Until the overturn of the entire water column in February 1979, triggered by a drop in water level, the Dead Sea was a meromictic lake(Steinhorn et al., 1979). Since then, the lake has been holomictic for most of the time, with two brief meromictic episodes(1980–1982 and 1992–1996)in which the infl ow of massive amounts of freshwater caused the formation of a pycnocline separating the diluted upper 40 m of the water column from the deeper waters with >340 g/L total dissolved salts(Anatiet al., 1987; Anatiand Stiller, 1991; Oren and Anati, 1996; Anati, 1997; Oren, 2003a).

The first modern survey of the Dead Sea was performed by the Geological Survey of Israel in 1959–1960, and the report published(Neev and Emery, 1967)remains the baseline study for our underst and ing of the physical and chemical properties of the lake. Many earlier chemical analyses of Dead Sea water exist, some even performed on samples collected in the 18 ^th century(Nissenbaum, 1970; Oren, 2003b, 2006), but almost invariably the water examined was taken from the lake’s surface. A few results of analyses of deeper water samples collected in the first decades of the 20 ^th century were reported(Production of Minerals from the Waters of the Dead Sea, 1925; Elazari-Volcani, 1940; see also below), but overall we have hardly any information about the structure of the lake’s water column before the 1959– 1960 survey.

Surprisingly, it is little known that data on the physical and the chemical structure of the Dead Sea water column had been obtained already in the middle of the 19 ^th century, and the information collected then is highly relevant for the reconstruction of the limnological properties of the lake in earlier times. The expedition of Lieutenant William Lynch in 1848 reported the presence of a temperature minimum at a depth of ~18 m, and also retrieved a water sample from a reported depth of 338 m, close to the bottom, for chemical analysis. Sixteen years later, the Dead Sea exploration by the Duc de Luynes and his crew yielded detailed density and salinity profiles for a number of sampling stations. I here discuss the results of these pioneering studies, as well as information about the sampling equipment and measuring instruments used by the 1848 and the 1864 expeditions.

The first two attempts to explore the properties of the Dead Sea waters by boat failed tragically. The Irish theology student Christopher Costigan(1835) and the British Lieutenant Thomas Molyneux(1847)both died from exhaustion and dehydration(Kreiger, 1997; Oren, 2003b). The first successful expedition on the Dead Sea took place in 1848 and was led by Lieutenant William Francis Lynch(1801–1865)of the U.S. Navy(Fig. 1, left panel). With two rowboats, one of iron and one of copper, Lt. Lynch and his crew sailed down the River Jordan, and between April 20 and May 10, 1848 ^they made 164 depth soundings in the Dead Sea. The members of the expedition performed numerous additional measurements such as temperature and barometric pressure, and collected water and sediment samples for analysis. The report prepared by Lt. Lynch(Lynch, 1849a, 1852) and the “Narrative” of the expedition written in a more popular style(Lynch, 1849b)still form the basis for our underst and ing of many of the characteristics of the lake. A monograph was recently dedicated to Lt. Lynch and his explorations in the Holy Land (Jampoler, 2005).

The details of the 1864 expedition by the Duc de Luynes are not widely known. Thus, the books by Kreiger(1997) and Jampoler(2005)on the history of the Dead Sea and its exploration fail to mention it. This is probably due to the fact that the report was published posthumously, in French, in a very luxurious and expensive edition of three large volumes of text and one volume of plates and photographs, printed in a small number of copies only(Duc de Luynes, 1871– 1877). While Lynch’s Dead Sea exploration was performed on a low budget—in his reports, Lt. Lynch stressed more than once how he did all he could not to burden the American tax payers, no such financial constraints existed for the 1864 expedition. Honoré Théodore Paul Joseph d’Albert, duc de Luynes(1802–1867)(Fig. 1, middle panel), was an immensely wealthy French nobleman with wide interests in archaeology, numismatics, and a range of other topics. His custom-built 9.5-m-long sailing yacht was transported from Marseille to Alex and ria and from there to Jaffo. There the parts were loaded on camels and carried via Jerusalem and Jericho to the Dead Sea, where the boat was assembled. The vessel was operated by a crew of four professional sailors, and there was room for up to fourteen people. The leader of the scientific crew was the young geologist and paleontologist Louis Marie Hospice Lartet(1840– 1899)(Fig. 1, right panel), who later became famous for his discovery of the Cro-Magnon skeletons in 1868. The expedition spent three weeks on the lake in March-April 1864. Lartet’s findings were published in one of the volumes of the report of the Duc de Luynes explorations(Duc de Luynes, 1871–1877) and in other works(Lartet, 1866, 1869).

Fig. 1 Portraits of William Francis Lynch (1801–1865) (left panel), Honoré Théodore Paul Joseph d’Albert, duc de Luynes (1802–1867) (middle panel) and Louis Marie Hospice Lartet (1840-1899) (right panel) From http://www.history.navy.mil/bios/lynch_wmf.htm, http://www.ramboliweb.com/HTML/info.asp?InfoID=1389, and http://en.wikipedia.org/ wiki/Louis_Lartet.

Figure options

The depth soundings reported by Lt. Lynch remained the only reliable source of information on the bathymetry of the entire Dead Sea for more than a century. The 1959–1960 exploration(Neev and Emery, 1967)only covered the south-western quarter, the only part that was then accessible for geopolitical reasons. Precision depth soundings of the entire lake in 1974(Hall and Neev, 1978; Hall, 1996, 1997)confirmed the bathymetry of the northern basin of the Dead Sea as a long flat-bottomed dish, as earlier assessed by Lynch.

The device used by Lynch and his crew for depth soundings, as well as for collection of sediment and deep water samples, was a simple but for the time very modern instrument, developed just a few years earlier by Henry Stellwagen(1810–1866), a lieutenant in the U. S. Navy who later comm and ed the Mediterranean Squadron and retired as a captain in 1865. Lynch misspelled the name of the inventor in his Narrative: “One of the deepest casts, the cup to Stelwagon’s lead brought up a blade of grass, faded in colour, but of as firm a texture as any plucked on the margin of a brook”. Stellwagen’s bottom sampler, mounted below the lead of the sounding line, consisted of a conical steel cup affixed to the line just below the lead. A stiff valve lined with fl exible leather dropped over the cup to seal it closed as the line was hauled in, preventing the sample from washing out(Fig. 2). Lynch’s map of depth soundings includes information about the aspects of the bottom sediments, such as: “S and and Mud”, “Mud and Crystals of Salt”, “Ashy Mud”, “Soft Blue Mud with Crystals”, “Saltish Crust”, etc.

Fig. 2 “Stellwagen’s cup”, the sounding and sediment sampling device used by Lt. William Lynch during his 1848 Dead Sea expedition, open (left) and closed (middle), and its inventor, Henry Stellwagen (1810–1866) From: http://stellwagen.noaa.gov/ about/cup.html and http://stellwagen.noaa.gov/about/discovery.html.

Figure options

It is interesting to note that most of Lynch’s reported depth values were consistently too high. Most soundings in the deepest part of the lake yielded values of 184–195 fathoms(336–357 m). Reconstruction of the Dead Sea water level in the past centuries suggests that at the time of the Lynch expedition the shoreline was at ~397 below mean sea level(Klein, 1961, 1982), i.e., more than 30 m below the current level. The data collected in 1974 by Hall and colleagues, encompassing the entire northern basin with 540 km of track and 5 148 depth soundings, show that the large depth values reported by Lynch are all in error: the deepest point at -730 m in 1974(today somewhat less due to the accumulation of halite crystals on the bottom)was ~333 m below the lake surface in the middle of the 19 ^th century. One explanation may be that currents more or less consistently produced a slant reading. Another possibility is that growth of salt crystals within the linen sounding line caused the line to shorten as the diameter increased. This shortening would have been compounded with each use.

One of Lt. Lynch’s soundings gave a far greater value, 218 fathoms, which equals 399 m. This depth was measured 4.5 km west of the Zara(Callirrhoe)hot springs. On the map it is flanked by two readings in the ‘normal’ range, 190 and 185 fathoms(347 and 338 m). This result may be attributed to an experimental error due to the extremely difficult working conditions under the oppressive heat. It is surprising that several encyclopedias still stated a maximum depth of the Dead Sea of nearly 400 m until the mid-1980s at least.

Although the fact was not explicitly stated in Lynch’s report, we may assume that the water sample “drawn up by Capt. Lynch himself from a depth of 185 fathoms”(338 m)was collected from Stellwagen’s cup. Judging from the density of the brine(1.227 42 g/mL at 60°F=15.6°C; solid matter 267 g/kg)it is clear that it had not significantly been contaminated with water from the surface layers(density 1.13 g/mL, as measured by Lynch). The water was analyzed by Professor Booth and Mr. Alex and er Muclé in Philadelphia. They reported(g/kg): chloride of magnesium, 145.897 1; chloride of calcium, 31.074 6; chloride of sodium, 78.553 7; chloride of potassium, 6.586 0; bromide of potassium, 1.374 1; sulphate of lime: 0.701 2, total 264.186 7; water, 735.813 3. This analysis was surprisingly accurate and much superior to many later Dead Sea water analyses(Oren, 2006), as shown by the calculated ionic concentrations(in g/L): Na ⁺, 37.93; K ⁺, 4.30; Mg ²⁺, 45.69; Ca ²⁺, 14.03; Cl ^–, 219.81; Br ^–, 1.60; SO ₄ ^2ˉ, 0.58. Modern analyses of Dead Sea water, based on data from Oren(2006) and calculated for water of the same density, give values of 35.76, 7.65, 44.95, 17.25, 219.91, 5.46, and 0.47 g/L, respectively.

The report of the Lynch expedition to the Dead Sea contains an interesting observation on the physical structure of the water column of the deep northern basin: the existence of a layer at a depth of 10 fathoms(18 m)of a lower temperature than the surface water and the deep waters:

(Friday, May 5)... “Made a series of experiments with the self-registering thermometer … At the depth of 174 fathoms(1 044 feet), the temperature of the water was 62° [16.7°C]; at the surface, immediately above it, 76° [24.4°C]. There was an interruption to the gradual decrease of temperature, and at ten fathoms there was a stratum of cold water, the temperature 59° [15°C]. With that exception, the diminution was gradual. The increase of temperature below 10 fathoms may, perhaps, be attributable to heat being evolved in the process of crystalization. ...

(Saturday, May 6)… “Sent [Midshipman] Mr. Aulick out again in the iron boat, to make experiments with the self-registering thermometer, at various depths; the result the same as yesterday and the day previous, the coldest stratum being at ten fathoms.”(Lynch, 1949b).

Such a temperature minimum in the Dead Sea water column was also recorded in many profiles sampled during the 1959–1960 survey. In May 1959 a temperature minimum(19°C)was found at depths between 25 and 40 m, between the warm surface water(27–28°C) and the deep waters(21°C below 75 m)(Fig. 3). Presence of a temperature minimum was confirmed in 1939 by measurements of the Palestine Potash Company and in 1955 and 1956 by measurements of the IsraeliMining Industries(Neev and Emery, 1967). Similar trends were recorded in the summer months of 1981(a minimum of 20.9°C at 15 m, with 23°C below 30 m), in 1982(a minimum of 20.7°C at 20–25 m depth, with deep water below 50 m at 23°C), and in August 1993(the surface waters 35°C, a minimum of 22°C at 14 m, increasing to 23°C at 16 m, then decreasing to 22.3°C at 35–50 m)(Anatiet al., 1987; Anati, 1997). Such destabilizing temperature profiles are possible when compensated by a stabilizing salinity profile in a stratified water column. Proof of the existence of strong salinity gradients in 1848 can be found in the difference in density between the deep water sample collected by Lynch(1.227 42 g/mL) and the surface waters.

Fig. 3 Temperature profi le recorded at station 28(Fig. 4)on May 21, 1959 by means of a bathythermograph Redrawn from Neev and Emery(1967).

Figure options

The intriguing question is, what kind of “selfregistering thermometer” did Lt. Lynch use for his measurements? The Official Report, the Narrative, and other literature available to the author of this essay do not contain any further information. The type of bathythermograph that takes a continuous record tracing it on a smoked glass slide and used to collect the 1959 data as reproduced in Fig. 3 was developed in the years 1935–1938. Reversing thermometers became popular in oceanographic studies not before the last quarter of the 19 ^th century. The prototype was constructed in 1841 by the Frenchman George Aimé, who tested it in the Mediterranean Sea in 1841–1845, and instruments for routine use in oceanographic cruises were not developed until the 1870s. Aimé was an independent observer and researcher with few ties to the scientific institutions of his time. If Lynch had possessed such a hypermodern Aimé-type reversing thermometer, he surely would have elaborated about its action and its use, but his reports contain no such information.

Therefore it must be assumed that the type of “selfregistering thermometer” used by Lt. Lynch resembled the minimum-maximum thermometers based on the invention in 1780 by the British scientist and businessman James Six. Thermometers constructed according to his principle are still in general use today. Such instruments record both the lowest and the highest temperature encountered during the period of measurement. But then the question must be asked how such a thermometer can detect the presence of a cold layer of water located between two warmer strata. When lowered to the deep waters connected to the sounding line and then hauled in again, the instrument has to pass the intermediate cold layer twice. This layer can have a considerable thickness: 1959 profiles(Fig. 3)show it to extend over 20–25 m at least. Unless the response of the device was very slow and the instrument was hauled in very rapidly after equilibration at the desired depth, it is not clear how it could measure the deep temperature of 16.7°C when it had to pass a layer of 15°C to reach the deep waters.

It also must be noted that the temperature of 16.7°C(62°F)recorded by Lynch for the deep waters is unexpectedly low. The large difference in salinity between the surface waters and the deep brines documented by Lynch shows that the lake was at the time meromictic, and there are no indications that this situation had ever changed until the overturn of the water column in February 1979. The deep waters thus had then been isolated for 130 years at least, and there is no reasonable explanation for the apparent increase in their temperature from 16.7°C to 21.3–22.6°C recorded at depths >100 m in 1959–1960(Neev and Emery, 1967) and 21.4°C as measured in the beginning of 1978, a year before the overturn(Steinhorn et al., 1979). It is impossible to ascertain today whether indeed the temperature of the deeper water layers of the Dead Sea was significantly lower in the middle of the 19 ^th century. Alternatively, faulty calibration of the instrument used by Lynch may explain the difference.

To retrieve water samples from different depths in the Dead Sea, the Duc de Luynes and his company used a state-of-the-art sampling device, custom-built for the expedition(Fig. 4). The principle of the sampling instrument was first used by the abovementioned George Aimé in a study in the Mediterranean Sea in the 1840s, but there are no records that such an instrument was used again until Lartet had his improved version built for the occasion of the expedition of the Duc de Luynes. The mode of operation of the water sampler, based on the replacement of a column of mercury by water from the desired depth with minimum chance of contamination with water from shallower layers, is explained in detail in the report(Lartet, 1866, 1869; Duc de Luynes, 1871–1877)(translation: A. O.):

Fig. 4 The Dead Sea sampling stations where the Duc de Luynes and his crew collected depth profi les of water: stations A, B, C, and D, sampled respectively on March 15, 16, 16, and 18, 1864 The representation of the contours of the lake is based on the map published in Vol. III of the report of the Duc de Luynes expedition. The open circles indicate the location of stations 28 and 22 where respectively the temperature profi le shown in Fig. 3 was measured on May 21, 1959 and the density profi le shown in Fig. 6 was recorded on May 20, 1959(Neev and Emery, 1967).

Figure options

“I watched with interest Mr. Vignes and Mr. Lartet sample the lake, examine the density and the temperature of the water and collect samples from different depths. These gentlemen made their observations with the help of an instrument invented by Mr. Aimé, who used it near the coast of Algeria, and later perfected by Mr. Froment. The apparatus contained a volume of mercury which, at a given moment, was replaced by an equal volume of water. The mercury also served as a shutter, so that the water collected by tipping the instrument at the desired depth would come up without mixing with the masses of liquid through which it passed. … We ordered the construction of a new apparatus based on the principle of Aimé’s instrument, but with important modifications that made its operation easier and more reliable. … Basically it is composed of two parts: the sampling apparatus and the triggering device, which turns the former upside down after it has reached the water layer one wishes to study. The sampling apparatus consist of an iron tube that contains a test tube full of mercury connected to a conical cup which collects the mercury from the test tube when it is inverted. This is done by the triggering device: a lead messenger is lowered until it hits a disc connected to a bent rod, which forms a bolt before the opening of a copper cylinder, in which it slides with little friction. The lowering of this shaft causes the release of the sampling apparatus, which makes a half turn and remains in an inverted position, thanks to an iron rod connected to the triggering device. The mercury fl ows into the cup and is replaced with water in the test tube. This is the instrument used by Mr. Aimé in his studies of the water of the Mediterranean, and it could be used more widely if it were improved. …. Considering the usefulness of this type of instrument, not only to sailors but also to hydrographic scientists and to geologists, let us describe it. Although based on the same principle as that of Mr. Aimé, it has a number of improvements which make it more reliable, less expensive, and easier to use. The modifications … mainly relate to the bowl. They eliminate certain errors, and decrease, or even abolish the considerable loss of mercury that made the operation of Mr. Aimé’s apparatus so difficult. In Mr. Aimé’s sampling apparatus, the bowl is conical, with its tip pointing upwards. While lowering the instrument, some water from the upper layers is trapped there, its density being lower than that of the deeper layers through which the instrument subsequently passes. When the device is inverted, the water trapped in the bowl gets into the test tube. The small error which ensues detracts from the precision of the instrument. This is corrected by allowing free circulation in the bowl during descent. This is stopped by a set of valves when the unit is reversed, so that in this position the bowl can retain the mercury. In Aimé’s instrument the conical shape of the bowl and the large holes for the water cause a considerable loss of mercury. Therefore we used a cylindrical bowl and a smaller hole to slow down the sudden pouring of the mercury into the bowl, while the water enters through two even smaller openings on the side. The improved apparatus enabled us to obtain high precision water samplings. The instrument … always functioned well and lost only insignificant quantities of mercury. … To use the instrument one starts by emptying test tube E, which is full of mercury, into the iron cylinder B, intended to protect it. This same cylinder B is screwed into the cylindrical bowl A. The system can be lowered in this position, retained by ring O. During the descent, the water circulates freely in the bowl, and the fl ap-valves S and S' are retained by a rim R away from the holes which they have to close. To activate the reversion mechanism, the lead messenger C is allowed to slide along the sounding line. It hits the disc D, the bent rod is lowered, the ring O is unfastened, and the sampling unit is reversed. Then the fl ap-valves S and S' are pushed down by their own weight to cover their respective openings, closing the bowl to receive the mercury. While flowing out through opening ithe mercury in the test tube is replaced with seawater, which is then drawn up to the surface perfectly isolated and without any significant loss of mercury.”

The density of the water samples recovered from different stations(Fig. 4)was measured on board with the use of sensitive densimeters. The values obtained varied between 1.160 g/mL for surface water to 1.230 g/mL. The latter remains constant below a certain depth(Fig. 6, left panels). Lartet concluded that the waters of the infl owing sources only mix with the water of the lake in the upper zone. The water samples were then sealed in glass tubes for later analysis in France, but most of the tubes were broken when the boxes were treated roughly at the customs office in Marseille; only five tubes arrived safely to the laboratory in Paris for measurement of the water density at a controlled temperature and for chemical analyses.

Fig. 5 The sampling apparatus used by Louis Lartet for the collection of deep water samples from the Dead Sea in 1864 For explanations see text. From Vol. III of the report of the Duc de Luynes expedition.

Figure options

Fig. 6 Density profi les of the Dead Sea as sampled by Louis Lartet and his assistants at four stations (see Fig.4) between March 15 and March 18, 1864 and measured at temperatures between 20 and 30°C For comparison the right panels present the first density profiles recorded after the 1864 Duc de Luynes expedition: measurements in 1919 (sampling station not specifi ed) (Production of Minerals from the Waters of the Dead Sea, 1925) and at station 22 (see Fig.4) in May 1959 (Neev and Emery, 1967).

Figure options

The first density measurements on deep water samples from the Dead Sea after the 1864 expedition probably date from 1919 when Reginald Brock explored the possibility to produce potash and other minerals from the Dead Sea(Production of Minerals from the Waters of the Dead Sea, 1925). The density profile he measured(sampling station and equipment not specified)is plotted in the upper right panel of Fig. 6. Elazari-Volcani, in his Ph.D. thesis in which he characterized the microorganisms of the Dead Sea, described the sampling device he used in the 1930s, consisting of an evacuated round-bottom fl ask with an attached capillary that is broken at the desired depth by a metal messenger(for depths up to 10 m where the density was 1.187 g/mL); to obtain a sample from 50 m depth(density 1.223 g/mL), water was pumped through a hose(Elazari-Volcani, 1940). Elazari- Volcanialso mentioned a sample collected in 1935 by the Palestine Potash Company from a depth of 53 m(density 1.224 g/mL; 316.6 g/L total salts). New depth profiles including deep samples down to the bottom of the northern basin were only obtained in 1959(Neev and Emery, 1967). These profiles were very similar to those reported by Lartet based on his 1864 cruise nearly a century earlier(Fig. 6, lower right panel).

We possess only few limnological data on the Dead Sea prior to the exemplary 1959–1960 survey by Neev and Emery(1967). A few isolated measurements were performed in the 1930s when the establishment of the Palestine Potash Company(now the Dead Sea Works Ltd.)needed information about the properties of the lake as the source of its raw material for the production of potash and other minerals. In view of the dramatic changes in the water level and the transition from a meromictic to a holomictic regime with occasional brief meromictic periods, the existence of data on density, salinity and temperature at different depths of the lake collected by two groups of explorers in the middle of the 19 ^th century is especially valuable.

Both the Lynch and the Duc de Luynes expedition used state-of-the-art equipment: the deep water and sediment samples were recovered using Stellwagen’s cup, a device invented only a few years earlier. The Duc de Luynes and his chief scientist Louis Lartet improved an existing but not very satisfactory and complicated water sampling instrument; the performance of their custom-built water sampler was excellent, as attested by the Duc de Luynes in his report. The 1848 and the 1864 expeditions gave us the first information about the bathymetry of the lake, about the density stratification in its meromictic state, and about the occurrence of a temperature minimum along the salinity gradient. Working conditions during the two explorations were quite different. While both the Lynch and the Duc de Luynes expeditions did not repeat the mistake of their unfortunate predecessors Costigan and Molyneux and went out on the lake in March–May and not in mid-summer, the crew led by Lt. Lynch worked under Spartan conditions, while the Duc de Luynes and his coworkers enjoyed all conveniences that money could buy at the time.

Some of the data collected by these 19 ^th century expeditions have to be interpreted with caution. Examples are the too-great depths and the too-low temperatures of the different water layers reported by Lynch. But overall the information published in the reports of Lt. William Lynch, the Duc de Luynes and Louis Lartet is still highly valuable today for our underst and ing of the limnology of the Dead Sea.

The author thanks an anonymous reviewer for alerting him to the 1919 Dead Sea water profile collected by Brock, and the Earth Sciences library of The Hebrew University of Jerusalem for enabling reproduction of the Lartet water sampling device.