-New Mexico field and in this country. The char- acteristic calcium sulphate of most marine evaporites is anhydrite however rather than gypsum. In the succeeding zones less complete agreement with the theoretical profile is revealed and in many areas there is profound disagreement. Salts such as astrakanite reichardtite kainite or hexa- hydrite are rare or absent ; vanthoffite loewite langbeinite and sylvine are found instead.For parts of the German field a succession can be tabulated l4 i Carnallite succession. ~ Hartsalz succession. i j Older potash beds Transition beds I I I I Halite with polyhalite Older rock-salt Halite with glauberite Halite with anhydrite Sylvine which does not appear in the theoretical profile a t 25" (in which kainite is the chief carrier of potassium) is a constituent of two of the l4 E. Fulde " Zechstein," Berlin 1935 47 139. PHILLEE'S OCEANIC SALT DEPOSITS 101 most important ores in most areas-the " Hartsalz " of the German miners a mixture of sylvine with kieserite and halite and the rich " Sylvinite," a mixture of sylvine with halite. The latter is of chief economic importance in the Texas-New Mexico district,15 in the upper parts of the German succession and in the Solikamsk region.Whether a primary bischofite layer was ever developed is a question which will be referred to later; certainly nothing resembling the thick bischofite zone of the theoretical profile has ever been encountered but its absence can be readily accounted FIG.. 4 Stability fields of oceanic salts at 55". for on the supposition that in most natural occurrences the evaporation of the mother liquor did not proceed to completion. The first suggestion to arise in an attempt to explain these discrepancies may be that the evaporation took place a t higher temperatures and it was to pursue this suggestion that van't Hoff carried out a second series of experiments a t 83" a t which the field of kainite disappears.In Fig. 4 is reproduced the isothermal diagram for a temperature of 55"; loewite is the first salt to separate from normal sea-water at this temperature and langbeinite also is found in the succession. Sylvine associated with kieserite would need still higher temperatures since below 72" the field of kainite intervenes. Such high original temperatures appear most unlikely t o most 16.J. W. Turrentine "Potash in North America," Reinhold 1943 p. 24. 102 QUARTERLY REVIEWS present-day students of the marine evaporites. Many lines of evidence such as the characteristic association with “ red beds,” indicate that most oceanic salt deposits of past ages were laid down in an evaporating basin under am arid continental climate but E.Fulda is almost alone amongst recent authorities in believing that as the rapidity of evaporation decreased with increasing concentration the temperature may have risen sufficiently high to allow the direct crystallisation of these higher-temperature associa- tions. Much of his evidence in support of the contention that the present- day profiles are essentially primary is of a geological character but reference may be made here to experiments by S. Lowengart 17 on the evaporation of water from the Dead Sea. When the solution had reached a density of 1.35 evaporation came almost to a standstill whilst irradiation served merely to effect a rise of temperature. Some modifications of the normal profile might be expected if other conditions during the evaporation apart from the temperature were different from those which we have assumed.It is possible that extensive resorption of earlier-formed salts may have occurred a t a later stage of crystallisation if they remained as a porous mass permeable by the solution. During the final stages only kieserite carnallite bischofite halite and anhydrite are in equilibrium together and any earlier products now brought into contact with the magnesium-rich liquor will be resorbed and pseudomorphed. It is also possible that it layering of solutions of different concentration may have arisen in the containing basin and the crystalline products separating from the more concentrated but hotter surface layer may have been resorbed on settling through tho underlying layers of differing concentration. A further effect of the setting up of currents of solutions of different con- centrations has been specially stressed by H.Borchert.18 In the course of an investigation of the reasons for the impoverishment of salt-deposits by lateral passage into less rich ores he has developed a “ dynamicrpoly- thermal ” study in contrast with the purely static considerations of van’t Hoff and d’Ans. Such features as the configuration of the floor of the basin and different rates of evaporation in different areas may set up tem- perature gradients and circulating currents of solutions of different con- centrations. Those compounds which tend to separate in the regions of higher temperature in such a circulatory system are termed thermophile whilst the cryophile salts will be precipitated in the colder regions. The fields gf formation of some salts under these conditions are considerably modified in comparison with the static system.Sylvine and langbeinite for example show much wider possibilities of formation; a t high tem- peratures the langbeinite area is considerably extended towards the Mg apex and a langbeinite-carnallite paragenesis may then be possible. It has so far been assumed that the deposition of marine evaporites of past geological periods must be explained in terms of %he evaporation of l6 2. deut. geol. Qes. 1924 76 ; Monatsber. 7-30. l7 2. pr. Qeol. 1928 36 86-89. la Kali 1933 27 97-100 105-111 124-127 139-141 148-150 ; 1934 28 290-296 301-305; 1935 29 1-5; Arch. Lagerstforsch. 1940 No. 67. PHILLIPS OCEANIC SALT DEPOSITS 103 a solution similar in composition to present-day sea-water.It is of course possible that the composition may have been very different and the question of the origin of the salt content of the sea is an interesting geochemical problem. On the assumption that the sea was originally fresh and has gradually acquired its present content of salts from the contributions brought down to it by- rivers geologists have in the past even attempfed to derive an estimate of geological time. The salt content of rivers however must have been derived from the chemical weathering of previously-existing rocks and it is particularly difficult to accept that the high chlorine content of the sea can have been produced in this way. V. M. Goldschmidt Is has examined the availability of a large number of elements in the earth’s crust. It would appear that the concentrations of sulphur chlorine bromine and boron in sea-water are such that these elements must already have been present in the primeval ocean.2o E.J. Conway 21 has recently discussed the probable course of the chemical evolution of the ocean both on the hypothesis that the halogen content was derived from the original atmo- sphere and also on the alternative assumption that all the halogen has come from volcanic sources. The probable differences of salt content between the Permian oceans and those of the present day thus revealed are quite insufficient to effect any radical change in the course of crystallisation which we have traced. Some salt deposits are believed to have been laid down from solutions which derived their salt content not directly from the ocean but by re-solu- tion of previously-existing marine evaporites.The conditions of deposition then approach those of non-marine evaporites the formation of which can be studied at the present day in natural salt lakes. In contrast with the uniform composition of oceanic waters the waters of such lakes show a wide variety of chemical characteristics. In particular it may be suspected that beds of rock-salt without an associated layer of gypsum or anhydrite below them are of secondary origin. The discrepancies between the theoretical profile and the natural succes- sions are not merely of a qualitative kind. A difficult quantitative problem is encountered in the vast thicknesses of gypsum anhydrite or halite often recorded. A bed of halite 100 m. thick would correspond to the evaporation of a column of water some thousands of metres in depth whilst thicknesses of anhydrite have been recorded which would involve the evaporation of an appreciable fraction of the whole present hydrosphere.Even if these large original volumes in the evaporating basin could be accepted there should be evidence of marked shrinkage of the ocean as evaporation proceeded; in actual fact the beds of anhydrite and halite maintain their thicknesa close to the margins of the area of deposition. These difficulties find at least a partial explanation in the accepted picture of the conditions under which evaporation in the basin took place. Most geologists adopt in a more or lB J . 1937 667. 20 H. U. Sverclrup M. W. Johnson and R. H. Fleming op. cit. p. 221 ; C. H. White a1 Proc. Roy. Irish A d .1943 48 B 9 161-212. Arner. J . Sci. 1942 240 714-724. 104 QUARTERLY REVIEWS less modified form the “ bar ” theory put forward by Bischof and developed by C. Ochsenius in the middle of the laat century. Salt deposition ia pictured as taking place from the waters of an-enclosed lagoon behind a permanent bar. So long as a constant or intermittent connection with the waters of the open ocean was maintained an inflowing cuprent of water of normal salinity across the bar would augment the supply of salts to the evaporating wafers of the lagoon. In this manner thick beds of salt could be formed but the high degree of concentration necessary to precipitate the more soluble salts must eventually have allowed the concentrated mother- liquor to collect in the more depressed parts of the basin.Even at this stage there may have been an occasional influx of new solutions either from the FIG. 5 D@pO8itiOn.d rhythm in two salt projELee (Lotze) 0 Deposition of non-saline sediments. 1 Formation of anhydrite. 3 Formation of potassium salts. 2 Formation of halite. sea itself from one basin to another or as a result of rainfall which dissolved and washed down the salts which had separated earlier on the marginal portions of the area and had been left exposed by the retreat of the water. Indisputable evidence of such periodic additions on both a major and a minor sc&le can be found in the deposits themselves. In many parts of the German field a succes8ion beginning with anhydrite and passing upwards through the deposition of halite to the stage of formation of potash salts is overlaid by clastic sediments (often a red saline clay) above which follows a more or less complete further cycle.A diagrammatic section by F. Lotze 23 reveals three such cycles (Fig. 5 ) in one area. In this country two partial cycles from limestone through anhydrite to salt followed by a third reaching as “Die Bildung der Steinsalzlager . .,” Halle 1877. “ Steinsdz und Kalisalze,” Berlin 1938 p. 151. PHILLIPS OCEANIC SALT DEPOSITS 105 the stage of deposition of anhydrite have been distinguished.24 Periodicity on a minor scale is seen in the so-called annual rings (" Jahresringe ") a series of thin layers or streaks of anhydrite or of polyhalite in halite. Their name is derived from the belief still accepted by some geologists that they represent changes in solubility of calcium sulphate consequent upon annual fluctuations of temperature but they may equally have arisen from periodic influx of further supplies of sea-water which reduced the concentration below the point of saturation for halite.A similar fine banding by layers of clay particles is sometimes observed and seems to speak conclusively in favour of the influx of muddy water possibly consequent upon increased rainfall due to climatic oscillation. The considerations which we havexo far advanced may help to explain in part the features presented by the natural profiles but it is generally accepted that they provide only a partial clarification and that an important branch of the study of marine evaporites involves the examination of changes effected subsequently to their deposition by the heat and pressure to which they have been subjected and by reaction with circulating solutions.The salts of the theoretical profile many of them highly hydrated will be specially susceptible to rise of temperature consequent upon burial under later over- lying sediments-salt deposits have been described as " the liveliest and most temperamental of rocks "-and the importance of subsequent therm&l metamorphism was stressed especially by F. Rinne 25 and Janecke.26 That many of the present mineral components of the salt deposits are the products of secondary changes seems to be abundantly clear. Lateral changes such as carnallite passing into Hartsalz have been frequently demonstrated and it has been found possible in some weas to trace a number of successive guide-horizons of beds of almost pure halite of great lateral extent which pass unchanged between different zones of potash salts.27 Examination of the detailed studies by K. Weber 28 of the Stassfurt region or by W. T. Schaller and E. P. Henderson 29 of the succeasive replacements in the Texas-New Mexico deposits will convince most readers that whatever may have been the ori@nal succession the salts now found are the products of profound alteration. A purely thermal metamorphism Consequent upon burial would bring about the successive '' melting " of various salts with the generation of solutions of various compositions. Under a depth of burial of about 3000 m. the change from gypsum to anhydrite would release almost pure water. If a primary bischofite layer were present this would melt at a depth of about 4200 m.t o an almost pure magnesium chloride solution. Jsnecke has traced these changes in detail and put forward the following scheme 24 S. E. Hollingworth Proc. Geol. As~oc. 1942 68 145. p6 See summary amount in " Handbuch der Mineralchemie," C. Doelter and H. Fmtschr. Min. 1920 6 101-136. Leitmeier 1929 IV 2 1283-1290. A. Tinnes Arch. Lagerstfur8ch. 1928 No. 38. asKali 1931 25 17-23 33-38 49-55 66-71 82-88 97-104 122-123. ae Geol. Surv. United States 1932 Bull. No. 833. 106 QUARTERLY REVlEWS Theoretical profile. Geothermally changed proflle. Bischofite zone + Yields MgC1 solution Carnallite zone Kainite zone + Hartsalz hone} Potwh-free magnesium 8ulphatc.i ___+ Polyhalite zone -p Polyhalite zone Glauberite zone Anhydrite zone] Anhydrite zonej Anhydrite zone Gypsum zone A specially important change is that by which Hartsalz would be derived Carnallite or -7 Hartsalz zone Loewite-Vanthoffite zone -+ Kieserite zone zone - { at 72" viz.Kainite + Carnallite + Kieserite + Sylvine + Solution but much of the Hartsalz shows an apparently primary lamination and would thus appear to be an original formation rather than the product of thermal metamorphism of previously-existing salts. If the solutions generated during these reactions remained in contact with the salts they would be available for a reversal of the reaction on declining temperature. Usually however they will have been pressed away to other regions changing their composition by further reactions such as the abstrac- tion of MgCl from carnallite and effecting further modifications in the composition of the various zones.The proponents of extensive metamorph- ism of this kind believe that such residual solutions (" Restlaugen ") have exercised a profound influence on the generation of the present-day profiles of marine evaporites. To those such as Fulda however who reject com- pletely the theory of a thermal metamorphism the only solutions which have been active are the connate solutions ('.' Urlaugen ") which represent portions of the mother-liquor enclosed with the deposits a t the time of their formation and percolating ground water the action of which will be con- sidered shortly. The details of Janecke's presentation have been criticised by H. B ~ r c h e r t ~ ~ who points out that the compositions of residual solutions postulated will in fact be reached only by stages and it is probably true to say that decreasing reliance is placed on a pure thermal metamorphism at the present time to explain the features of the actual profiles of marine evaporites.Consideration of the changes effected during deformation and earth movement and of the significance of the plasticity of salt deposits would lead into a purely geological field but we may note here that earth movements have .probably been responsible for producing the " Hasel- gebirge," an intimate mixture of salt gypsum and clay which is the common salt-producing rock of the Alps. By reason of their ready solubility the salt deposits are likely aIso to suffer further changes in the upper parts of the succession under the action of circulating ground water unless effectively sealed 08 by overlying clays.Here belong a number of reactions classified by German geologists as " Hutsalzbildung " t h e formation of an altered " cap " analogous with the secondary changes often found close to the stdace in other types of ore 30 Kali 1938 32 132-135 143-146 169-172. PHILLIPS OCEANIC SALT DEPOSITS 107 deposits. A re-formation of kainite is especially chsracteristic of them later changes KCl,MgCI2,6H,O + MgSO,,H,O + zHZO = KCl,MgSOa,3H,O + (Ma + yH,O) Carnallite + Kieaerite + Water = Iceinite + Solution The sylvine and kieserite of Hartsalz may combine to yield kainite KCI + MgS04,H20 + 2H20 = KCl,MgS04,3H,0 Sylvine + Kieserite + Water = Kainite Carnallite not accompanied by kieserite may yield sylvine (which with the associated halite is the sylvinite ore) its MgCl content passing into solution.Posthumous kainite and sylvinite thus formed are usually massive and unlaminated. Kainite itself may suffer a further extraction of its chlorine yielding schoenite 2(KCI,MgS0,,3H,O) + zHZO = KzSO,,MgS04,6H,O + (MgCl2,yHSO) Kainite + Water = Schoenite + Solution In the earlier days of the mining industry in Germany large quantities of these rich kainite rocks from the shallower zones were worked and sold for use as an artificial fertiliser ; the name has tended to linger on in application to a product which is now often a mixture of salts deriving its potassium content mainly from sylvine and kieserite. Anhydrite brought within the influence of circulating ground water will be converted into gypsum and the resultant increase of volume is usually accepted as the cause of the folded and contorted appearance (“ enterolithic structure ”) of many gypsum beds intercalated in salt deposits.This structure is occasionally found in beds which are now composed of anhydrite but it is clear that in the course of the complex history of some.of these deposits an anhyclrite rock converted into gypsum at one stage may have been again dchydratecl later in its history. In their studies of metamorphic changes in rocks geologists have for long been staunch supporters of the axiom “ Corpora non qunt nisi jluida,” looking always for the prcscnce of a solvent t o act as a medium for recrystal- lisation. During the last two decades however more attention has been paid to the possibility of reaction by diffusion in the solid state and this aspect of the study of oceanic salt deposits has been explored particularly by Leonharclt and his associate^.^^ Experiments on mixtures of powdered salts which were compressed and subsequently heated showed that even in the absence of solutions compounds such as langbeinite and vanthoffite may begin to form a t a temperature of 80° and this type of reaction ma-y well be important during metamorphism related to earth movements.Thus far wc have considered only those elements present in normal sea- water in relatively high concentrations. Amongst those of bedium concen- tration considerable interest has centred on boron. Reckoned as H,BO it ranks fifth amongst the anions with a concentration of 27 mg./kg. and as we have noted it must apparently be accepted 8 s an original constituent of the primeval ocean possibly derived from the presence of BC1 in the 31 J.Loonhnrdt Fortschr. Min. 1935 19 37-39; H. Ide Kdi 1935 29 83-86 93-96 103-105. 108 QUARTERLY REVIEWS original atmosphere.32 The work of W. Biltz and E. Marcus 33 showed that the boron content of the commoner minerals of the German deposits varies widely. Sporadically however a concentration is reached sufficient t.0 allow the formation of boron minerals of which the most important is boracite. When found as well-developed crystals boracite has every appearance of being a primary product (though this conclusion has been questioned),54 but it presents an interesting genetic problem. The external habit of the crystals is in agreement with cubic symmetry but in section the crystals are seen to be composed of doubly-refracting lamella? with the symmetry of an orthorhombic structure.When heated in the laboratory the crystals become truly cubic only at 265" ; it seems unlikely that such a temperature could ever have been reached during burial and metamorphism of the salt deposits and the explanation must be accepted at present that the crystals are pseudo-cubic polysynthetic twins. Boracite is found also in a massive form originally named " stassfurtite," either interbedded with carnallite or as concretionary nodules and most of this material is clearly secondary. Although stassfurtite is readily changed further by the action of circulating ground water giving rise to other boron minerals such as ascharite kali- borite and pinnoite which are found in the Hutsalz magnesium chloride being carried away in solution yet the well-crystallised boracite seems to be much more resistant to such changes.Lueneburgite another rare mineral in the German deposits has been recorded also from the Texas-New Mexico district .35 Of other rare minerals occasionally found in marine evaporites it will suffice to mention two examples. Tachhydrite may arise from carnallite as a secondary product under the action of solutiona containing calcium chloride 2(KC1,MgCI2,6H20) + CaCI = CaC1,,2MgC12 12H,O + 2KC1 in agreement with the observed paragenesis tachhydrite-sylvine-carnallite- halite. The calcium chloride solution may arise from the action of magnesium chloride solutions on anhydrite CarnaUite Tachhydrite Sylvine and tachhydrite has beer recorded in association with kieserite. Rinneite NaC1,3KCI,FeC12 is likely to form only where there is a high local concen- tration of iron.The small amount of iron present in sea-water is commonly represented in the marine evaporites by the particles of hematite which impart to most of the carnallite and many of the other salts a characteristic red colour (good colour photographs illustrating this feature are given by G. R. Man~field).~~ A regular zonal arrangement of these particles of iron 33 H. wettenberg 2. a w g . Chem. 1938 236 355. '3 I-. 1911 72 302-312. H. Werner Kali 1930 24 129-132. 36 W. T. Schaller and E. P. Henderson Geol. Surv. United States 1932 Bull. s6 J . Chem. Educ. 1930 7 737-761. NO. 833 pp. 47-48. PHILLIPS OCEANIC SALT DEPOSITS 109 oxide within the host crystal 37 suggests that they have developed after crystallisation the iron having originally been present as an isomorphous replacement of some part of the magnesium.In the American potash field hematite as a red pigment is most abundant in the potash minerals. The insoluble residue left on dissolving away the salt has a stringy structure and it has been suggested that the hematite is here the result of bacteriologic action.38 More rarely carnallite and tachhydrite are coloured yellow by ferric chloride or the iron may be present as FeS2 (pyrite) or as Fe30 (magnetite) colouring carnallite black. During the secondary reactions which have given rise to the cap salts the iron content of the primary salts seems to pass into solution and there is often a sharp dividing line between the brightly-coloured primary salts and the colourless overlying secondary products.For the direct formation of a compound such as rinneite it is possible that the presence of organic compounds retarded oxidation. In all 4 4 elements excluding dissolved gases have been demonstrated to be present in the water of the oceans though for some the demonstration is indirect by way of the examination of the ash of marine organisms. In the marine evaporites these rarer elements are seldom or never represented by individual compounds but occur as isomorphous mixtures or as trace elements in the commoner salts. Our present knowledge of their distribu- tion is still very uneven a systematic investigation of all the minor and trace elements by modern methods having yet to be undertaken. The following table gives the concentration in ocean water of a few of these minor elements to which we may direct attention here Element.I 1 Bromine . . . Strontium . . * . Fluorine . . . Rubidium . . . Lithium . . . Mg./kg.(Cl = lQ*ooo/oo). 65 13 1.4 0.2 0.1 Element. Iodine . . . Cssium . . . Silver . . . Gold . . . . Mg./kg.(Cl = lQ.OOO/,,). 0.05 0.0003 0*000006 (0.002) I I Bromine in spite of its relatively high concentration does not give rise t o distinct bromine compounds but is found isomorphously replacing chlorine in salts such as carnallite sylvine and kainite (ionic radii Br- 1.95 a. C1- 1.81 a.). Bischofite and tachhydrite may also contain appreci- able amounts but much less replacement is found in halite. Previously to the development of methods for the extraction of bromine from sea-water Germany possessed an almost complete monopoly based on her Permian salt deposits using carnallite and sylvine with a bromine content up to 0.4%.An early investigation by H. E. Boeke 40 showed clearly that a t Stassfurt the bromine content of a given profile varied regularly with the content of carnallite and a similar relationship has been demonstrated in the Russian A. Johnsen Zentr. Min. 1909 168-173. W. T. Schaller and E. P. Henderson Zoc. cit. pp. 11 38. a@ H. U. Sverdrup M. W. Johnson and R. H. Fleming op. cit. pp. 176-177. roZ. KTiBt. 1908 45 346-391. 110 QUARTERLY REVIEWS deposits near Solikamsk where J. Moratchevsky and A. Fedorova *l found the bromine content to be independent of the depth but directly related to the carnallite content of the rock. Iodine with its still larger ionic radius (2.16 A.) does not readily replace chlorine in these salts and is found only in very small amounts in salt deposits.A detailed study of the iodine content of the German deposits has been made by J. R ~ e b e r . ~ ~ Rubidium and caesium have also been extracted from carnallite in which they replace the potassium (ionic radii K+ 1-33 A, Rb+ 1.48 A. Cs+ 1-69 A.). Rubidium has been found in the waters of the present-day oceans but the caesium content of 0.002 mg./kg. quoted in the table above is calculated on the basis of the ratio Cs Rb in carnallite.43 G. Heyne 44 found rubidium also in sylvine and langheinite but none in the kainite and polyhalite which he examined. The presence of lithium has been demonstrated spectroscopically and its occurrence has a practical application. Although the.majority of salt mines are quite dry solutions occasionally break into the workings. If these are connate waters (" Urlaugen ") enclosed with the salts a t the time of deposition they will be limited in volume and will oventually drain away harmlessly. If however surface water breaks into a salt mine through fissures the consequences are likely to be disastrous and may entail the abandonment of the mine. The observation that the harmless Urlaugen are usually notably rich in lithium may hence be put to practical 1,188.~~ The very different proportions of the alkali metals in sea-water and in the marine evaporitcs when compared with their concentrations in igneous and sedimentary rocks are to be refcrrcd to the readiness with which they are adsorbed in fine-grained sediment^.^^ Strontium with a concentrat'ion of 13 mg./kg.in sea-water is the fifth most abundant kation and in agrcenicnt with this relatively Pigh concentra- tion the sulphate celcstinc has been occasionally recorded from salt deposits both in Europe and in Amcrica. More usually strontium is found replacing calcium (ionic radii Cat+ 0.99 A . Sr++ 1.13 A . ) either in the carbonates especially aragonite or in the sulphates. The fact that a higher percentage seems to enter the anhydrite structure compared with the small amounts usually found in primary may be used to study the vexed question of the original form in which the calcium sulphate of oceanic salt deposits was laid down. Anhydrite which has resulted from the dehydration of primary gypsum may be expected to show a low strontium content in comparison with primary anhydrite which separated directly from solution.Some occurrences of gypsum and anhydrite in the Permian deposits of Russia have been investigated from this point of view.48 4 1 Abstract in Neues Jahrb. Min. Ref. 11 1029 686. 4a Jahrb. Hallesch. Verb. Erf. mitteldtsck. Bodensch. 1938 16 129-196. 4 3 H. Wattenberg 2. anorg. Chem. 1938 236 346. 4 4 Abstract in Neues Jalrrb. Min. 1913 I 365. 4 5 E. Fulda 2. pr. Geol. 1039 47 11-14. O 6 W. Noll Chenz. Erde 1831 6 573. 48 L. M. Miropolsky and S . A. Borovick Compt. rend. (Doklady) Acad. Sci. U.R.S.S. 47 Idem ibid. 1934 8 559. 1943 38 33-36; 41 382-383. PHILLIPS OCEANIC SALT DEPOSITS 111 The high figures sometimes quoted for the supposed gold content of sea- water would suggest that this element also might become strongly concen- trated in the residual liquor and thus enriched during the deposition of the oceanic salts.It would appear however that these high figures are partly due to faulty analytical methods and also that much of the gold actually present is not in solution as ions but exists as discrete particles or in organic matter.4s Such gold would be adswbed on the finer clastic sediments rather than concentrated in the salts. J. Goubeau and L. Birckenbach 5O found the highest content of precious metals amongst the salt minerals in those which were notably coloured by inclusions of fine clay particles and showed also that the average content of the associated clays was higher than that of the salt minerals themselves. An attempt has been made in this review to trace the development of the study of marine evaporites and to outline some of the geochemical problems involved. For recent more extended accounts written from various points of view reference may be made to the works of Lotze (1938) and Borchert (1940) cited above or to two further works by F ~ l d a . ~ l ID H. Wattenberg 2. anorg. Chem. 1938 236 352. 61 “ Steinsalz und Kalisalze,” Stuttgnrt 1938 ; “ Die Salzlagerstatten Deutsch- 6O Ibid. pp. 37-44. lands,” Berlin 1940.