Popular Science Monthly/Volume 85/August 1914/The Origin of Nitrate Deposits




THE origin of all nitrate deposits was at one time accounted for by oxidation. The theory was held that the production of nascent nitrogen through the decomposition of organic matter caused a union to take place between the oxygen of the air and the nitrogen of the organic matter. Since then it has been shown that nitrates may be produced in various ways, and new theories are still being advanced from time to time to explain the origin of the nitrate deposits which occur in various parts of the world. The fact that different views have often been advanced to explain the origin of the same deposit has given rise to a great deal of discussion, and there still exists a wide difference of opinion as to the source from which the nitrogen may have been derived. By the source in this case is usually meant the preceding form in which the nitrogen appeared rather than the ultimate source of the nitrogen since this is generally admitted to be the atmosphere. The object of the present writing is to give a review of the various theories which have been advanced in this connection.

The fixation of nitrogen, that is, the transformation of elementary nitrogen into a combined state, may be brought about in the laboratory in a great many ways. Some of these processes have been shown to be profitable commercially and large quantities of nitrates and other compounds of nitrogen are now being manufactured in Europe and likewise in Canada, at Niagara Falls, but up to the present no commercial plant for the manufacture of "atmospheric nitrogen" products has yet been established in this country. These various processes of fixing nitrogen artificially may be grouped into four classes according as there is produced (1) nitrates or nitrites, (2) ammonia, (3) nitrides, or (4) cyanamid and its related compounds.

The fixation of nitrogen in nature also takes place in a number of different ways, but, unlike the technical operations just referred to, this is brought about principally by organic processes. Small amounts, however, are also fixed in nature by inorganic processes in a way analogous to some of the artificial methods, and in this way there are formed nitrates or nitrites, ammonia and nitrides.

The best known case of the inorganic fixation of nitrogen in nature occurs when nitric acid is formed in the air by lightning discharges at the time of thunder storms. The quantity of nitrogen which is combined in this way seems to differ in different parts of the world. In British Guiana during a period of twenty years the quantity so formed amounted each year on an average to 1.88 pounds per acre, while in Utah only 0.356 pounds was obtained annually for a period of three years. As a result of the decay of organic matter on the earth there is usually present in the air a sufficient quantity of ammonia to combine with all the nitric acid formed, although in the tropics the latter may sometimes be in excess. The ammonium nitrate which results from the combination dissolves in the snow and rain and in this way is carried to the earth along with other ammonium salts. Except in some parts of the tropics the total nitrogen recovered in this way is usually several times greater than the nitric nitrogen formed by the electric discharge. At Rothamsted, England, it amounted on an average during a period of eight years to 3.37 pounds per acre annually. At Ottawa, Canada, the average for the past five years amounted to 6.18 pounds per acre.

The quantity of nitrates which is thus formed in the air is small when considered locally, but in the aggregate the amount of combined nitrogen which is thus restored by nature to the surface of the earth is very great and is estimated at about 100,000,000 tons.

Other examples of nitrogen compounds which are not of organic origin are the metallic nitrides and ammonium salts which are found in the vicinity of volcanoes at the time of an eruption. It is probable that they are not carried as such in the fumes of the volcano, but are formed near the surface of the earth through exposure of molten rock to an atmosphere of nitrogen in a way analogous to the manner in which nitrogen is fixed artificially by the Serpek process, which consists in heating an ore of aluminium with carbon in an atmosphere of nitrogen. According to this view, nitrids would be formed first, and then ammonia when the former would come in contact with water. Many analyses have been made of the gases given off by volcanoes in different parts of the world, and free nitrogen has always been an important constituent, but so far as known combined nitrogen has never been reported.

A nitrogen compound of inorganic origin is also to be found in the case of a rare mineral which occurs in some parts of Arizona, and which has the distinction of being the only insoluble nitrate occurring in nature. This mineral to which the name Gerhardtite has been given is a basic nitrate of copper and is supposed to have been formed in the earth by water charged with air percolating over copper ore. It affords an illustration of an unusual way in which the fixation of nitrogen may be brought about.

The organic processes, however, are by far the most important in bringing about the fixation of nitrogen and the formation of nitrates. That nitrogen is one of the principal constituents of plants has been known since the beginning of the last century, but the source of the nitrogen was a matter of controversy for a long time afterwards. The experiments of some investigators showed that with sterilized soil and with all sources of combined atmospheric nitrogen cut off, the free nitrogen takes no part in the food supply of the plant. Other investigators arrived at just the opposite conclusion. These opposite views led to a great deal of discussion, and it was not until 1888 that Hellrigel was able to account for these conflicting results by growing leguminous plants in nitrogen-free soils. One set of plants was watered with distilled water, while to the other set was added in addition small amounts of leachings, containing only a trace of nitrogen, from a cultivated field. The plants watered with distilled water made but a small growth and soon died of nitrogen starvation, but those watered with the leachings reached a full growth and were found to contain about one hundred times more nitrogen than the seed sown. It was observed that the roots of the latter plants were covered with swellings, or nodules, which contained characteristic organisms while those which were watered with distilled water only had none. Furthermore, no nodules appeared and the plants did not develop when the soil leachings were sterilized before using. The experiments thus showed that the plants which were provided with nodules must have obtained nitrogen through the agency of the microorganisms; that these must have come from the soil leachings; and that they must have the property of fixing the nitrogen of the air. For some unknown reason these bacteria, to which the name Bacillus radicicola has been given, do not develop on the roots of non-leguminous plants; consequently, when plants of this kind are grown in the soil and harvested the total quantity of nitrogen present gradually becomes less. The advantage of rotating leguminous plants with crops of this kind thus becomes clear, because when a crop of the former is grown or plowed under as green manure the total nitrogen in the soil is increased. As the plants decay a part of the protein nitrogen of the plant again passes into the elementary state, part changes into ammonia, and a third part changes into nitrates. These changes are brought about by different bacteria, those responsible for the formation of nitrates being called nitrifying bacteria.

The amount of protein nitrogen which is converted into nitrates in the soil by these bacteria varies with conditions and depends on the physical condition of the soil, the quantity of organic matter present, the moisture content and the temperature. A basic element as potash, soda or lime must also be present with which the nitric acid formed may unite. On a limited scale these conditons may be so controlled that large quantities of protein nitrogen may be converted into nitrates, as is still done in India for the production of potassium nitrate to be used in the manufacture of gun-powder. The action of the nitrifying bacteria in thus leading to the formation of nitrates does not bring about any increase in the quantity of nitrogen combined, but simply brings about a transformation from one form to another.

Shortly after the discovery that the bacteria which are associated with the roots of leguminous plants are able to fix nitrogen, a great many experiments were made to determine if other bacteria have a similar function. Very conflicting results were obtained. Some investigators reported that certain bacteria are able to fix nitrogen, while others arrived at an opposite conclusion with respect to the same bacteria. Experiments along this line are still being made. It seems that most of the bacteria in the soil do not bring about any fixation of nitrogen, but there is good evidence that there are two or three kinds in addition to the Bacillus radicicola which are able to do this. Unlike the latter, these bacteria do not require the medium of leguminous plants, but may be active in soils which are devoid of growing vegetation. The most important of these bacteria have been called Azotobacter chroococcum. It has also been shown that various fungi, algaæ and other organisms possess greater or less power of fixing atmospheric nitrogen.

A special study of the amount of fixation which may take place through the activities of these organisms, particularly the Azotobacter chroococcum, has been made by Headden[1] and others at the Colorado Agricultural Experiment Station. It was observed that in certain parts of Colorado dark-brown spots occur in which nothing will grow. From the color of the spots it was natural to assume that the cause of the spots was due to the presence of black alkali, or sodium carbonate, which is so commonly met with in the soils of arid and semi-arid countries. An analysis of the soils, however, showed little or no sodium carbonate present, but instead surprisingly large quantities of sodium, calcium and magnesium nitrates. The spots observed were not fixed, but spread rapidly and sometimes covered an area of several acres in extent. Moreover, new spots were noticed to appear in old and new localities. In some cases the amount of nitrogen accumulated, as shown by analysis, amounted to as much as one hundred tons to the acre foot. Arguments were given to show that the nitrates could not have originated with the irrigating water nor from concentration of ground water in the surface layers of the soil. That the spots are the remains of great herds of extinct animals which perished from some unknown cause, was likewise considered improbable for the reason that the areas involved are too large, and that the spots are increasing in size and appearing in localities where they were never before noted. The theory was therefore advanced that the excessive quantities of nitrates were formed in situ, through the nitrogen-fixing activity of micro-organisms in the soil.

It was actually shown that the power to fix atmospheric nitrogen is a property common to many cultivated Colorado soils, and that this power is not confined to nitrogen fixation in solution, but is manifested in soils as well. The principal nitrogen fixing organisms in the soil were identified as Azotobacter chroococcum. These have the power of developing a dark brown color in cultures containing organic matter and a nitrate, but give no color when the nitrate is replaced by organic nitrogen. The rate at which the fixation took place was considered sufficient to account for the formation of the nitrates found in the soil.

Exception has been taken to this view regarding the origin of these nitrate beds by Stewart and Greaves[2] who made a study covering a period of eight years at the Utah Experiment Station of the influence of irrigating water upon the production and movement of nitric nitrogen in the soil. Although the soils upon which the investigations were made were ideally adapted both chemically and biologically to support a rapid biological action, no unusual amounts of nitrates were found. A recalculation of the results reported by Headden showed that in the samples which were taken from the surface of the soil at different points in the same locality and from the same point at different depths, the nitrates and chlorides varied in the same ratio, and that whenever an accumulation of the former took place during a given period, the latter also increased during the same time in the same general proportion. It was therefore concluded that the excessive quantities of nitrates formed in the soils of Colorado were not formed in situ, but owe their origin to the same source as the other water-soluble salts.

Further investigations by Headden[3], however, showed that while large amounts of chlorine generally occur with excessive nitrates, this is accidental rather than necessary, and that on the whole there is no ralation between the amount of nitrates and that of any other class of salts present. Additional evidence is given to show that the concentration in nitrates in brown spots in which nothing will grow is not due to the accumulation of preexisting nitrates, but to the action of microorganisms which are able to bring about the fixation in the soil of atmospheric nitrogen, and that the dark-brown color which is characteristic of the spots is due, not to black alkali, but to the development of pigment by the organisms.

The occurrence of nitrate deposits in caves has long been known. During the War of 1812 and again at the time of the Civil War, the "saltpeter" deposits in the Mammouth Cave, and in other caves in Alabama and Georgia, formed an important source of nitrates required in the manufacture of gunpowder. The origin of these deposits is commonly ascribed to the decomposition of animal remains, and particularly to the excrements of bats. In the southwest small deposits of guano are still to be found in some caves, but almost all deposits which are sufficiently large to be of commercial value have been removed for use in the manufacture of fertilizers. Samples taken from some of the caves where the guano has undergone decomposition have been analyzed by the writer and found to run as high as 75 per cent, of potassium nitrate. Other samples have been examined which consisted of very pure elongated white crystals of calcium nitrate, and which had been taken from crevices in a cave. When placed in a humid atmosphere these soon melt in the moisture which they absorb from the air, but may be kept indefinitely in a desiccator.

While guano is usually considered to be the source of the nitrates found in caves, other theories are occasionally advanced to explain the origin of some of the deposits. Thus Hess[4] considers that guano could not be the source of the large store of nitrates which have been taken from the Mammouth Cave at distances of over five miles from any opening which leads to the surface, since bats go, as a rule, but a short distance from the entrance to the cave. Moreover, in the bottom of many caves there are to be found earths from which nitrates can be extracted, but which do not contain any animal remains as would be expected if the nitrates were derived from guano. To explain the occurrence of nitrates in caves of this kind, the view is put forward that the nitrates do not come from guano, but originate in the surface soil above the eaves through the oxidation of organic matter by nitrifying bacteria. As the soil in limestone regions is usually loose and porous, the nitrates are considered to be carried down by percolating water and deposited in the floor of the caves. Air currents in and out of the caves would remove the water by evaporation, and the nitrates would consequently remain and would not be washed away so long as the inflow of water did not exceed that lost by evaporation.

A similar explanation is given for the origin of nitrate deposits under overhanging cliffs. Thus, water carrying nitrates dissolved from the soil percolates through the earth and a portion finally oozes out at the surface underneath where it evaporates and leaves the nitrates behind. Being protected from the rain in this position, the nitrates in this way are enabled to accumulate.

The theory advanced by Hess to account for the origin of nitrates in caves has not met with universal acceptance. Nichols[5] has argued that bats do frequent remote parts of caves; that cave earth does contain organic matter; and that the proportion of phosphates to nitrates in the cave earth is much too great to be accounted for on the supposition that they were brought in by percolation from the surface soil. It is considered that while small deposits may be found in the way described by Hess, the great bulk of the nitrates that are found in caves results from the decomposition of bat guano. By leaching of the soluble salts from the guano, the nitrates are removed and may be concentrated in other parts of the cave or distributed elsewhere.

Because of their solubility, it is not to be expected that any large accumulation of nitrates can take place excepting in protected places such as caves, or in arid countries where there is very little rain. In such countries, however, there is very little vegetation, and consequently the organic processes which are of such importance in bringing about fixation of nitrogen in humid countries are able to operate to a much less extent in desert regions. We thus find that in the soil of such regions the total nitrogen present is usually much less than in soils which support a good vegetation.

On the other hand, the soils of desert countries have marked nitrifying powers, with the result that a large percentage of the nitrogen actually present is converted into nitrates. Furthermore, owing to the low percentage of organic matter present and to the porous nature of desert soils the anaerobic dentrifying bacteria are not so active in changing nitrates into free nitrogen, while the lack of vegetation prevents their convertion into protein nitrogen. The nitrates which are formed are thus enabled to accumulate, and either remain in the soil or are transferred by underground waters to other localities where they may be concentrated by evaporation of the water at the surface of the ground. If there has been any introduction into desert localities of organic matter from external sources, as may be brought about by the droppings or remains of animals, the accumulation of nitrates may be correspondingly increased. It thus happens for the reasons given that the largest nitrate deposits are found in desert regions.

In this country few nitrate deposits are to be found apart from those of cave origin. The most extensive so far known occur in San Bernardino and Inyo Counties, California, along the shore lines or bed beaches of what was supposed to be a former sea, but which is now geologically known as Death Valley. The region popularly known as Death Valley is that portion of the valley proper which is below sea level. The territory covered by nitrate beds has been estimated to cover an area of about 35,000 acres. Through erosive agencies the clay beds in which the nitrates were deposited have been worn into buttes and ridges of characteristic shape and color. The hills so formed vary from only 50 feet high to over 300 feet. Samples taken from the niter-bearing strata in the hills, and exposed by erosion, vary all the way from a trace to more than 50 per cent, of nitrates. It is generally agreed that these deposits have not been formed in situ, but have resulted from the concentration of nitrates formed from the decomposition and nitrification of animal and plant life which must have existed in the region at the time that the valley was filled with water. Owing to the limited distribution of these nitrates, they are not considered of much commercial importance, and the same may be said of all other deposits so far discovered in this country.

Small nitrate deposits are also to he found in various other parts of the world, as in the Sahara, in Russian Turkestan, and in Egypt, where nitrate earths occur which contain about 15 per cent, of calcium and sodium nitrates. The earth has long been used locally as a fertilizer, and its use is supposed to be increasing. The source of the nitrates in this region is not known.

All known deposits, however, which occur, like the ones just referred to, in various desert regions throughout the world are insignificant compared with the well-known deposits in the deserts of Atacama and Tarapaca in the north of Chile. These deposits command a great deal of interest, not only on account of their commercial importance, but also for the many attempts which have been made to explain why the quantity of nitrates in this particular region should be so large compared with any other known deposit.

The first shipment of nitrates to Europe from Chile was made in 1825. Since then the annual exportation has continuously increased until in 1912 the total quantity exported amounted to 2,485,860 tons of which 1,925,590 tons went to Europe, 469,100 tons to the United States, and 91,170 tons to other lands.

The arid region in which the nitrates are found extends for about 430 miles between 13° and 25° south latitude and lies between the Andes in the east and the Coast Range on the west. This area lying between the two mountain ranges does not form a continuous valley, but is broken up by transverse ranges into a series of elevated basins or plateaus. These plateaus are generally flat or undulating, and have an elevation from less than 2,500 feet to more than 5,000 feet. They have a general slope from the foot of the Andes towards the Coast Range, and as a result the lowest part of this plateau region, or pampa as it is called in Chile, lies along its western border where it joins the foothills of the Coast Range. It is along this zone that the nitrate deposits occur. The surface of the surrounding region is dry and sandy and vegetation is totally absent.

The nitrate beds as they occur in different parts of this region vary in thickness up to about six feet. They are usually found at or near the surface, but may in some cases be covered with an overburden to a depth as great as thirty feet. The nitrate deposits are never found pure, but are always mixed with sodium chloride and other salts, and are impregnated with insoluble earthy material. Crude nitrate may sometimes run as high as 60 to 70 per cent. of sodium nitrate, but a deposit running 50 per cent. is considered high-grade material. Material containing less than 16 per cent. is too low grade to be mined at a profit at present.

The source of these deposits is a subject which has given rise to a great deal of discussion. Many theories have been advanced to account for the origin of the nitrates, but all appear to fall short of adequately accounting for all the conditions under which the nitrates are found in Chile. It is generally considered that an organic source is the most probable, but there have not been lacking explanations for the formation of these nitrates which have been based on inorganic agencies.

Thus one of the theories advanced is that the nitrates may have resulted from electric storms occuring in the Andes. It has been suggested that the nitric acid which is formed in this way by the oxidation of the nitrogen of the air becomes changed into calcium nitrate on coming in contact with the limestone of the mountains, and that this in turn on being washed down into the pampa region has been converted into sodium nitrate in coming in contact with the sodium salts already existing there. It has also been stated that at certain seasons of the year there is a great deal of static electricity in the air over the desert region, owing to the strong winds and the extremely dry climate, and that the nitric acid which is formed as a result of this condition is carried to the ground by the moisture in fogs which drift in from the sea.

The view has also been advanced that the nitrogen in the Chilean nitrate may have come from nitrogenous fumes given off by volcanoes in the Andes. It has already been pointed out that nitrides and ammonium salts are sometimes found in the vicinity of volcanoes after an eruption, but whether these compounds result from the direct fixation of atmospheric nitrogen near the mouth of the volcano, or from some combined form of nitrogen already present in the earth is not known, but the former view is the more probable. It has been shown, however, that the source of the nitrogen is not organic.

It has been claimed by some that alkali carbonates are able to bring about the fixation of atmospheric nitrogen into nitric acid in the presence of oxidizable matter, and Pissis[6] expressed the opinion that the niter beds in Chile were formed in this way. It was pointed out that the decomposition of feldspar rock in the region of the Andes supplied alkali carbonates, while the protoxide compounds of iron which are common in the rocks of the pampa are easily oxidized under ordinary conditions to form peroxide compounds of iron. The view was accordingly put forward that the alkali carbonates in contact with rocks of this kind brought about the oxidation of the nitrogen of the air with the ultimate formation of nitrates.

Perhaps the most far-fetched attempt at an explanation of the origin of these deposits was that presented by a writer[7] in the Comptes Rendus a few years ago. It was observed that pieces of glass left on the ground in the vicinity of the saltpeter mines in the Province of Aconcagua, Chile, became colored blue in a short time, while samples of the same glass exposed on the roofs of buildings to the direct rays of the sun remained colorless. This suggested the possibility of the soil in the vicinity being strongly radioactive, which was thought to be confirmed by the action on photographic plates properly protected and subjected to an exposure in the ground for a month. It was suggested that the radioactivity of the soil as indicated by these experiments might have had something to do with the formation of nitrates in this part of Chile. It is now known that all soils are slightly radioactive and to approximately the same extent.

None of these views which have suggested an inorganic mode of formation for the Chilean nitrates have received very general acceptance. Much more credence is given to the theories that the nitrates found in the deserts of northern Chile have resulted from the decomposition of organic matter brought into the basins in which they are found from outside sources.

One of the most popular of these theories suggests sea-weeds as the source of the nitrates. The explanation is given that in past ages the pampa regions were sea beaches, and that an enormous amount of seaweed was piled up on them. In course of time the beaches were elevated above sea-level, and the collected sea-weeds in decaying under arid conditions decomposed in such a way that the nitrogen present was converted into nitrates, and the iodine into iodates. It may be pointed out in this connection that immense groves of giant kelps are now to be found along the Pacific Coast of North America, and that the proportion of iodine to potash in the dry plants is about the same as is to be found in the crude niter of Chile. The ratio of nitrogen to potash in the former, however, is very much less than in the latter.

There are many objections which may be offered to this theory. Thus, if the niter came from sea-weed it must necessarily contain bromine as well as iodine, since both are present in this source, and there is no known natural process which can bring about the separation of bromides and iodides. So far as known, however, bromine has not been found in any of the nitrate deposits of Chile; whereas, from analyses made by the writer, the bromine in the giant kelps of the Pacific, for example, is of the same order as the iodine.

Again, sea shells are never found in the nitrate beds, and the stones in the neighborhood are sharp and jagged and show no signs of being worn by water as they must necessarily have been if they had at one time existed on a sea beach.

Perhaps the best known theory which has been advanced to explain the origin of these nitrates is that they have been derived from the decomposition of ancient guano deposits. In defending this theory Penrose[8] has assumed, in the same way as those who favor a marine origin for the nitrates, that the pampa region was once a part of the ocean bottom, but as the region gradually rose it became a more or less enclosed basin. At this time guano beds were supposed to have been deposited along the borders of these waters, just as they are now deposited in the neighboring shores of the Pacific. It is considered quite possible that marine plants might also have collected in the basin at the same time, and that these constituted the source of the iodine, although it is pointed out that this element might also have originated in minerals, or mineral springs occurring in the region. The formation, as suggested, of inland basins of sea-water, which would ultimately evaporate, would furnish also a source for the common salt associated with the nitrates, as well as for the soda of the nitrates.

The guano theory, however, has been objected to on the ground that no accumulation of phosphate has ever been found in the nitrate country, and such must necessarily occur in amount corresponding to the nitrates if the latter have been derived from guano. It is argued, on the other hand, that such phosphates may actually exist, but that they have not yet been discovered, and it is further explained that the absence of the remains of birds and of sea shells may be accounted for on the ground that sufficient time has elapsed since the beds were deposited to admit of the decay of all such materials.

There is still, however, a further objection which applies to both the sea-weed and guano theories. Thus, if the region was at one time a sea-beach it must have taken ages, as Newton has pointed out, for the nitrate pampa to be elevated to its present level. During these ages the region must have passed through varying climatic conditions, including most probably rains. It has, therefore, been argued that the nitrate deposits are, geologically speaking, of very recent origin.

In suggesting another organic source from which the nitrates may have been derived, Kuntze has called attention to the fact that vicunas, and llamas, which are at home in this portion of the Andes, have the peculiar habit of always depositing their manure in one and the same place. Immense herds of these animals are supposed to have roamed over the region from time immemorial, each herd having a definite dunging place at some convenient point. As the manure accumulated its nitrification would progress rapidly under the prevailing arid conditions. The common salt would be derived from the urine and excrements, while the decomposition of rocks throughout the region is considered sufficient to account for all other salts occurring in the crude niter.

Newton[9] is of the opinion that the source of the nitrates is the organic matter occurring in the soil of the great plain lying between the nitrate beds and the Andes. It is pointed out that there exists in this region all the conditions which favor the rapid conversion by nitrifying bacteria of the nitrogen of organic matter into nitrates. The soil is porous and basic in its nature, and contains organic matter chiefly of ancient vegetable origin; the temperature is high and, on account of the absence of rain, there is no growing vegetation to absorb the nitrate, and therefore it must accumulate. The mountain floods which swamp the plain once in every seven or eight years are considered chiefly responsible for transporting and concentrating the nitrates from the superficial layers of the pampa soil to the lower western part of the pampa region where the deposits are found. The nitrate deposits are thus looked upon as the concentrated fertility of the thousands of square miles of land between the watershed of the Andes and the Coast Range. It is admitted that electrically generated atmospheric nitrate may also be present.

Headden[10] has suggested that the nitrates of Chile may have been formed by the direct fixation of the nitrogen of the air by nitrogen fixing bacteria in the same way as accumulations of nitrates have been shown to have been formed in certain soils of Colorado.

It is apparent from the views which have thus been advanced to explain the origin of the Chilean nitrates that no single theory has yet been proposed which is adequate to account for all the conditions under which the deposits are found, and it seems most probable, as some have suggested, that instead of being formed in one way only, the nitrates owe their origin to several sources.

  1. Col. Agri. Expt. Sta., Bull. Nos. 155, 178 and 179.
  2. Utah Agri. Expt. Sta., Bull. No. 114.
  3. Col. Agri. Expt. Sta., Bull. No. 186.
  4. Jour. Geol., 8, 129, 1900.
  5. Jour. Geol., 9, 236, 1901.
  6. "Nitrate and Guano Deposits in the Desert of Atacama," A. Pissis, London, 1878.
  7. Bordas, Compt. rend., 147, 924, 1908.
  8. Jour. Geol., 18, 1, 1910.
  9. Jour. Soc. Cliem. Ind., 19, 408, 1900.
  10. Col. Agr. Expt. Sta., Bull. No. 155.