Geology and Ground-Water Virgin Islands Hy D. J. CEDERSTROM GEOLOGICAL SURVEY WATER-SUPPLY PAPER 1067 Ground-water resources V a moderately dry Caribbean island This e6py is PUBLIC PROPERTY and is not to be removed from the official files. PRIVATE POSSESSION IS UNLAWFUL (R- S. Sup. Vol. 2, pp. 380, Sec. 749) "NITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1950 UNITED STATES DEPARTMENT OF THE INTERIOR Oscar L. Chapman, Secretary GEOLOGICAL SURVEY W. E. Wrather, Director For sale by the Superintendent of Documents, U. S. Government Printing Office Washington 25, D. C. - Price $1.25 (paper cover) CONTENTS Page Abstract .................................... . . - 1 Introduction ............................................................ ...........-....-...... ... . 2 Purpose and scope of investigation........................................-. ... .. 2 Acknowledgments................ .. . .. -. 3 Location .................. . ........................... . .... ................ 3 Topography and drainage............................................................................ 4 Historical sketch ............................. ....................................................... 6 Population ............ .................................................................................... 7 Agriculture and industry. . . 7 Flora and fauna . . .. . - . . . 9 Climate ..................... ............. ..... .. ...... . .......................... 9 Rainfall ..... .................... .............. ... . .................................. 9 Accumulated departure of rainfall from normal.......................... 12 Relation of rainfall to sugar production........................................ 12 Geology......................................................................-..-...--...-...--.. ............ 12 Work of previous investigators..............-.... . . 12 Summary of geologic history... .. . ..... .. 14 Geologic formations .................... .......... ................... ... ..... .............. 16 Mount Eagle volcanics.............. . ......... .................................. 16 Diorite intruded into Mount Eagle volcanics . ..... 17 Jealousy formation .. ........................... ................................... 19 Kingshill marl ............... ........ ..................................................... 20 Description and distribution................. ............. 20 Geologic age ................................. ....... ................................ 22 Factors influencing deposition of Jealousy formation and Kingshill marl .. .. ............ 26 Alluvium ................................................................................................ 27 Structure ..... .................... ..... ................... ... ....... ............... .......... 29 Physiography ................. . . . .. ................... ........................................... 34 Early Tertiary erosion cycles..................... .. ............... 34 Drainage pattern .......................................................................................... 35 Land forms ................................ .. ......... ... .......................... 36 Submerged coral reef ..... .......... ..................................................... 36 Coastal terraces and buried valleys..-......... . ......... 37 Emerged coral reef and beaches... 38 Sources of water supply and use of water......... .... . . 38 Rain water ...................................................................................................... 38 Stream flow ...................... . ............................................. 39 Wells .-- .-.----...---...----...--. . . . .................... 39 Springs ........................- ............................................ 40 Water supplies at Central Factory............................................................ 40 Water supplies at Bethlehem Factory.. ....... ................ 40 Summary ................. ............. ........... .............................................. 41 Well construction .................................................................................................. 41 Dug wells ........ .. .. . . .. ..................................................... 41 Drilled wells ................................ ............................................................... 41 ill IV CONTENTS Page Occurrence of ground water................................................................................ 43 Previous investigations of ground water.................................................. 43 General conditions ...............~.. .................................................................. 45 Ground water in the Mount Eagle volcanics............................................ 47 Ground water in the diorite... .................................................................. 50 Ground water in the Kingshill marl.......................................................... 51 Ground water in the alluvium.................................................................... 56 North Side ............................................................................................ 56 East End ................................................................................................ 56 Central plain .......................................................................................... 58 Ground water in beach sand........................................................................ 64 Test-drilling program .................................................................................. 64 Quality of water.................................................................................................... 69 Comparison of surface water with ground water.................................... 69 Contamination by sea water.....................-...-.......-..................-.- .............. 69 Chloride contamination resulting from organic pollution.................... 73 Saline minerals associated with limestone of Tertiary age.................... 74 Effects of alkali soils................................................................... 75 Effects of salt spray..... .... ..... .. ........... . . . 77 Base exchange .. . ... ........... .. ....... .. . .................. 77 Summary ............. ..................................... ......................................... 78 Quality of water in the various formations............................................ 78 Mount Eagle volcanics................. .................................................... 78 Dioritic rocks .................. ........ .... . ... . ............... ........ 79 Kingshill marl ........ .. ............ ......................................................... 79 Alluvium . ........... .................................................................... 82 Wells from alluvium within areas of Mount Eagle volcanics -. . .. 82 Wells from alluvium within Tertiary limestone areas. . 85 Beach sand .............................................................................................. 86 Suitability of water for various uses.. . ..................................................... 86 Human consumption ... .... ................................................ 86 Cattle - .-----.-...--.-...--...-..-. ...... ..____ ....................... 87 Irrigation ........................................................................................................ 87 Sugar and rum production........... ...................................................... 89 Summary of ground-water resources................................................................ 90 Water-bearing formations .......................................................................... 90 Adequacy of present supply.. .................................................. 90 Quality of water....................................... .................................................. 91 Records of wells and springs . ............................................._..... 91 Recommendations .. .... ..... .................. ................................................... 91 Appendix, National Park Service drilling project ....................................... 103 Introduction . ....................................... .......................... ........ .......... 103 Selection of well and test-hole sites . . . .. . ...... 104 Summary of results.. ... .. ....... .............................................. 104 Description of wells and test-holes............................................................ 105 Whim homesteads .................. ....................................................... 105 Paradise .................................................................................................. 106 Proposed underflow dam at Fair Plain............................................ 106 Fair Plain ................. ........................................................................ 107 CONTENTS v Appendix Continued Description of wells and test-hole sites Continued Page Golden Grove......................................................................... ...---.. 107 Fredensborg ................. ....... ....................... .. . 108 Colquhoun-Mount Pleasant homesteads ................. ............ ... . 108 Rattan ...... ........ . ........ .. ......... ......... ............. . .. . 109 St. John-Princess homesteads........................................ . .. . 109 Central Factory ........ .. .... ............. ...._................ .. ...... HO Christiansted .......................................................................................... Ill Cotton Valley ........................................................................................ Ill Cottongarden..... .... ............. .................. ..... . .... . Ill Resume of National Park Service drilling project..... .. . H2 Index ...................................................................................................................... 117 ILLUSTRATIONS Page PLATE 1. Geologic map of St. Croix............................................. ...... In pocket 2. A, Basinal area in vicinity of Fountain estate; B, Central alluvial plain. View toward Mount Eagle. . . 10 3. A, View across inner lowland; B, View across narrow alluvium-filled valley . . . 34 4. A, View across western plain; B, Emergent coral reef at Canebay, north side... . . 35 5. A, Artificial catchment basin at Mary's Fancy; B, Municipal well, Christiansted.. 50 6. A, Old well tower at Strawberry; B, Rotary rig at test well near Jealousy... ....... ......... .... . . ..... . 51 FIGURE 1. Map showing location of St. Croix.............................................. 4 2. Chart showing annual rainfall and accumulated departure from normal at St. Croix, 1852 to 1938 .. 11 3. Generalized geologic section showing structure in the eastern part of the Tertiary marl plain.. 31 4. Cross section of central St. Croix showing location of test wells with reference to geology.................................... 32 5. Section showing development of early Tertiary erosion surfaces along east edge of marl plain...................... 35 6. Diagrammatic section showing the zone of saturation and the three belts of the zone of aeration.............................. 46 7. Diagrammatic sections showing the movement of ground water: A, In times of normal or exces- sive rainfall; B, In times of deficient rainfall ................... 46 8. Diagrammatic section showing artesian conditions. .. ... 47 9. Diagram showing results of pump test on the developed well at Fair Plain.......................................... 62 10. Graphic representations of analyses of some St. Croix ground waters and surface waters ........ 71 11. Diagrams showing relation between salt water and fresh water in homogeneous water-bearing sands making up a small island: A, Under nonpumping conditions; B, Where pumping is taking place...................... 72 VI CONTENTS TABLES Page TABLE 1. Annual mean rainfall on St. Croix from 1852 to 1938............ 10 2. Average monthly rainfall on St. Croix, 1852-1935.................. 10 3. Logs of test wells drilled in St. Croix, 1938-39........................ 68 4. Analyses of ground water in Leper Asylum we1!, St. Croix, showing chloride contamination and analyses of ocean water and of water from the Tertiary limestone area for comparison.................................. 72 5. Analyses of waters from wells showing organic pollution...... 74 6. Mean monthly evaporation on St. Croix from 1920 to 1935.... 75 7. Analyses of water from well at Kobel in dioritic rock.............. 79 8. Analyses of waters from wells in the Kingshill marl.............. 80 9. Analyses of waters from wells in the alluvium within areas of Mount Eagle volcanics..... ... .................... 83 10. Analyses of waters from wells in the alluvium within Tertiary limestone areas................................................ 84 11. Records of wells and springs in St. Croix.................................. 94 12. Logs of wells and test holes drilled by the National Park Service in 1940-41............................................ 112 13. Records of wells and test holes drilled by the National Park Service in 1940-41............................................ 115 Geology and ground-water resources of St. Croix Virgin Islands By D. J. CEDERSTROM ABSTRACT Field work in the island of St. Croix, V. I., was carried on from December 1938 to April 1939 in connection with a test-drilling program for water sup- plies. The island is 21 miles long and has a maximum width of 6 miles. Its western part consists of a range of mountains flanked on the south by a rolling plain; its narrower eastern part is entirely mountainous. There are only a few small streams. The rolling and flat lands are cultivated or are in grass, and the mountainous areas are either wooded or in grass. The average rain- fall of the island is 46.34 inches, but severe droughts and periods of excess precipitation are not uncommon. The island is made up of rocks of Upper Cretaceous age, mostly voTcanic tuffs and limestones known as the Mount Eagle volcanics; diorite intruded into the cretaceous rocks; and Oligocene to Miocene blue clays and yellow marls (the Jealousy formation and Kingshill marl, respectively). Alluvium is widely distributed. The Mount Eagle rocks were strongly folded in early Tertiary time and the Kingshill strata gently folded in post Lower-Miocene time along an east-northeast axis. Three early Tertiary cycles of erosion are recognized. After the folding of the Kingshill marl, streams followed the strike of the folded rocks in a westerly direction, but they gradually asrumed southward courses across the marl plain and as a result a western ar^a of old-age topography, a central area of late-mature topography, and an erstern area of early-mature topography have been created. Submerged reefs and emergent reefs and beaches indicate several fairly recent stands of the sea. Water for human consumption is obtained by collecting rain water in cis- terns, but water for other purposes is almost entirely supplied by wells which are generally less than 100 feet deep. Many dug wells are used, but in recent years drilled wells have been constructed. Most of them are discharged by wind-powered pumps of small capacity. Wells are developed in all the rocks mentioned (except coral reef), but the best yields are obtained from the alluvium. A maximum yield of 80 gallons a minute was obtained from'a gravel-packed well in the alluvial valley at Fair Plain. Further exploration of the alluvium is recommended. The weathered diorite also appears tc be a fairly good water-bearing formation. Test drilling showed that deep v^ater- bearing formations should probably not be expected beneath the Tertiary rocks. Most of the ground waters of St. Croix contain a moderately high mineral content owing to the solution of rock-forming minerals and the deposition of alkali and salt spray in the soil. Only a few wells are contaminated ty sea water. The low hardness of some highly mineralized waters is believe! due to base exchange. The most highly mineralized waters are found in the allu- vium in areas with alkali soil and in some places in the Tertiary limestones where presumably soluble salts were deposited in those strata. The least 2 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS mineralized waters are found in shallow wells in the alluvium: near the foot of the mountains and in the areas of dioritic rock. Many well waters in St. Croix, if properly protected from contamination might be entirely suited to human consumption. Although many waters are hard, th^y are used for domestic purposes. Most waters, even those high in chloride, are reported to be excellent for cattle consumption. Most ground waters in St. Croix cannot be used for boiler feed without treatment but are used for other purposes in the manufacture of sugar and rum. A brief discussion of the results of test drilling by the National Park Service in 1940-41 is also given. INTRODUCTION PURPOSE AND SCOPE1OFJINVESTIGATION A program of test drilling for additional water supplies on lands owned by the United States Government, in the island of St. Croix, V. I., was carried out in the period of December 1938 to April 1939 through an appropriation made by the Public Works Administration. These lands have been put to economic use by the Virgin Islands Co., a quasi-government corporation. The serv- ices of the Geological Survey were requested for technical advice, and the writer was accordingly assigned to the project. Sites for drilling were selected, samples of rock penetrated by the test wells were collected, and other duties attendant upon the drilling pro- gram were carried out. In the remaining time available the geology and ground-water resources of the island were studied and a geo- logic map of the island was prepared. (See pi. 1.) An account of the physiography of St. Croix, accompanied by a geologic map, was published in 1941,1 but in that publication much of the detailed geology and most of the hydrologic data were not considered. The investigation was made under the general direction of O. E. Meinzer, then geologist-in-eharge of the Division of Ground Water. M. D. Foster, Nathaniel Fuchs, E. W. Lohr, L. W. Miller, and W. N. Noble made the analyses of water samples. C. S. Ross made petrologic examinations of thin sections of the rocks. T. W. Vaughan and L. G. Henbest made determinations of Foraminifera and W. P. Woodring and T. W. Vaughan made determinations of macrofossils. Cushmanla has published a description of the smaller microfossils found in drill cuttings from the test wells described in this report. During a drilling project carried on by the National Park Serv- ice on St. Croix in 1940 and 1941, data on wells and test holes 1 Cederstrom, D. J., Notes on the physiography of St. Croix, V. I.: Am. Jour. Sci.. vol. 239, pp. 553-576, 1941. lm Cushman, J. A., Tertiary foraminifera from St. Croix, V. I.: U. S. Geol. Survey Prof. Paper 210-A, 1946. INTRODUCTION O were compiled by Donald G. Hazlett working under the direction of H. E. Rothrock. These data are discussed in an appendix to this report. ACKNOWLEDGMENTS The field work was made possible through the kind offices of Mr. Boyd Brown, president, Virgin Islands Co., and willing assist- ance was given by Messrs. C. Hunt, A. N. Gray, and James Tily of that company. Mr. Jensen and Mr. B. Nelthropp supplied much valuable information on wells throughout the island. Mr. Paul Schweitzer, of the Layne-Atlantic Co., who was in charge of the test drilling, supplied equipment for making pumping te?ts on existing wells and later donated several photographs, which are included in this report. A great number of local residents con- tributed information and in other ways helped to make the work in St. Croix a pleasant task. Thanks are due Brent S. Drane of the National Resources Plan- ning Board and his associates for their efforts in making available much data on the Park Service wells. LOCATION The island of St. Croix (or Santa Cruz) is about 95 miles south- southeast of San Juan, P. R., 1,400 miles southeast of New York, 1,000 miles east of Key West, and lies between latitude 17° 41' and 17° 46' and between longitude 64° 34' and 64° 54'. (See fig. 1.) It is one of the Virgin Islands group, of which St. Thomas, St. John, and St. Croix belong to the United States. St. Croix is the largest of these three islands. It is about 21 miles lorg and 6 miles wide at the center and has an area of 84 square miles. St. Thomas, 40 miles north-northwest of St. Croix, has an area of 28 square miles, and St. John, a few miles east of St. TJ omas, has an area of 20 square miles. The town of Frederiksted, with a population of about 2,000, is located on the west end of St. Croix, and Christiansted, with a population of about 3,000, is situated on the north coast in the eastern part of the island. The island has semiweekly steamship service from San Juan, P. R., by which passengers and freight are accommodated. Much freight is also carried by small sailing vessels. Frederiksted is a regular port of call for a passenger steamship line between New York and South America, and irregular calls are made by other merchant ships. A large number of vessels from many parts of the world call at St. Thomas. Weekly airplane service connects St. Thomas and Miami, and more frequent airplane service is avail- able at San Juan. GROUND WATER OP ST. CROIX, VIRGIN ISLANDS TOPOGRAPHY AND DRAINAGE The northwesteni part of the island is a mountainous area flanked on the south by a rolling plain. The mountains are broken by many narrow steep-sided valleys, through which intermittent streams flow and continue south or southeastward across the plain. A few deeply incised streams flow directly westward in the area between Frederiksted and Ham Bluff. The highest peaks on the island, Blue Mountain and Mount Eagle, are 1,090 and 1,165 feet above sea level, respectively. Other peak? attain eleva- INTRODUCTION 5 tions of 700 to 900 feet above sea level. Near Annaly is a very small area of flat upland and near Mount Eagle and Blue Moun- tain streams have carved out two small basinlike depressions. (See pi. 2 A.) Elsewhere high peaks and ridges and steep-sided valleys predominate. On the west near Frederiksted a gently undulating lowland surface borders the sea. In the central part of the islard the lowland is wider and is characterized by rounded hills. West of Christiansted the lowland is submountainous. High hilh and elongated ridges are common, and near Christiansted, where ero- sion has been least effective, a north-south belt of high limestone hills cut by narrow valleys appears superficially to be part of the mountain rather than of the lowland area. That part of the island, extending eastward from Christiansted, locally referred to as "East End," is also mountainous. Here, however, the maximum elevations are less, the stream valleys are not so sharply incised, and intermontane alluvial areas are of more importance. Two lowland areas separated from one another by a low narrow ridge extend inland from Southgate Pond and Great Pond. The northern range of mountains is drained by streams that flow to the west and south. With the exception of Salt River the streams flowing northward are very short and unimportant. Most of the streams discharging from the northern rar miles northeast of Jealousy in the stream bed at the crossroads between Morningstar and Concordia. A small ex- posure opposite the abandoned pumping station about two-tenths of a mile northeast of the crossroads may be of considerable significance. Here, in a small pit, is a succession of buff lime- stone, overlain by. 3 feet of limy conglomerate carrying pebbles 2 inches in diameter, and 1 foot of clayey sandstone, which in turn is overlain by white limestone. These strata strike east and STRUCTURE 31 dip 85°N. The significance of this outcrop is not understood, but it certainly suggests that more deformation has taken place, locally at least, than is generally apparent. From the mouth of Salt River to St. John, limestone of the Kingshill marl lies upon a warped plane that slopes gently to the southwest. As far as could be determined, north of Dolby Hill the plane slopes about 400 feet per mile, or slightly more tran 4°. Between St. John and the Leper Asylum the limestone descends below sea level, but from the Leper Asylum to Cane Garden on the south coast the contact lies above sea level and its location was determined with fair accuracy in a number of places. From the Leper Asylum to Beeston Hill the line of contact rises to an elevation of 350 feet above sea level and lies at a conn: arable elevation on Work-and-Rest hill to the southeast. Exposures on the north side of Work-and-Rest hill indicate that here the plane of contact slopes westward about 530 feet per mile, or 5Va°. Gentle inclination to the southwest of the surface of the base- ment rock cannot account for the low elevation of the lirrestone between St. John and the Leper Asylum, and it is apparent that either the basement rock surface was channeled by erosion or diastrophically deformed before the deposition of the limestones, or the basement rock and the limestones were strongly folded after the deposition of the limestone (Kingshill marl). Data on the elevation of the base of the limestones of the Kingshill in the central part of the island (fig. 3) indicate that these rocks are -250- FIGURE 3. Generalized geologic section showing structure in the eastern part of the Tertiary marl plain. not greatly infolded, and it is therefore believed that they were laid down upon a highly irregular surface. Exposures of limestones of the Kingshill within the central plain are few. Except along the coast line, where the action of waves has produced fine cliff faces, along roads and trails, and in the several small quarries dotting the hillsides naturally occurring outcrops are very scarce indeed. Even there many of the rocks exposed are marly or massive, and determination of their attitude is impossible. Enough data have been obtained to deliner.te the main structural trends in different parts of the limestone plain, 32 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS and the data are considered sufficient to make possible a charac- terization of the plain as a whole. The two small hills east of Golden Grove have been quarried for their blocky limestones, and a more or less continuous series of outcrops girdles each hill a short distance below its crest. These two hills are synclinal and the breach between them, in which New Works lies, is an anticline. (See fig.4.) The axes of N. NW. MT. EAGLE -500' -1000 BLUE MOUNTAIN JEALOUSY FR£DENSBOR6 NEW WORKS FAfR PLAIN EXPLANATION Alluvium Kinqshill marl Jealousy formation FIGURE 4. Cross section of central St. Croix showing location of test wells with reference to geology. the folds appear to extend in a west-northwesterly to east-south- easterly direction. The two hills at Adventure and Fox have a similar origin. The strata are even more gently folded, but synclinal structures are certainly present and form the hills. The axes of folding trend more nearly in a northeast-southwest direction parallel with the Center Line road. At Kings Hill police station the strata dip southeastward. There are not sufficient outcrops along the hill extending to the south to state definitely that the rocks are folded rs to the two localities discussed above, but a reversal of dip (dip to the north- west) near Anguilla suggests folding. Sion Hill to the northeast is also a synclinal hill. TTie structural axis trends northeast-southwestward. A more or less continuous series of observations taken along the road from Bellevue through Rattan to Dolby Fill indicates that Rattan lies within a much broader syncline whose axis trends north-northeastward to south-southwestward. As noted above, this structure was mapped by Quin. Other field observations strongly suggest folded structures but do not define them as conclusively as those already described be- cause exposures are lacking at critical points. The fine series of exposures along Cane Garden Bay show that the strata dip to STRUCTURE 33 the northwest, and it is thought that these rocks form part of a synclinal limb. The prominence on which Cane Garden estate is located may mark the axis of the syncline. Other scattered obser- vations suggest that Work-and-Rest hill and Anna's Hope are on synclinal axes and that the course of the stream flowing be- tween them follows an anticlinal crest. The structure of the area underlain by limestones of the Kings- hill may therefore be characterized as follows: The rocks in the central and eastern part of the area have been compressed into a series of closely spaced open folds, the axes of which generally trend northeast-southwestward, but departures from this trend are seen at Rattan and at New Works. In the eastern and east- central areas the folds plunge from 6° to 14° to the southwest. The axes of the folds at New Works and at Adventure appear to be horizontal. Very little of the structure of the underlying rocks extending from Spratt Hall through Fredricksburg to En- field Green could be determined. The writer agrees with Meyer- hoff,40 who questions the presence of the anticline passing south- west through Hannah's Rest, which was mapped by Quin. In the deeper portions of the Oligocene sedimentation are?., more than 1,400 feet of clays accumulated. The basin gradually filled, and as waters shallowed an extensive growth of coral covered the area of clayey sedimentation. The contact between the clay of the Jealousy formation and the calcareous Kingshill marl is abrupt and appears to denote a marked change in conditions of sedimentation. The line of contact between the limy rocks and the Mount Eagle volcanics from the mouth of Salt River to Cane Garden indicates that the basin was troughlike. The de^th to bedrock is unknown at Little Princess, where the limestone extends below sea level. The trough probably extends west-southwertward under the plain and near Bethlehem. In the central part of the island the base of the Kingshill marl lies 60 feet above sea level at Jealousy, 10 feet above sea level at Castle Burke and Bethlehem, and 180 feet below sea level at Fair Plain. These facts may be taken to indicate that the rocks have not been greatly deformed but were only gently folded in the manner already described. It is believed that if the trough had resulted from deformation after sedimentation the base of the limestones would lie at least several hundred feet lower at Bethlehem and Castle Burke than at Jealousy and Fair Plain. The basal conglomerate of the under- lying Jealousy formation is 60 feet thick, and some of the clay strata higher in the series contain imbedded pebbles and boulders. 40 Meyerhoff, op. cit., p. 164. 34 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS This fact seems to indicate that steep slopes were already present at the time of deposition of the Jealousy formation. The gentle folding of the sediments along west-southwest to east-northeast axes has already been described. Mild compressive forces threw the Tertiary strata into a series of folds in most of which the dips are less than 12°. Steeply dipping recks along the foot of the mountains may be the result of a low dip due to de- formation and a high initial dip. That component of the dip due to deformation may be greater than usual along the foot of the mountains, where the Tertiary strata were forced against a resistant old land mass. It has been noted that the direction of plunge of the axes of folds varies. It has also been brought out that at-Work-and-Rest hill and Anna's Hope the folds plunge to the west-southwest at a minimum of 5°, or 400 feet per mile. If this plunge were unin- terrupted the base of the Kingshill marl at Fair Plain should lie 2,500 feet below its present position. Therefore, the plunge to the west has been interrupted by faulting or cross folding. There are no data at hand that tend to show which of thes? possibilities is more probable. There is very little evidence of faulting, although Quin figures some normal faults, of not more than a foot or two of displacement, which are exposed in the cliffs along Cane Garden Bay, and the writer has noted a small normal fault north of Rattan which trends north-southward and is inclined 65° to the west. However, evidence of the young scarp topography along the north coast shows that faulting has profoundly affected the island, and as a working hypothesis the eastern limestone area is re- garded as a fault block that has been tilted toward the south- southwest. The central and western parts of the island may be relatively undisturbed or may plunge gently in the sr.me direction. PHYSIO GR APH Y The physiography of St. Croix has been dealt with in a previous publication41 and need be only summarized in this paper. EARLY TERTIARY EROSIONfCYCLES The present land surface of St. Croix began to be shaped shortly after the accumulation and deformation of the Mount Eagle rocks of Upper Cretaceous age. In early Tertiary time a land mass much greater in size than the present island of St. Croix was beveled to a rolling plain. Remnants of this surface are seen in tt Cederstrom, D. J., Notes on the physiography of St. Croix, V. I.: Am. Jour. Sci., vol. 239, pp. 553-576, 1941. GEOLOGICAL SURVEY WATER-SUPPLY PAPER 1067 PLATE 3 A. VIEW NORTH-NORTHEASTWARD ACROSS INNER LOWLAND FROM FOX ESTATE. Blue Mountain is in left background. This lowland is largely veneered with alluvium, and here the courses of more deeply buried valleys leading out of the mountains can be determined only by shallow test drilling. B. VIEW ACROSS NARROW ALLUVIUM-FILLED VALLEY SOUTH OF BETHLEHEM. Narrow valleys of this type are excellent areas for wells of moderate yield. GEOLOGICAL SURVEY WATER-SUPPLY PAPER 1067 PLATE 4 A. VIEW SOUTHWARD ACROSS WESTERN PLAIN SOVTHEAST OF FREDERIKSTED. Topographic relief is low, and the locations of more deeply filled valleys leading from the mountains, if present, are not apparent. IT" B. EMERGENT CORAL REEF AT CANE BAY, NORTH SIDE. PHYSIOGRAPHY 35 the high upland at Annaly.42 Eocene (?) uplift and subsequent erosion resulted in a mountainous topography characterized by high relief and moderate ruggedness. The gently rolling basin floors near Fountain (pi. 2A) were created at this time.43 The beveled spurs at elevations less than 250 feet above sea level from Little Fountain at Windsor may likewise have been formed during this cycle of erosion. In figure 5 part of this surface of the SE - DIAMOND BEESTOM HILL WORK AND REST KETURAH SECOND CYCL_C^L-r,'y*!EPL. f //V ~- ?- " VALLEY Of THIRD CYCLE KingshHlmarl FIGURE 5. Section showing development of early Tertiary erosion surfa.ces along eastern edge of marl plain. second cycle is shown to be covered by Tertiary sediments. A still later uplift initiated a third cycle of erosion. This erosional cycle is represented by the channel in the Cretaceous rocks between St. John and the Leper Asylum. The records of wells indicate that the channel extends westward through the central part of the limestone plain (fig. 4). From geological similarities to Puerto Rico, where a more com- plete record is available, Meyerhoff44 postulated that the peneplain of the first cycle was developed in Paleocene-Eocene(?) time, the second partial peneplain in Eocene time, and the third-cycle partial peneplain in lower Oligocene time. In middle or upper Oligocene time submergence occurred, and in the deepest valleys clayey sediments accumulated. As the waters cleared and shal- lowed, great reefs girdled the island, and extensive limy sediments were deposited above the clays. The Oligocene and Miocere sedi- ments thus formed were then gently folded and faulted, ?,nd the island was uplifted. DRAINAGE PATTERN Streams flowing southward from the mountains tended to follow the old valleys, but upon entering the area of folded lime- stone they were deflected to the west-southwest. This west-south- 42 Meyerhoff, H. A., The physiography of the Virgin Islands, Culebra and Vieques: New York Acad. Sci., Scientific survey of Porto Rico and the Virgin Islands, vol. 4, pt. 2, pp. 152-156, 1927. 43 Meyerhoff, H. A., op. cit., p. 100. 44 Idem. 36 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS westward drainage was encroached upon by streams originating along the south coast, and by a series of stream captures the drainage was deflected generally southward.45 Vestiges of the early drainage pattern are present in the eastern part of the limestone plain. LAND FORMS As a result of the drainage pattern outlined above three types of topography have been developed in the area underlair by the lime- stone of Tertiary age. The western area of old-age topography is almost featureless. The central area of late-mature topography is characterized by synclinal valleys and rounded anticlinal hills and ridges, and the eastern area of early-mature topography, is predominantly a region of ridges and valleys trending southwest- ward or south-southwestward. The hard-rock mountain high- lands have resisted erosion and are only a little more advanced in erosional stage than at the beginning of Miocene tirre. Erosion of the gently folded strata which plunge gently south- westward has created steep northeastward-facing scarps ex- tending from Cane Garden to St. John. The distinct terracing seen in places along this scarp has been ascribed to th« alternation of strata of unequal hardness, but possibly some of these high terraces may be due to higher stands of the sea46 for which other evidence has not been recognized. SUBMERGED CORAL REEF The island of St. Croix is almost entirely surrounded by a fringing reef, 2 miles in width along the southeast but generally narrower elsewhere. From Ham Bluff to the mouth of Salt River the reef is very poorly developed along a series of young fault scarps.47 Submerged reefs or banks are also present and h?.ve been dis- cussed by Vaughan48 and Meyerhoff.49 A bank generally from 6 to 10 fathoms deep and about 1 mile wide skirts the present reef 46 Cederstrom, D. J., op. cit, p. 567-571. 46 Meinzer, O. B., Geologic reconnaissance of a region adjacent to G^antanamo Bay, Cuba: Washing-ton Acad. Sci. Jour., vol. 23, pp. 256-258, 1933. Woodring, W. P., and others, Geology of the Republic of Haiti: Dept. Public Works, p. 374, 1924. Stearns, H. T., Ancient shore lines on the island of Lanai, Hawaii: Geol. Soc. America Bull., vol. 49, pp. 615-628, 1938. 47 Meyerhoff, H. A., op. cit, p. 181. 48 Vaughan, T. W., Some littoral and sublittoral physiographic features of the Virgin and northern Leeward Islands and their bearing on the coral reef problem: Washington Acad. Sci. Jour., vol. 6, p. 56, 1916. "Meyerhoff, H. A., op. cit, pp. 170-175. PHYSIOGRAPHY 37 along the south coast. For 10 miles east of St. Croix this platform broadens and deepens and lies generally 11 to 16 fathoms below the surface. On the eastern extermity of this submarine plat- form Lang Bank forms a higher elongate crescent, the terminals of which point westward and enclose East Point, the eastern extremity of St. Croix. Lang Bank lies 8 to 10 fathoms below the surface and appears to have been a barrier reef at one stage in the development of the island. A still lower reef is also present. It is very narrow and lies generally 12 to 20 fathoms below the surface and rims the next higher bank. The narrowness of this bank suggests that it is covered in part by the higher and younger bank, which is quite well developed. Meyerhoff has suggested that the higher bank was largely formed by subaerial erosion.50 Beyond the reefs the sea floor drops steeply. Along tY "mZWliim « « « * A. ARTIFICIAL CATCHMENT BASIN AT MARY'S FANCY. Windmill pumps water from underground cistern. Old wind-powered cane-mill tower and stack of more recent steam-powered cane mill is at right. Photograph by Paul Schweitzer. B. MUNICIPAL WELL, WATER GUT, CHR1STIANSTED. GEOLOGICAL SURVEY A. OLD WELL TOWER AT STRAWBERRY. Modern wind vanes have been installed. Photograph hy Paul Schweitzer. WATER-SUPPLY PAPER 1067 PLATE 6 B. ROTARY RIG AT TEST WELL NO. 2 NEAR JEALOUSY. Note proximity of well to mountain at left. Photograph by Paul Schweitzer. OCCURRENCE OF GROUND WATER 51 obtained at this place by drilling additional wells or by construct- ing a well in such a" way as to provide a greater intake area. If the weathered diorite is too rotten to allow firm seating of the casing at a depth of about 20 feet, it might be advantageous to install perforated or slotted pipe of a smaller diameter than the hole below 20 feet and fill the space between the hole and casing with sized small gravel. It is also possible that fissures and cracks which would further increase the capacity of wells developed in the dioritic rock might be encountered at greater depths, and additional wells in this formation should be carried to a de^th of about 300 feet unless it is found by careful testing that increases in yield do not occur with increases in depth of hole or that water from depth is highly mineralized. It is reported that in the drought of 1940 this well continued to yield about 2,000 gallons a day. In view of the pumping test referred to above it would seem likely that the installation of a low-capacity power pump on this well and substitution of a 3- or 4-inch pipe line for the existing ^-inch line would be profitable. By this means the supply at Bethlehem Factory might be increased from 5,000 to 10,000 gallons a day. At Southgate a drilled well (100) 80 feet deep probably draws upon weathered diorite. It is believed that the diorite here is as good a water-bearing formation as the rock penetrated by Kobel well, but no data on the capacity of the well are at hand. How- ever, it is reported to have furnished the wind-powered pump with ample water in the 1940 drought. There is good reason to believe, considering the record of the Kobel well and the performance of wells in coarsely crystalline rock areas in general, that a moderate quantity of water could be developed by properly constructed wells in the Fountain Basin and to a lesser extent in the Hermitage Basin to the east. Like- wise it may be expected that water in moderate quantity could be obtained from properly constructed wells penetrating the dioritic rock in the region around Southgate and southeast of Sally's Fancy in East End. [GROUND WATER IN THEfKINGSHILL MARL About 25 wells in St. Croix obtain water from the limestone of Tertiary age which makes up the plain between Frederiksted and Christiansted. The wells vary considerably in depth. A few wells less than 20 feet deep may obtain water from the limy rod's, but most of the wells are deeper and at least two wells are mor? than 200 feet in depth. Most of the wells are dug wells and eight of the 52 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS dug wells are more than 100 feet in depth. The dug well at Con- stitution Hill is more than 225 feet deep. The wells are protected near the surface by natural stone masonry but are not cased below the surficial alluvium. In recent years a few 8-inch open-end wells have been drilled. The limy Kingshill rocks lie at lower elevations n^ar the coast than they do at the foot of the mountains, but details on the structure of the central plain are incomplete and rlthough it is believed that only one to two specific strata yield the water en- countered in wells, this cannot positively be determined. Water in wells in the Tertiary limestones stands from a few feet to 20 feet above sea level near the south coast. Northward, water generally stands at higher elevations, and along th^ foot of the mountains the water level in times of normal rainfall may lie within relatively few feet of the surface. At Mary's Fancy, in the higher area in the eastern part of the plain, the static water level is as much as 250 feet above sea level. Water-bearing limestones of the Kingshill marl are character- ized by small solution holes at Paradise (well C-5). Rock of this type has not been reported in other wells. Strata with small solu- tion holes have been noted in a very few places in road cuts and stream banks. It is possible also that in many wells water is derived from openings along bedding planes and fissures which have resulted from structural deformation. The water-bearing limestones of the Kingshill marl may be largely replenished along the foot of the mountairs by surface streams and by underflow from the mountains. Recharge across the limestone plain is probably hindered in many places by a cover of clayey soil, but elsewhere direct downward percolation of rain- water may take place. A detailed description of representative wells is g:ven below. At Frederiksted seven dug wells (9a-9g) supply the greater number of the residents with water for some domestic uses. The wells along the shore are only 10 feet deep, but those on higher ground are deeper. Along the coast the water level is within one- half foot of sea level, but at a distance of 800 feet east of the seashore the water level may be as much as 4 or 5 feet above sea level. As the water here is a body of fresh water floating on sea water it might be expected that there would be danger of salt- water contamination, particularly during those times when the water-bearing formations receive little fresh-water recharge or when relatively large amounts of water are taken from the well. OCCURRENCE OF GROUND WATER 53 An analysis of water from well 9g (table 5) shows that more magnesium is present than calcium and suggests that the mineral- ization is in part due to the mixing of fresh water with sea water. However, the high nitrate content of this water indicates con- siderable pollution, and it is more likely that the chloride is the result of pollution rather than mixing with sea water. The water level in the dug well (12) at Camporico, on the limestone "plain southeast of Frederiksted, stands a few feet above sea level. The well is reported to be a dependable sourc? of a moderate amount of water even in periods of severe drought. At Whim (14), Concordia (13), and Hope (15), shallow wells in which the water level stands within a few feet of the surface in periods of normal rainfall are developed in the limestone area. However, these wells are too shallow to be dependable in times of drought. Thus far, according to local reports, no efforts have been made to deepen these wells at such times of low water levels. It should be noted that an attempt to secure an adequate water supply by a deeper drilled well, made in the vicinity of Hop% was unsuccessful. It is reported that the well was carried to a depth of more than 100 feet, but no water-bearing strata were en- countered. A drilled well 42 feet deep at Cain (22) obtained an entirely adequate supply of water for stock. The water level stood within 2 feet of the surface when the well was visited by the writer. The well isjocated at an elevation of 90 feet above sea level within one-half mile of the foot of the northern range of mountains. This well is reported to have been abandoned in 1940, but it is not known whether clogging or excessive lowering of the water table was the cause of failure. The well at Diamond (24), 90 feet deep, lies at a slightly greater distance from the mountains than the well at Cain. Water rises in this well to within 32 feet of the surface or about 53 feet above sea level. At Enfield Green, which is Jocated about midway between Dia- mond and the south coast, water rises to within 31 feet of the surface, or probably less than 10 feet above sea level. With respect to location and water level, it is similar to the well at Camporico. Data on the maximum yield of these wells is not available. The well at Diamond (24), however, is reported to continue tc yield an adequate supply even in periods of drought. The amount dis- charged from the well is limited by the capacity of the wind- powered pump and probably at no time exceeds 3 gallons a minute. The dug well 37 feet deep at Paradise (29) is reported to yield less than 2 gallons a minute. 54 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS A spring issuing from limestones at Envy (31) along the south coast has a flow estimated to be 30 or 40 gallons a minute. The occurrence of this spring suggests the presence of channels in the limestones capable of yielding moderately large quantities of water. The well (32) at Manning Bay, less than one-half mile north- east of Envy, is probably dug in alluvium. However, the presence of the spring at Envy suggests that the alluvium here may be receiving water from the limestone rocks rather than from the underflow passing down the valley at Fair Plain. Wells in the limestone rocks are also present along the south coast at Jerusalem (61) and Cane Garden (62). The water level in these wells is approximately 221/2 to 18J/2 feet atove sea level, respectively. The former well is reported to yield an ample supply in periods of drought. A number of wells in the more easterly part of the plain obtain water from the Tertiary limestone for domestic use and for stock. Well 54 at Mary's Fancy extends to an elevation of 190 feet above sea level, and water rises in the well to 93 feet below the surface or to about 245 feet above sea level. At other places the wells reach strata nearer to sea level and water does not rise as high, generally less than 100 feet above sea level, and, as indicated above, as the coast is approached the water may stand less than 20 feet above sea level. At Sion Farm it is reported that, with the present installation, well 68 yields 2,000 gallons a day even in periods of severe drought. Mr. Robert Skeoch reports that for the past 30 years the well at Diamond (55) has always yielded an ample supply, estimated as at least 2,000 gallons a day. The well extends nearly to sea level, and the water stands probably as much as 35 feet above sea level in the well in seasons of abundant rainfall. The Ruby well (57), one-half mile north of Diamond, is 103 feet deep and extends to about 100 feet above sea level. Water stood at 94^/2 feet below the surface (110 feet above sea level) when the we1! was visited by the writer. If the wells at Diamond and Ruby are developed in the same ground-water body, the slope of the water table is more than 110 feet per mile. A similar apparently high gradient is present in the vicinity of Enfield Green and Cain. Certainly this is a very high gradient, and it seems possible that parched bodies of ground water exist above the body of fresh water occurring a short distance above sea level. OCCURRENCE OF GROUND WATER 55 There is reason to believe that in wells located some distance from the coast, where the water level is normally from: 50 to 100 feet above sea level or higher, the water level varies widely from season to season and year to year. Measurements or est*mates of depth to water in a number of wells were made by Mr. H. Rydeen on June 4 and 5, 1919, in conjunction with a report on the water supply of the island by T. W. Vaughan. At Strawberry, where the water level stood at 116 feet above sea level when measured by the writer in December 1939, the water level vras re- ported by Rydeen to have been 27i/2 feet higher in June 1919. The water level at Ruby stood about 116 feet above sea level in 1939 and was reported by Rydeen to have been 201^ feet higher in June 1919. Rydeen also notes that at Constitution Hill the water level was 190 feet from the surface or 110 feet above sea level. In 1939 this well was dry at 225 feet below the surface. On the other hand, only slight differences in the depths to water in coastal wells at Jerusalem and Cane Garden are noted. Ry- deen's figures indicate that the water levels in these wells border- ing the coast were 3i/> and 7% feet higher in 1919 than in 1939. It is further reported by Mr. Hazlett, Associate Geologist of the National Park Service, that on January 1, 1941, the water levels at Jerusalem and Diamond (55) were 3i/^ feet lower than when measured in December 1939, on which date water levels stood at 221/2 and 35 feet above sea level, respectively. Measurements by Hazlett were taken during a period of severe drought. It is re- ported that at Spanish Town the water level declines only a foot or so in times of drought. The water level is normally only a few feet above sea level. The chloride content of the waters in these wells range? from 150 to 950 parts per million except for two wells, which ar^ even higher in chloride content. The quality of water will be discussed in detail in another chapter. A few wells (72, 74) are developed in the Tertiary limestone rocks along the coast northwest of Christiansted. The drilled well at La Grande Princess may encounter limestone beneath a shallow alluvial cover. It is reported that in 1936 the water level was 3 feet below the surface, or about 37 feet above sea level. The dug well at La Grande Princess is located at a somewhat higher elevation. This well is 33 feet deep, and water stood at 28 feet below the surface, or about 27 feet above sea level. The well had not been pumped in 3 months. In the unused dug well at St. John the water level is 321/2 feet below the surface, or about 4 feet above sea level. 56 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS The wells at La Grande Princess yield highly mineralized water. It appears probable that the mineralization is due tc salts in the water-bearing formations rather than to contamination by sea water since the water levels are quite high above sea level. In general, the Kingshill marl is a poor water-bearing forma- tion, although several wells with very low yields are of consider- able value in furnishing water to livestock. GROUND WATER IN THE ALMJVIUM NORTH SIDE Only a few wells from the mouth of Salt River westward are developed in the Quaternary deposits. In one well (77), located just south of Greig Hill on low swampy land, water stands slightly above sea level. This well is pumped every day by a low-capacity gasoline-powered pump and furnishes ample water for cattle. The water is highly mineralized. A second shallow well (78), pumped by a wind-powered pump, is located on the coast north- west of Greig Hill. As closely as could be determined, water stood about 1 foot above sea level in this well. The chloride content of the water is relatively low 378 parts per million. EAST END Several shallow wells (95-97) are located at the east end of Altoona lagoon at slight elevations above sea level. Water levels stand as high as 6 or 7 feet above sea level in these wells. When these wells were visited the chloride content of the water from two wells was relatively low, but the chloride content of the water from a third well was about 950 parts per million. The reason for this difference is not understood, but it is thought that since the windmill pump was operating in the third well and not in the first two, more highly mineralized water was being drawn into this well from the immediately adjacent salt pan bordering Altoona lagoon. Another well (98) in this general area, just south of Darby Hill, yields water containing more than 1,000 parts per million of chloride. In a dug well (99) at All-For-The-Better, at the head of the broad valley carved in dioritic rock embracing Southgate Pond, water stood 29y2 feet below the surface, or about 60 feet above sea level. This well is probably developed in the alluvium but may reach the weathered bedrock beneath. OCCURRENCE OF GROUND WATER 57 Two wells are present in a similar topographic and geologic situation at the head of the broad southward-facing valley em- bracing Great Pond. In well 109, 35 feet deep, the water stands 33 feet below the surface, and in periods of drought the water level falls below the bottom of the well. Well 108 is a drilled well 76 feet deep and obtains water from sandy strata about that depth. In constructing this well, much of the sand was cas^d off to prevent sand clogging. The chloride content is somewhat higher in the shallower well water, being 544 parts per million. In a shallow well (101) at Green Cay, along the north coast, water stands about 2 feet above sea level. This well supplies more than 300 cattle with water even during periods of drought. The chloride content is relatively low. This well may be in the path of greatest underflow down the vflley, otherwise it might be ex- pected that the mineralization would be higher. The well (102) at Mary's Fancy is poorly situated on th« east side of Cotton Valley and yields a water containing abort 600 parts per million of chloride. If it were located more westerly in the path of greatest underflow the water might be lower in chloride. Three wells are located at Great Pond estate. Well 105 is de- veloped in a narrow strip of beach sand just west of Mount Fancy. The well near the great house (106) is developed in alluvium and encounters rock at 24 feet below the surface. The water here has a chloride content of more than 1,700 parts per millior. The water level had declined at the time the well was visited to within a foot of the bottom of the well, and the yield of the well was de- creasing. A third well (107), 40 feet deep, is in the broad valley one-half mile northwest of the great house. Water stood 6% feet from the surface at the time the well was visited. The well is not used and is reported to furnish only a limited amount of water. In summary, small supplies of rather mineralized water are obtained from shallow wells developed in alluvium along the. coast in the East End area. In some of these waters (wells 78, 95, 96, 97) contamination or the danger of contamination by sea water is apparent. Other well waters (wells 70, 71, 77, 98) are highly mineralized because of salts in the sediments from which the water is taken. Wells (93, 94, 101) which are located on or near the main channels of underflow from large drainage areas yield more dependable supplies of water, and the water obtained is relatively low in mineral content. Water levels fluctuate more widely in response to local seasonal rainfall in the wells on higher 58 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS ground. Wells that are too shallow (109) go dry in periods of drought. CENTRAL PLAIN The alluvium that occupies the valleys and broac1 depressions in the central marl plain supplies fair amounts of water of rela- tively moderate mineral content to a number of shallow dug wells and to a few deeper drilled wells. In most of these wells the water level is within 15 feet of the surface. Further, most of the wells are so shallow that in times of drought the water level may fall below the bottom of the well and the well go dry. The deeper drilled wells are generally depend- able sources of water even in times of drought. North of Frederiksted three wells take advantage of the under- flow coming down from the large drainage area in the mountain- ous western part of St. Croix. The wells at Willir.ms (4) and Wheel of Fortune (10) supply water for domestic use and for cattle. The chloride content of the water in the well at Williams is only 112 parts per million but is 410 parts in the veil at Wheel of Fortune. In the dug well just south of La Grange factory (8) the water level was 8 feet below the surface, or about 30 feet above sea level. This well, 15 feet deep, is reported to yield 200 gallons a minute for periods of several hours. It seems likely that this area just north of Frederiksted, which is at the mouth of a valley carrying a fairly large urderflow, is a particularly favorable locality for the development of moderate to large supplies of water. The underflow coming down the valley varies, of course, but except in periods of protracted droughts it may be as much as several hundreds of gallons a minute. If moderately large amounts of water are to be pumped in this area, as for a municipal supply for Frederiksted, care should be taken not to draw the water level down to sea level, as a general lower- ing of water levels will tend to bring in salt water and endanger the quality of the supply. For maximum yield and efficiency and to lessen the danger of chloride contamination, any moderately large quantity of water discharged should be pumped from sev- eral widely spaced wells rather than that the total amount be taken from one well. It seems evident that a fairly large supply of water might be also obtained between Spratt Hall and Williams sir^.e this area too receives underflow from a drainage area extending eastward to Annaly. From Whim to Mount Pleasant very little is know:^ concerning OCCURRENCE OF GROUND WATER 59 the thickness of the alluvium and its quality as a water-bearing formation. Most of the wells obtaining water from the alluvium here are shallow and fail in periods of drought. However, if favor- able thicknesses of alluvium are found in this area, it seems likely that supplies of water may be obtained which would be an im- provement over many of the existing supplies. Dependable sup- plies of small quantities of water are reported to be obtained from wells at Cain (22), Diamond (24), and Enfield (23). Although these wells are said to penetrate the marls and limestone of the Kingshill it is conceivable that much of the water available to them has been carried down from the mountains as underflow in the alluvial cover. It seems possible that deep alluvium-filled channels in these underlying rocks from which somewhat larger amounts of water could be obtained might be located. Areas re- ceiving the greater part of the underflow from the mourtain valleys terminating at Grove Place, St. George, and below Cane Valley will probably be the most favorable places to prospect. The fact that streams from the mountains become dry ever be- fore they reach the alluvium but reappear as the coast is ap- proached indicates that subsurface flow is normally of importance in this area. The inner lowland from Plessen to Mon Bijou and the valleys extending southward from the inner lowland to the sea are filled with a variable thickness of alluvium. In this area three shallow wells, four deeper wells, and two springs furnish water for several villages and for cattle. The streams discharging across the alluvial plain from the mountains are lost beneath the surface a short distance below the mouth of the valleys. Water appears in their courses from place to place; for example, the stream near Fair Plain flows on the surface. A short distance below the South Side road the flow was roughly measured at 50 gallons a minute. This measurement was taken late in March in the dry season. A short distance below the place at which the above mentioned measurement was taken the surface flow increased considerably. The discharge south of the South Side road represents only a small part of the water taken into the ground at the foot of the mountains inasmuch as important losses are sustained all along the stream courses by evaporation where the water level is at or close to the surface. Further, it is known from logs of wells that deeper sands and gravels, separated from the surface sands by relatively impervious clays, are present in the major valleys traversing the central plain, and these undoubtedly carry off much of the drainage from the mountain area. 842683° SO S 60 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS At Plessen (28b) and near Castle Burke (34b) rrater is pumped from springs issuing from the surface sands. The spring at Plessen yields a water containing 316 parts per million of chloride. It is said never to go dry. At the Castle Burke sump from 10 to 20 gallons a minute is pumped continuously in seasons of nor- mal rainfall, but in periods of drought the yield is considerably less. However, since water (in somewhat smaller quantities than normal) was available from the Castle Burke sump in the 1940 drought, it seems indicated that the water levels may not fall very greatly in this vicinity in periods of deficient rainfall and that moderate quantities of water should be available to properly located shallow wells in this area at all times. As the water from Castle Burke sump is of very low mineral content (see table 10) and as ground water occurs near the surface e^en in times of drought, this area should be explored to determine the location of the deepest part of the ancient-filled valley leading southward from Upper Love. Here the greatest thickness of coarse sands will be found, which should provide an excellent supply of water to properly screened wells. A dug well, not more than 20 feet deep, is located along the stream bed near Fredensborg. Normally this well yields about 7 gallons a minute per 8-hour day. In times of drought, how- ever, it is practicable to pump the well only about 3 hours a day. The chloride content of the water is high but not excessive, being 452 parts per million. The extent of variation in water-level is illustrated by several measurements made on the Golden Grove dug well. Rydeen noted that on June 2, 1919, the water level was 10 feet below the surface. On December 9,1938, the water level war 14i/2 feet below the surface. This was in a season of normal light rains. Dur- ing the following dry period, water levels fell, and on March 25, 1939, the water level was about ISVs feet below the surface. Hazlett reports that in January 1941 the water level in the well had declined to about 24 feet below the surface. It seems apparent that since the level of ne&r-surface water is subject to wide variations in this locality (and perhaps in other localities some distance from the foot of the mountains) every effort should be made to continue shallow wells through the water-bearing sands to the'underlying impervious formation, either clay or marl, in order that water be available to pumps at all times. This can be accomplished easily by drilled wells using casing and screen but may be impossible with dug wells. OCCURRENCE OF GROUND WATER 61 Deeper sand strata are developed in the axial part of the valley between Fredensborg and Fair Plain. Drilled wells at New Works (44) and at Fair Plain (45b) encounter coarse gravel about 58 feet below the surface. Water is present under artesian head and rises to within 14 feet of the surface in the latter well. These wells are not screened, nor are they developed to the fullest ex- tent possible. Yet the latter well, at least, yields about 3 gallons per minute per foot of drawdown, and both wells continued to yield ample water to low-capacity windmill pumps during the drought of 1940. Plate SB is a view across the alluvium-fillec" valley south of Bethlehem. A test well (45a) was drilled within 100 feet of the well at Fair Plain discussed above. The log of this well shows that in addition to near-surface water occurring under water-table conditions, two strata of artesian water-bearing gravels and sands occur from 57 to 62 feet and from 70 to 73 feet below the surface, respec- tively. The near-surface water, which occurs in fine-to-coarse sand extending from the surface to a depth of 33 feet, stood at an elevation of 21 feet below the surface (about 2 feet above the level of the nearby stream), but the water from the deeper strata rose to within 17 feet of the surface. A pumping test was made on well 45a after it had been fully developed. The results of this test are shown in figure 9 ard may be briefly summarized as follows: Before pumping began the static water1 level was 16.83 feet below the top of the casing; at the end of 231/2 hours of pumping at 61 gallons a minute, the water level was 30.02 feet below the top of the casing. Tl ° rate of pumping was increased to 70 gallons a minute, and pumping at this rate was continued for 2414 hours, at the end of which tirne the water level was 32.31 feet below the top of the casing. The rate of pumping was increased to 80 gallons a minute and continued at that rate for 2i/2 hours. The water level fell to 35.33 feet below the surface. The rate of pumping was decreased to 53 gallons a minute and maintained at that rate for 17y» hours, at the end of which time the water level was 26.S9 feet below the surface. The specific capacity of the well (gallons a minute per foot of drawdown) was indicated to be 5.09, 4.52, and 5.55, respectively, in the several pumping tests. The chloride content of the water fell from 515 parts per million to 435 parts per million during the first 23% hours of pumping. When the rate of pumping was increased to 70 gallons 62 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS ^ // 1 / //\ 1 s tsa -t-> A on S ^ Its of pi " § a> R> 03 6 1s * t o> S tl0£ 03. IO £ 5 U33J) ONISVO JO dOl M0138 Oi Hld3Q o o o 00 * O '109) oo dO 3JLtfb 1N31NOO IQlbOlHO OCCURRENCE OF GROUND WATER 63 a minute llhe chloride content rose to 660 parts per million, de- clined to 565 parts per million, and then increased to 710 par^s per million at the end of 48 hours of continuous pumping. The chloride content was still about 700 parts per million at the con- clusion of the 70-hour test. It seems likely that, as a cone of depression is created around the well, water from other sources, either from the limestones or from the near-surface alluvium, is brought into the well causing an increase in the chloride content of the water. Analysis of a sample (table 10) collected in April 1940 shows that the water contained 500 parts per million of chloride at that time. The dug well (32) at Manning Bay is 10 feet deep. According to a field test the water contained 685 parts of chloride on December 8, 1938. In 1939 Manning Bay was taken over by the United States Army. The well was used as a source of water supply, but it was found that when the withdrawal increased, the chloride content increased from 500 to 960 parts per million. It is stated that for several months, at least, the well was pumped at a rate of 30 gallons a minute, a remarkable yield for a dug well in St. Croix. Very few wells are developed in the alluvial valley system discharging at Hope and Blessing. A well (50) at Lareine 42 feet deep supplies water to a wind-driven pump, as does a well 28^ feet deep at Hope (60). It is of interest to note that the chloride content of the water from the well at Hope was relatively low, about 110 parts per million. The Castle Coakley well (63) was thought to be developed in limestone, but the low chloride content (79 parts per million) of the water suggests that it may be drawing water from the alluvium. The drilled well at Anna's Hope (64) is 180 feet deep anc1 may penetrate the weathered bedrock. However, it seems probable that the overlying alluvium is contributing a large part cf the water available to the well and the well should be considered as being developed in Quaternary material. The well is said to yield ample water to a windmill pump even in times of protracted drought. Several wells are located along the coast northwest of Christian- sted. Two shallow wells at Little Princess are reported to yield small amounts of water at all times. In normal seasons water stands about 5 or 6 feet from the surface. One of the wells (69) yields water containing about 700 parts of chloride, but the 64 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS other (70) yields water containing over 1,800 parts of chloride. The water from well 70 is used for drinking purposes occasionally by some of the native Cruzians and is described by them as "heavy" but quite palatable. It seems established from the record of wells in the alluvium, particularly the Fair Plain developed well (45a), that there exists in St. Croix a potential supply of moderately large quantities of water which has hitherto been barely touched. It seems likely that the storage capacity of the alluvial sands and gravels may be relatively large and that with proper development several hun- dreds of gallons of water a minute might be easily obtained from the alluvial sands and gravels in the valley north and northeast of Fair Plain and in similar narrow alluvial valleys. Further, one or both of the deeper water-bearing strata may be present in relatively deep channels leading seaward from the mountain gorges across the alluvial lowland. Here the mireral content of ground water is lower than at localities nearer the sea. The location of these channels is not so apparent wher^ the valleys are broad and in the general area north of Center Line road (pi. 3A), and it is not possible to predict their location, but their presence may be established by inexpensive test drilling with a light rotary or jetting rig. GROUND WATER IN BEACH SAND A well at Cane Bay (80) and another at Great Pond (105) obtain water from beach sand. Beach sand, including emergent beaches, is of limited extent in St. Croix except at Southwest Cape and cannot be counted on for more than a limited quantity of water. A well in beach sand generally yields a rather highly mineralized water, owing in large part to excessive salt spray and possible contamination by sea water resulting- from over- pumping. At Cane Bay water stands 1^2 feet above sea level. Although containing more than 1,000 parts of chloride the water is satis- factorily used in a cattle dip. At Great Pond water from beach sand contains over 5,000 parts of chloride but it said to be entirely satisfactory as a source of drinking water for cattle. TEST-DRILLING PROGRAM Through funds made available by the Public Works Administra- tion, a program of test drilling on lands controlled by the Virgin Islands Co. was initiated in the fall of 1938 to determine the presence or absence of deep water-bearing strata in the area OCCURRENCE OF GROUND WATER 65 underlain by mid-Tertiary limestone. The wells were to be drilled on the property of the Virgin Islands Co., and a depend- able supply of water for use at the Bethlehem sugar mill was to be obtained, if possible. The writer was assigned to the project and selected the sites for drilling. The drilling itself (pi. 65) was done by the rotary clay seal method by the Layne Atlantic Co., of Norfolk, Va. A study of the geology in the area north and northwest of Bethlehem Factory indicated that the marl and limestone of the Tingshill rested upon the older Mount Eagle volcanics, as had b^en described in the literature. The actual contact between the two formations was not displayed, and the dip of the surface of older rock plane was not known. However, it was thought that the plane nrobably sloped more or less gradually seaward in a mariner analogous to the basement rock beneath the Atlantic Coastal Plain. A site was selected for test drilling just east of Bethlehem Factory and slightly less than 1 mile south of the contact be- tween the Kingshill and Mount Eagle rocks. It was hoped that information on the following points might be forthcoming upon the completion of the test well: 1. The presence or absence of water-bearing beds in the Ter- tiary limestones at greater depths than had been previously attained by wells in St. Croix. 2. The determination of the water-bearing characteristic? of the basal conglomerate. 3. The presence or absence of water-bearing alluvial material or weathered bedrock beneath the basal conglomerate. In the first test well (41, table 3, and pi. 1) 18 feet of soft yellow marl was first penetrated, beneath which lay 7 feet of s^.nd and gravel. The next 83 feet of material consisted of soft (but probably indurated) yellow marl of the Kingshill exactly similar to that seen outcropping in many places on the island. However, at a depth of 108 feet a greenish-gray clay (blue when wet) was encountered, and at a depth of 1,508 feet from the surface the drill was still in this same material. Hard, thin layer? of limestone, not more than 1 foot thick, occurred between 728 and 802 feet and between 1,291 and 1,448 feet, and a stratum of hard limestone conglomerate 16 feet thick was penetrated between 1,095 and 1,111 feet below the surface, a similar stratum occur- ring between 1,196 and 1,201 feet below the surface. The near-surface sand and gravel was water-bearing and was being drawn upon by a nearby well, and hence no attempt was made to develop this stratum. The underlying limestone and clays were entirely barren of water-bearing beds. 66 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS It was apparent that the bedrock surface did not slope gradually seaward but plunged sharply downward from the contact near Jealousy, and, accordingly, to obtain the information sought in this program it was necessary to select a site nearer the foot of the mountains. The location for the second test was one-eighth of a mile west of estate Jealousy and 1,000 feet soMth of the con- tact of the Kingshill and Mount Eagle rocks. In the second test well (39, table 3, and pi. 1) yellow marl ex- tended from the surface to a depth of 91 feet. Hard layers of limestone, each one only a few inches in thickness, were encoun- tered in the yellow marl. Below the marl gray clay of the Jealousy formation was penetrated which extended to a depth of 398 feet. The gray clay here contained much more detrital material than was found in the section at the first test well. A stratum con- taining boulders was encountered between 224 and 228 feet. Between 245 and 255 feet the clay contained many pebbles and small boulders, and from 255 to 398 feet below the surface the clay contained a variable but small amount of sard and pebbles. Cores of this material showed that the inclination of the bedding planes was not constant and the dip of individual strata ranged from 12° to 35° within a foot o'f core. The presence of thin cross-bedded sandy strata suggests that these are not true dips but are the result of cross bedding on a larger scaK Hard limestone conglomerate extended from a c1c*pth of 398 to 459 feet below the surface. It is exactly similar to the rock seen cropping out in the stream bed a short distance northwest of the well. The conglomerate was well cemented and, r.s far as could be determined from cores, it was not characterized by solution cavities or fractures. Weathered bedrock was encountered beneath the limestone con- glomerate and cored for 11 feet, making the total depth of the test well 470 feet. The weathered bedrock was fractured but in the samples brought up the fractures were sealec1 by secondary deposits of calcite (lime). Water was found only in the basal part of the limestone ex- tending to a depth of 91 feet below the surface. The yield was less than 16 gallons a minute, and the water contained more than 1,600 parts per million of chloride. No water was yielded by the conglomerate or by the basement rock. The information obtained by these test wells may be sum- marized as follows: 1. The limestones and marls of the Kingshill, which are seen at the surface in many places, are of limited thickness and OCCURRENCE OF GROUND WATER 67 are underlain by a grayish clay of unknown maximum thickness. The limestones and marl rocks penetrated in these wells contained no stratum which* under any conditions might be considered a particularly good water-bes ring formation. The clay, as might be expected, yields no water. 2. The basal conglomerate, where not weathered, is a tight and impervious formation and will probably not yield water anywhere. 3. Bedrock lies at relatively great depths a short distance from the base of the northern range of mountains. Although fractured and open where weathered, it may become im- permeable through the deposition of mineral matter by percolating ground waters. The weathered bedrock has proven to be an excellent water- bearing formation in several places, as at Little Grange, Fobel (north of Jealousy), Hermon Hill, Longford, and Southgate. Whether fissures and cracks tend to be filled with secondary mineral matter at depth in most places or only in very few places is not known, but in any event the weathered bedrock, and pos- sibly also the fresh harder rock beneath if encountered in wells at shallow depths, should be considered as possible sources of at least small supplies of water. This point has been previously discussed, but is repeated here to avoid the implication that the evidence gained in the second test well is applicable to all r.reas underlain by hard rocks in St. Croix. It was decided to develop a well at a selected location wher^ the alluvium would most likely be of maximum or nearly maximum thickness and subject to continuous and ample recharge. A site was selected (45a) near the existing drilled well (45b) at Fair Plain. In this place the surface and subsurface drainage of a wide area extending up into the mountains is concentrated in a narrow channel. The existing well demonstrated the presence of other than near-surface water-bearing formations. In addition it was felt that the location was far enough removed from the sea to avoid salt-water contamination. In addition to near-surface water-bearing sands, water-bearing strata were encountered from 57 to 62 feet and from 70 to 73 feet below the surface. The relatively large yield of water obtained from these strata as compared to-yields of other wells on the island has been discussed in a previous chapter. Yellow marl lay below the alluvial sands and clays and extended to a c'epth of 162 feet, and alternating harder limestones and soft narls extended to a depth of 208 feet below the surface, at which c^pth 68 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS gray clay was encountered. The well was continued to a depth of 225 feet. No water-bearing strata were encountered in the limy rocks. Tiie test well was converted into a gravel-packed well and at present is supplying water to the Bethlehem Factory. It is reported to have furnished 18,000 gallons a day during the severe drought of 1940 and was pumped continuously at a rate of 40 (?) gallons a minute in 1943. TABLE 3. Logs of test holes drilled in St. Croix, 1938-39 Well 41, Virgin Islands Co. at Fredensborg [Altitude, 90 feet] Alluvium (Quaternary) : Kingshill marl (Miocene-up- (?) Oligocene): Greenish-gray clay (thin, hard layers of limy rock at 728, 732, 762, 770, 778, and 802 feet).-.... Thickness (feet) 18 4 3 83 987 Depth (feet) 18 22 25 108 1,095 mentedbylime. Clayey Gray clay with hard streaks at 1,291, 1,305 to 1,307, and 1,448 Thickness (feet) 16 85 5 305 Depth (feet) 1,111 1,196 1,201 1,506 Well 39, Virgin Islands Co. at Jealousy [Altitude, 150 feet] Kingshill marl (Miocene-up- per (?) Oligocene): Yellow marl (contains hard, limy layers, each a at 81, 83, 84, 85, 88 feet); Jealousy formation (upper (?) Oligocene) : Gray and black clay con- 91 133 4 17 91 224 228 245 Jealousy formation Cort. taining a considerable amount of rounded pebbles and sms.ll Gray clay containing a amount of sand and Hard basement rock.___- 10 143 61 11 255 398 459 470 Well 45a, Virgin Islands Co. at Fair Plain [Altitude, 30 feet] Alluvium (Quaternary) : Fine sand gradually be- Clay somewhat sandv or Clay. - .. Sand and gravel, water.. .. 33 24 5 8 3 33 57 62 70 73 Kingshill marl (Miocere- upper (?) Oligocene): Yellowish to white marl . 89 5 8 10 13 17 162 167 185 195 208 225 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS 69 QUALITY OF WATER COMPARISON OF SURFACE WATER WITH GROUND WATER Samples of surface waters that were taken at Two Friends and at Jolly Hill may be characterized as moderately hard bicarbonate waters containing less than 400 parts per million of total dis- solved solids, in which sulfate and chloride are low. So far as the two available analyses may be used as bases for characteriza- tion of St. Croix surface waters, the waters are rather simple chemically and about what would be expected from the solution of limestone rocks by percolating waters. Sodium and chloride may have originated in greater part as ocean spray carried over the islands by the constant winds. The ground waters are in part derived from the infiltration of surface streams but are more highly mineralized than the surface waters. Most of the ground waters of St. Croix thf.t .were analyzed contained from 400 to 2,400 parts per million of total dissolved solids, and a few waters had more than 4,000 parts per million. The higher mineral content of the ground waters may be due to solution of some of the constituents of the materials through which the waters percolate. A brief study of the r.nalyses of the ground waters shows, however, that the increase is not a simple proportionate increase of the elements found in 8. simple calcium and magnesium bicarbonate water, but, as shown graphi- cally in figure 10, in proportions of constituents present, the ground waters generally contain a high, sometimes very high, content of sodium chloride and may also contain moderate amounts of sodium sulfate. Analysis C in figure 10, a sample of surface water taken near the mouth of Salt River appears more like some of the ground waters analyzed than like either of the two other surface-water samples. Analyses from Castle Burke sump and Kobel well con- tain less sodium salts than most of the other ground waters analyzed because these waters have not traveled any considerable distance underground and have passed through relatively in- soluble rock material only. CONTAMINATION BY SEA WATER One of the first questions which arises when the quality of the underground waters of an oceanic island is considered is the relation of the fresh water to the salt sea water. There- fore, this relation will be considered in detail. It has been common knowledge for many years that fresh 70 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS water falling upon a sandy coast or upon a small sandy island will percolate into the sand and move laterally toward the shore to mingle ultimately with the sea. It might be supposed that in such places the salt water surrounding the island or adjoining the coast would penetrate the sand to mean sea level and immedi- ately absorb all the fresh water that might percolate downward to its surface. For several physical reasons this does not happen. Such islands are found, in reality, to contain a dome-shaped lens of fresh water floating upon a concave surface of salt water, as shown in figure 11 A. The fresh water is enabled to float upon the salt water because the fresh water has a considerably lower density. This principle was first applied to the hydrology of seacoasts by Baden Ghyben61 in 1887 as a result of investigations made in Holland and was also noted by Herzberg62 of Berlin in 1900, and in more recent years has been applied in the United States.63 In figure 11, A, the weight of the higher column of fresh water, H, should be equal to the weight of the smaller column of salt water displaced, h. Thus it has been found that in places where the law is applicable, for every foot which fresh wrter extends above sea level it also extends approximately 40 feet below sea level. Thus, in a well along the coast at St. Croix where the water level in wells stands 3 feet above sea level, salt water should be encountered in that well at 120 feet below sea level. However, a corollary of this principle is that if the height of fresh water above sea level (t, in fig. 11, A) is reduced 1 foot, then the salt-water boundary moves upward 40 feet, as shown in figure 11, B. Consequently, if wells along the coast are heavily pumped, water levels in the immediate vicinity of the wells may be lowered to sea level or below sea level and cause an upward intrusion of salt water into the wells. In most wells in St. Croix the water level stands more than a few feet above sea level and the danger of salt-water con- tamination appears to be negligible. There are a few places along the coast where ground waters have a high chloride con- 81 Baden Ghyben, W., Nota in verband met de voorgenomen put. boring nabig Amsterdam: K. Inst. Ing. Tydschv., p. 21, 1888-1889, The Hague, 1889. 42 Herzberg, Baiirat, Die Wasserversooging einigir Nordseebader: Jour. Gasbelench- tung und Wasser versorging, Jahrg. 44, Munich, 1901. 83 Stearns, H. T., Clark, W. O., and Meinzer, O. E., Geology and water resources of the Kau district, Hawaii: U. S. Geol. Survey Water-Supply Paper 616, pp. 10-21, 1930. Stearns, H. T., and Vaksvik, K. N., Geology and ground-water resources of the Island of Oahu, Hawaii: Div. Hydrography, T. H., Bull. 1, pp. 346-364, 1935. Barksdale, H. C., Sundstrom, R. W., and Brunstein, M. S., Supplementary report on the ground-water supplies of the Atlantic City region: New Jersey Water Policy Commission, Special Kept. No. 6, pp. 10-26, 1936. EQUIVALENTS PER MILLION \ \ \ mm:^t m 10.- Graphic representations of analyses of some St. Croix grrouna waters and surface waters. 72 GROUND WATER OF ST. CROIX, VIRGIN ISL/NDS Nonpumping well \ Water table Pumping well Water table water B FIGURE 11. Diagram showing- relation between salt water and fresh water in homo- geneous water-bearing sands making up a small island: A, Under norpumping condi- tions ; B, Where pumping is taking place. tent; these places are (1) in the city of Frederiksted, and (2) along the coast northwest of Christiansted. A few wells else- where may also be contaminated by sea water. TABLE 4. Analyses of ground waters in Leper Asylum well, St. Croix, show- ing chloride contamination and analyses of ocean water and of water from the Tertiary limestone area for comparison. [Results in parts per million] Silica (SiO2)_---__ ____-_-,______-----__-_- Sulfate (SOO-- ---------------------------- Chloride (Cl)_ ______ ----------------------- Range of composition of ground waters in Tertiary limestone area 1 21 to 49 4 .3 to 1.7* 10 to 133* 6.9 to 101 4 615 to 971 02 to 500 84 to 1,985 Tr to 25 0 to 1 .44 48 to 747 812 to 4,4864 6 6 Lep?r Asylum3 12 12 38 64 / 1,639 I 4.4 712 638 1,895 1.3 1.0 357 4,708 EWL 7 July 7, 1936 Ocean water3 /fl Q 1,300 10,710 390 150 2,690 i Q ^3 cn 6,380 35,000 1 Wells are Bethlehem, Bonne Esperanee, Constitution Hill, Sion Farm, Profit, and spring at Envy. (See table 11.) 2 Well 81a. 3 Mean of 77 analyses of ocean water collected by Challenger Expedition. 4 Not determined in some analyses of wells in Tertiary marl. 5 Sodium and potassium content of waters from wells at Sion Farm and Bonne Esperanee (1940 sample) were calculated. « See table 8. ' EWL, E. W. Lohr. QUALITY OF WATER 73 In Table 4 ground water from the Leper Asylum well (81a), believed to be contaminated by sea water, is compared to ocean water and to water from wells in the Tertiary limestone area. In this water magnesium is high in relation to calcium, in addi- tion to having a high chloride and fairly high sulfate contert. The analyses of water samples collected in 1919 from wells 77 and 80, located near the mouth of Salt River and at Cane Bay, respectively, suggest that little if any contamination by sea water had taken place at that time. However, field determinations made in 1939 (see table 11) show that the chloride content of the water from these wells is now much higher than it was in the samples collected for T. W. Vaughan in 1919. Well 77 yielded water containing 344 parts per million of chloride in 1919, but in December 1938 the chloride content was almost 2,000 parts per million. Likewise, the chloride content of well 80 at Cane Bay increased from 303 to over 1,300 parts per million. Well 77 discharged daily by a low-capacity gasoline pump. Well 80 supplies water to a cattle clip and is discharged somewhat less heavily by a hand pump. Both these wells are located at low elevations near the sea. As complete analyses of the wr.ters yielded by these two wells in 1939 are lacking it is not possible to determine definitely the cause of the increased chloride con- tent, but it is suggested that they may have been contaminated by sea water as a result of moderate discharge by pumping. Wells 105 and 106 at Great Pond are located at low elevations near the sea and are pumped more or less continuously at a low rate. These wells yielded water containing 5,150 and 1,740 parts per million of chloride, respectively, in January 1939. It is suggested that here too the wells are probably somewhat con- taminated by sea water. The other ground waters of St. Croix for which analyses are at hand either show reasonably low chloride content or are of such chemical character or so located that contamination by sea water is most unlikely. Therefore, since it has been shown that prob- ably only a few wells in St. Croix are contaminated by sea'water, the source of the relatively high mineral content of waters from many wells calls for some other explanation. CHLORIDE CONTAMINATION BESULTING FBOM ORGANIC POLLUTION The analysis of the water from well 9c (table 5) in FrederiJ-sted shows a chloride content of 3,549 parts per million and, consider- ing the location of the well near the coast, it might be thought that this represents mixing with sea water in part. However, 74 GROUND WATER OP ST. CROIX, VIRGIN ISLANDS the analysis also shows that the water has a ver;r high nitrate content, generally an index of organic pollution, and it is probable that the very high chloride content has been derived from the same source. Further, the proportion of magnesium to calcium is lower than would be expected if sea-water contamination had taken place. Well 9g at Frederiksted is located on higher ground some distance from the sea. When the analysis is examined, it is seen that the total mineralization is only about half that of the water from well 9c, near the shore, although it is still high. Chloride is less than half that in the water from well 9c, 1,609 parts per million. The nitrate content, 57 parts per million, is less, and it seems likely that here the high chloride content may be "ascribed partly to organic pollution but the high ratio of magnesium to calcium and the very high sulphate content indicate sea water contamination. It may be added that the water in well 90 at No. 1 Strand Street, Christiansted, which contains 507 parts per million of chloride as compared to 204 parts per million of chloride in well 91 in Water Gut, also shows high organic pollution since the analysis shows a nitrate content of 305 parts per million. TABLE 5. Analyses of waters from, wells showing organic pollution [Results in parts per million] Silica (SiO2) ___ ...... _ . _ - ___ ---- -- Sodium (Na) ____ - _____ ... ___ ... ------ Sulfate (SO4) ------------ _ ------------ Chloride (Cl) ___ - __ ..-- __..- ____ .- Fredericksted Well 9c 26 .60 401 123 } 2,119 0 390 273 3,550 291 1,507 7,443 NF June 6, 1919 Well 9g 23 .67 83 115 1,099 0 57i3 63.^ 1,609 57 68^ 3,64,5 NF June 6, 1919 Christiansted Well 90 25 .17 86 65 552 0 656 175 507 305 482 2,060 MDF June 4, 1919 1 NF, Nathaniel Fuchs; MDF, M. D. Foster. SALINE MINERALS ASSOCIATED WITH LIMESTONE OF TERTIARY AGE It has been found in St. Croix that in the area near Jealousy and Bethlehem the limy rocks yield a salty water. An analysis (43, table 8) of water from the 245-foot well drilled near Bethlehem in 1934 (?), shows that the water had a total mineral QUALITY OF WATER 75 content of 4,486 parts per million with 1,985 parts per million of chloride. A 105-foot well drilled one-fourth mile southwest of Jealousy yielded water containing over 2,400 parts per million of chloride. This water was obtained from a cased weF in the marl (Tertiary) at a depth of 90 feet below the surface. Another water with high chloride content, 1,700 parts per million, was obtained from well 39 nearby at a comparable depth below the surface, and it seems evident that in this area the marl contains salt. Marly strata containing saline minerals may be responsible for the high mineral content of water obtained from wells at Bonne Esperance and Paradise and are undoubtedly a contribut- ing factor to the generally high mineral content of water from wells in many other places in the limestone plain. EFFE C TSZOFjALKALIfS OILS It is believed that the solution of alkali salts which have been concentrated at the surface in practically all low-lying parts of the island accounts in large part for the mineral content of many of the ground waters of St. Croix. In the discussion of the occurrence of ground water it was noted that where the zone of saturation lies close to the surface, the capillary fringe may extend to the surface. Where this condition exists, ground water is continually being brought up to the surface and evaporated. Evaporation is undoubtedly high as cool ocean breezes are warmed in their passage across the sunny island, and their drying effect, coupled with periods of low rainfall, is evi- denced by the type and degree of verdure of the vegetation. The mean evaporation at St. Croix from 1920 to 193564 was as follows: TABLE 6. Mean monthly evaporation at St. Croix from 1920 tc 1935 Month laches 4.81 5.05 6.52 7.09 7.06 , .. ,., Month July....... _ ..... Inches 6.74 7.01 6.96 6.12 5.66 Month Total _ ___ .. Inches 4.64 4.65 72.31 As evaporation proceeds, the salts carried in solutior by the ground waters are deposited at or near the surface and in time appreciable concentrations of mineral matter are built up. The concentration of salts at the surface varies inversely Y"ith the rainfall. As rain falls upon the surface and percolates into the 64 Johnson, Arthur P., op. cit., p. 64. 842683° SO 6 76 , GROUND WATER OF ST. CROIX, VIRGIN ISLANDS ground it tends to dissolve these salts out of the earth and decrease their concentration at the surface, but in so doing th?. mineral content of the water in the zone of saturation is increased. Thus, under conditions of high evaporation and intermittent rainfall a process of concentration of salts at the surface and Ir.ter trans- ference of these salts to the zone of saturation may go on progres- sively where the ground-water levels are close to the surface. In the evaporation of ground water the water itself passes off into the air, but the greatest part of the dissolved chemical con- .tent remains in the surface soil and is deposited in lari;e part as less soluble calcium and magnesium carbonates and sulfates, and as highly soluble sodium carbonates, bicarbonates, sulfates, and chlorides. The concentration of these salts increases with time. During periods of heavy rain, the water percolating into the soil dissolves some of the salts deposited there and transports them downward into the zone of saturation in relatively concentrated solutions. The amount of calcium and magnesium corbonates and sulfates dissolved is probably not great and is limited by the amount of free carbon dioxide available. However, the sodium (and potassium) salts are freely soluble, and water passing through alkali soils containing these salts becomes heavily charged with them. By this process, the ground water maintains its calcium bicarbonate content relatively unchanged but becomes more highly mineralized in soaium (and potassium) snlfate and chloride. In a reconnaissance soil map of St. Croix, Thorp65 har delimited areas of alkali soils. Many alkali-soil areas are found at low elevations along the coast; for example, from Fredricksted to Spratt Hall on the west coast, many places along the south coast from Hope to Cotton Grove, along the north coast at Rustoptwist, and in patches from the mouth of Salt River to Cottongarden. The alkali soils are largely confined to areas below canyon mouths, or surrounding mouths of rivers, or small embayments of the sea. However, they are not entirely confined to the coast. Areas of alkali soils also border several streams and extend as much as 2 miles inland, as from Enfield Green to St. George, from Fair Plain to New Works, from the mouth of Salt River to Glynn, from Surlaine Point nearly to Catherine's Rest, and surrounding Great Pond and Southgate Pond and extending inland to Mount Wash- ington and All-For-The-Better. Other areas of alkali soil occur inland as isolated patches sur- rounded by unaffected or alkali-free soil. Those which have been 65 Thorp, James, op. cit. QUALITY OF WATER 77 plotted are in the vicinity of the spring at Plessen, between Castle Burke and Lower Love, and along the streams west and north of Bethlehem. EFFECTS OF SALT SPRAY In a great many places the concentration of salts at the surface by evaporation of ground water will be augmented by salt spray blown for considerable distances inland. Evidence of the im- portance of salt spray is seen in the fact that cistern water, which had been collected from the roof at Central Factory near Chris- tiansted, was found to contain 26 parts per million of chloride. Legget66 states that "as. evidence of what rainwater may contain may be mentioned the fact that for England the average chloride content is about 2.2 ppm . . ." and "rain falling at Land'? End during strong southwest winds, blowing in from the sea, har been found to contain one hundred times this amount." Vaksvik67 notes that on Oahu, Hawaii, all natural waters contain salt. "As no point on the island is more than 12 miles from the ocean, all of it is reached by fine ocean spray or minute particles of salt that remain suspended in the air when the spray evaporates. The material is blown over the land, and some of it is brought down by rains." A table is given showing the rainwater collected dur- ing showers at various points on Oahu contained from 6 to 42 parts per million of chloride. Thorp68 notes that white alkali, mostly sodium chloride and sodium sulfate, is "most concentrated near the numerous lagoons, and the salt concentration rapidly decreases toward the inland/' whereas the black alkali, sodium carbonate, and bicarbonate, "was found affecting a few fairly large areas near Great Pond, the Leper Asylum, Southgate, Bethlehem, Castle Burke, and Jerusalem." This appears to indicate that along the coast the effect of salt spray upon the soil is dominant, whereas at more inland localities the salts in the soils have largely resulted from the evaporation of ground water. BASE EXCHANGE Inasmuch as some highly mineralized soft waters found in the Tertiary limestones are sodium bicarbonate waters rather than 66 Legget, Robert F., Geology and engineering: McGraw-Hill Book Co., New York, p. 459, 1939. H7 Stearns, H. T., and Vaksvik, K. N., op cit, p. 345. 68 Thorp, James, op. cit., p. 19. 78 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS sodium chloride waters, the sodium may best be explained as having resulted from base exchange.69 By the process of base exchange, hard bicarbonate waters which come into contact with certain materials bearing exchange- able sodium, exchange their calcium and magnesium for equivalent amounts of sodium. No change in the total amount of dissolved solids occurs. There seems to be no reason to believe that the marine marls forming the greater part of the Tertiary marl plain could not furnish exchangeable sodium to hard bicarbonate waters passing over them. SUMMARY The mineral content of most St. Croix ground waters is de- rived from (1) the solution of rock-forming minerals, (2) salt spray blown inland from the ocean and carried down into the zone of saturation by rainwater, and (3) concentration by evaporation at the ground surface of salts in ground water, originating from (1) or (2) above, and subsequent solution and addition of these concentrated salts to the zone of saturation below by rainwater. A very few wells have been contaminated by sea water, and a few owe a major part of their mineral content to organic pollution. QUALITY OF WATER IN THE VARIOUS FORMATIONS MOUNT EAGLE VOLCANIC8 No analyses of waters from the few wells obtaining water from the Mount Eagle rocks are at hand. However, field tests for chloride were made of waters from a number of wells penetrat- ing these rocks and a brief discussion may be based on these determinations. In general, those wells which penetrate the Nount Eagle volcanics yield water with a moderate to high chic ride content. The lowest chloride content, 172 to 250 parts per million, was found in water from wells located at high elevations at Orange Grove (83), Hermon Hill (85), and Catherine's Rest (117), whereas wells encountering bedrock beneath alluvium at lower elevations generally have a somewhat higher chloride content. Drilled wells at Longford (114) and Nugent (112) yield water containing more than 600 parts per million of chloride, and the dug well at Hartman (111), likewise entering the zcne of weath- 69 Renick, B. C., Geology and ground-water resources of central an-? southern Rose- bud County, Mont.: U. S. Geol. Survey Water-Supply Paper 600, pp. 41-42, 1929. Foster, Margaret D., The chemical character of the ground waters of the South Atlan- tic Coastal .Plain: Washington Acad. Sci. Jour., vol. 27, No. 10, pp. 405-411, 1937. QUALITY OF WATER 79 ered rock, yields a water containing 1,275 parts per mill'on of chloride. In these wells at Longford, Nugent, and Hartman, how- ever, it is believed that the source of mineral content is in the overlying alkaline soils. A dug well on the mountainside at Barons Bluif (79) yields water containing 340 parts of chloride. Inas- much as these cliifs lie in the path of the prevailing wind* they are in a position to receive much salt spray, and the ground water here may be high in sodium chloride and relatively low in other constituents. DIORITIC ROCKS Only two wells in St. Croix are known to be developed in the dioritic rock. Water from the Kobel well (38), north of Jealousy estate, contains 647 parts of total dissolved solids, as shown in the following analysis made in the spring of 1940: TABLE 7. Analyses of water from well at Kobel in dioritic rock [Results in parts per million] Silica (SiOs) _ _ .. . ___ ...... Iron (Fe) .. . . ......... Sodium (Na). .... __ ......... Bicarbonate (HCOs) __ _ .. Carbonate (CDs). _ ...- _ . ... . 572.9 AQ 50 104 3.0 398 26 Sulfate (SOO . __ ... _ - _ Chloride (Cl)---.-. ...------- Nitrate (NO»)__ ---------- ___ 18 148 .05 .4 378 647 May 10, 1940 It is a hard bicarbonate water in which the magnesium con- tent in only slightly less than the calcium. Sodium and chloride are farly low as compared to most other St. Croix ground vraters. In a pumping test made on this well on February 3, 19f9, the chloride content had increased from 160 to 278 parts per million after pumping 20 gallons a minute for 6 hours. This ir crease may be due to the flow of more highly mineralized water from near the surface into the well, induced by the pumping. The well (100) at Southgate in eastern St. Croix also penetrates dioritic rock and yields water containing 280 parts per million of chloride. As the well is located in a "slightly alkalin0" soil area, it would seem that the water did not enter the well through the immediately overlying soil. KESTGSHTLL MARL The following table contains analyses of waters from wells developed in the Tertiary limestones of the plain extending from Frederiksted to Christiansted. An analysis of water from one spring issuing from the limestones is also included. These wells, with the exception of the well at Castle Coakley, are locr.ted in 00o TABLE 8. Analyses of waters from wells in the Kingshill marl [Results in parts per million] Silica (SiO2) ------- _ ---- ___ -- _ ---- _ -------- Sulfate (SO4) _ ------ _ .. _-__ -------- _ ------- Chloride (Cl) ....-...-_-__.-_----......-..----.--- Envy1 42 1.6 42 35 442 0 655 130 375 9.3 229 1,421 MDF June 4, 1919 Bethlehem2 21 1.7 133 101 1,385 0 615 442 1,985 .4 1.4 747 4,486 EWL July 7, 1936 Anguilla3 27 .12 61 39 426 0 551 139 428 45 312 1,472 MDF June 4, 1919 Profit4 *23 1.0 26 8.9 478 0 . 812 89 269 Tr 101 1,320 MDF June 4, 1919 Bonne Esperance5 49 .3 10 10 673 0 971 189 364 25 66 1,842 MDF June 4, 1919 Bonne Esperance5 91,391 890 10500 1,400 20 1.6 "216 MDF, WMN, LWM April 15, 1940 Castle Coakley6 755 180 79 21 .8 "26 MDF, WMN, LWM April 15, 1940 Constitution HilF 23 7t; 7t; 21 6 9 287 0 622 62 84 14 81 MDF 1919 Sion Farm8 9510 j 658 10170 97 z. 5.1 0 "48 MDF, WMN, LWM April 15, 1940 1 Spring. 2 Well 43. s Well 46. 4 Well 47. 'Well 53. 6 Well 63. 7 Well 67. " MDF, M. D. Foster; WMN, W. M. Noble; LWM, L. W. Miller; EWL, E. W. Lohr. 8 Well 68. 9 Calculated. 1 By turbidity. ^.^ u Determined. QUALITY OF WATER 81 places where seepage of water to the wells from Quaternary alluvial deposits seems unlikely, and hence the analyses are con- sidered as truly representative of water from the Tertiary lime- stones. Brief inspection of the analyses (table 8) shows that tvro rather well denned types of water are yielded by the Tertiary rocks. One type of water, yielded by a spring at Envy and by wells at Anguilla, Bethlehem, and Bonne Esperance (1940 sample), is a hard water in which the total hardness ranges from 216 to 747 parts per million and in which sodium is present in large part as sodium chloride and sodium bicarbonate. The bicarbonate con- tent falls within about the same range as that of the softer waters from the Kingshill, however. It is interesting to note that the sample taken from Bonne Esperance well in 1940 is much harder than the water yielded in 1919, but in the 1940 sample the bicar- bonate content is less and the additional hardness appears to be present as carbonate hardness. There is also much more sodium chloride. The other type of water is yielded by wells at Prof t, Bonne Esperance (1919 sample), Castle Coakley, Constitution Hill, and Sion Farm. It is relatively soft and contains generally less than 100 parts per million of total hardness and from 551 to 970 parts per million of bicarbonate. The Castle Coakley sample, contain- ing only 26 parts per million of total hardness, is soft IT almost any standard but may not be representative. This type may be classed, therefore, as soft to slightly hard sodium bicarbonate water. The sample from the Bethlehem "salt well" and the 1940 sample from Bonne Esperance are high in chloride 1,985 and 1,400 parts per million, respectively but the chloride content of other samples analyzed ranges from relatively low to moderate. The sulfate ranges from 62 to 500 parts per million and is highest in those samples containing a high chloride content. Botl sulfates and chlorides are thought to have their source in soluble salts in the limestone which were deposited syngenetically. The difference in mineral content of the samples collected from the Bonne Esperance well in 1919 and 1940 is believed due to the lack of recharge in a period of low rainfall and consequent inflow of highly mineralized water, induced by more or less continuous pumping at a low rate, from limestone areas contain- ing high concentrations of soluble salts. 82 GROUND WATER OF ST. CROIX, VIRGIN ISLArDS AULUV1UM WEIIS FROM AILUVIUM WITHIN AREAS OF MOUNT EAGIE VOICAWICS Analyses (table 9) indicate that the mineral content of waters from wells located in Quaternary alluvial deposits within areas of Mount Eagle rocks differs widely, those analyzed ranging from 659 to 2,329 parts per million total dissolved solids. It is thought that the difference in total mineral content may generally be ascribed to the location of wells with respect to drainage channels along which relatively large underflow takes place. It seems apparent from the analyses at hand and from field determinations of chloride content that in wells located along main drainage lines the waters are less highly mineralized than those from wells in other locations. It is believed that the relatively grea ter amount of water moving along such channels, both on the surface or underground, tends to reduce the concentrations of raits.in the alluvium along those channels, or that wells in such locations receive larger accretions of less highly mineralized surface waters than wells located elsewhere, or both. Wells at Anna's Hope (64) and Christiansted (91) are located along drainage lines. These wells have a mineral content of 659 and 896 parts per million total dissolved solids, respectively. Conversely, water from the Longford well, located in the more arid eastern part of the island and not along a main line of drainage, is nore highly mineralized and contains 2,329 parts per million of total dis- solved solids. Well 90 at Christiansted is located a short distance away from the minor drainage channel in which well 91 is situated. Water from the former well contains more than twice the amount of total dissolved solids found in water from the latter. The waters from these wells are quite hard with the exception of the water from the Orange Grove well, which has a hardness of 87 parts per million. The hardness ranges from 173 to 519 parts per million. The waters may be classed as hr.rd calcium and magnesium bicarbonate waters containing variable amounts of sodium salts. In a few waters sulfate is present as calcium sulfate. The chloride content of the waters mentioned above ranges from 56 to 794 parts per million. A field determination of chloride content shows that the water yielded by the well at Salt River may be more highly mineralized at times than is indicated by the analysis in table 8. The increase in dissolved constituents may be due to contamination by sea water as the well is pumped, but the greater part of the mineral content is thought to be derived locally from the adjacent "salt pan" or alkali-soil areas. TABLE 9. Analyses of waters from wells in the alluvium within areas of Mount Eagle volcanics [Eesults in parts per million] Silica (SiOi)-- ------------------- Chloride(Cl)--. ----_--------- - Nitrate (NOs) ____ . ____ _____ .. Anna's Hope1 65 .10 40 37 \ 134 502 0 20 56 38 252 659 MDF 1919 Anna's Hope2 35 .93 30 24 / 814 I 10 828 0 351 625 17 1.6 173 2,300 WMN April 15, 1940 Salt River' 35 .13 82 48 \ 234 449 0 50 3444.3 402 1,040 NF June 3, v 1919 . Cane Bay4 28 .25 106 62 253 672 77 303 11 519 1,218 NF 1919 Orange Grove5 "369 698 0 80 1652.1 1237 WLL March, 1933 Chris tian- sted6 25 .17 86 65 552 656 175 507 305 482 2,060 MDF 1919 Christian- sted7 24 .11 82 39 188 489 71 204 12 365 896 MDF 1919 Green Cay* 57 .62 54 41 319 606 77 283 24 303 1,176 NF 1919 Cotton- grove9 34 .76 93 46 410 499 108 544 27 421 1,520 MDF 1919 Long- ford10 32 .23 53 43 780 868 151 794 33 309 2,329 MDF 1919 'Well 64. 2 Well 65. s Well 77. * Well 80, « Well 82 or 83. « Well 90. ' Well 91. « Well 101. Well 104. 10 Well 115. "Calculated. 12 Determined. ls MDF, M. D. Foster; WMN, W. M. Noble; NF, Nathaniel Fuchs; WLL, W. L. Lamar. 00 CO CO TABLE 10. Analyses of waters from wells in the alluvium within Tertiary limestone areas [Results in parts per million] Silica (SiOa)-- -- -- - ----- - - Carbonate (CDs) __ --- __ ____ __ -. ____ Chloride (Cl).... ._______.....__.__-____.__...-___ Nitrate (NO8). ___ ._ _ . ..... .. __ ..... ... . Plessen1 362 20 2001.6 .8 "195 MDP, WMN, LWM April 15, 1940 Golden Grove2 47 .13 50 33 286 709 0 74 1529.6 260 1,031 MDF, * WMN', LWM June, 1919 Castle Burke3 46 .18 80 39 95 526 .6 10 86 .16 1.2 360 595 EWL July, 1936 Castle Burke3 530 3 92 .75 .8 207 April 13, 1940 Fredens- borg4 771 "60 3901.0 1.2 "273 MDF, WMN, LWM April 15, 1940 New Works5 50 .04 21 17 528 748 28 132 3358.8 1.6 122 1,497 EWL July 7, 1936 Fair Plain6 36 .22 42 29 435 714 123 318 12 .04 224 1,175 EWL July 7, 1936 Fair Plain' 40 102.3 64 41 f 514 \ 3.2 662 0 157 500 121.0 328 1,631 WMN April 15, 1940 Lareine8 688 9300 510 15 .4 »i!05 MDF, WMN, LWM April 15, 1940. 1 Sprins "WK 2 Well 33. s Sump 34b. * Well 42. » Well 44. 6 Well 45b. 7 Well 45a. 11 Determined. 1J EWL, E. W. Lohr; MDf, M. D. Foster; WMN, Y7. M. Noble; LWM, L. W. Miller. >> Well 50. By turbidity. 10 In sediment. QUALITY OF WATER 85 WELLS FROM ALLUVIUM WITHIN TERTIARY LIMESTONE AREAS The waters from wells and springs in the alluvium within the areas of Tertiary marls, for which analyses are shown below (table 10), generally contain more than 1,000 parts per million of total dissolved solids. The least highly mineralized water is obtained from springs in which water has traveled a relatively short distance underground through surficial gravels, whereas water obtained from somewhat deeper wells a greater distance from the foot of the northern range of mountains may contain a considerably greater amount of total dissolved solids. The analyses indicate that the waters are moderately hard or hard bicarbonate waters generally containing fairly high sodium bicarbonate, chlorides, and sulfates. The sample from the Castle Burke sump had a carbonate hard- ness of 360 parts per million and no noncarbonate hardness. The nonhardness-forming constituents are relatively low, and the total dissolved solids are only 595 parts per million. The mineral content of this water may be derived in large part through simple solution of the surficial sediments free of excessive alkr.li or other salts through which this water passed. The water from the spring at Plessen also has a relativel;r low mineral content; more than half the bicarbonate is present as calcium and magnesium bicarbonate, and the remainder is present as sodium bicarbonate. The chloride content is fairly high 200 parts per million. Inasmuch as the spring is located in an alkali- soil area, as noted by Thorp70 it is believed that much of the mineral content of the water is derived from the solution of those salts from the soil. The samples from New Works and Lareine have a relatively high content of total dissolved solids, yet the hardness is relatively low, only 122 and 105 parts per million, respectively. It is sug- gested that here too base exchange may have been an important process in creating a fairly soft high sodium bicarbonate water. However, the moderately high chloride and sulfate content of these waters suggests that there has been some solution of alkali salts from the soil. The proportions of sulfate and chloride in samples from the wells at New Works, Golden Grove, Fredensborg, Fair Flair, and Lareine suggests that the solution of salts from the alkal ; soil may have been effective. 70 Thorp, James, op cit., p. 19. 86 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS BEACHfSAND Water from well 105 at Great Pond contained over 5,000 parts per million of chloride when visited by the writer. It is thought that this well is overpumped and has been contaminated by sea water. (See fig. 11, B.) At Canebay a field determination in- dicated that the water from well 80 contained more than 1,000 parts of chloride, although an analysis of a sample collected in 1919 (included in table 9 with waters from wells in the alluvium, for convenience) contained only about 300 parts of cl loride. The increase may also be due to sea-water contamination. SUITABILITY OF WATER FOR VARIOUS USES HUMAN CONSUMPTION Vaksvik71 has pointed out that, although most people can detect the presence of salt when th-s chloride con- tent exceeds 200 parts per million, water containing considerably higher quantities of chloride can be consumed by human beings without harmful effect. People who are accustomed to drinking water containing as high as 350 parts per million of chloride have stated that they prefer their own water to that of the Honolulu municipal supply (chloride ccntent from 45 to 72 parts per million), which they consider tasteless. Several families of Japanese living on the west side of the entrance to Pearl Harbor use waters from a spring and several shallow dug wells that have a chloride content of 950 to 1,100 parts per million. In several places in Tidewater Virginia water containing from 200 to 400 parts per million is used for drinking purposes by municipalities or military posts. Water from one well near Nor- folk contains 1,080 parts per million of chloride and is said to be used from time to time by some individuals. In St. Croix one individual stated that water from well 69 at Little Princess was occasionally used as drinking water. This water contains more than 1,800 parts per million of chloride. It seems probable that the water from many wells in St. Croix would be quite satisfactory as drinking waters. In r. great many wells the chloride content is moderate, and these waters would probably not be unpalatable. It may be pointed out, however, that many wells may be polluted, as the nitrate content, which is derived from the de- composition of organic matter, is high in waters frorr many wells. Waters containing 1 or 2 to 45 parts per million of nitrate are generally from wells where cattle are watered, but nitrates in excess of 200 parts per million are found only in wells in the towns. Conversely, it is interesting to note that ritrate is ex- 71 Stearns, H. T., and Vaksvik, K. N., op. cit., p. 345. SUITABILITY OF WATER 87 tremely low in waters from wells located in cane-field areas, namely, Fredensborg, the sump north of Castle Burke, and the spring at Plessen. Only 0.05 part per million of nitrate was found in the water from the well at Kobel. CATTLE Well No. 105, developed in beach sand adjacent to the ocean at Great Pond, yields water containing more than 5,100 parts per million of chloride. This water is the sole source of drinking water for a herd of cattle. According to the owner of the cattle, this water is of excellent quality for this purpose and hs.s been used successfully for many years. IBBIGATION The question of suitability of ground waters in St. Croix for irrigation is a pertinent one. Large volumes of water ?,re not available for the purpose, but on the other hand smaller volumes which might be used to irrigate small tracts of sugar care, par- ticularly in seasons of severe drought, may be obtained in some places. Scofield72 has discussed the limits of permissible sulfste and chloride content in irrigation waters and has stated that 480 or more parts per million of sulfate or 355 or more parts per million of chloride in irrigation water are unsafe. In regard to sulfate content the water from only a fer? wells exceeds 480 parts per million, notably at Cottongrove and1 Long- ford in the East End area and at Bonne Esperance (1940 srmple). However, with respect to chloride the water from many yrells is not considered desirable, notably Cottongrove, Longford, Bonne Esperance (1919 and 1940 samples), Lareine, Fredensborr, Fair Plain (developed well), Enfield Green, Paradise, Castle Burke, and others in various parts of the island. In Hawaii73 water containing more than the permissible 355 parts per million set by Scofield is used to advantage although excessive chloride content of water has a bad effect on sug-ar cane, the amount that it will tolerate varies with the type of soil, the terrene, and the quantity of water applied to the fields. Several plantation officials have mentioned figures ranging from 700 to 900 parts per million of chloride as the maximum permissible without impairing the yield of sugar. On one plantation it was found that best results were obtained when the wter dis- charged from a group of wells containing as high as 850 parts per irillion of chloride was mixed with high-level spring and tunnel water very low in 72 Scofleld, C. S., Quality of irrigation waters: California Dept. Public Works, Div. Water Resources, Bull. 40, 1933. 78 Stearns, H. T., and Vaksvik, K. N., op. cit., p. 345 88 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS chloride. On another plantation brackish water from spring-fed sea-level ponds is used for the reason that for certain fields it is the only water available. Most of the waters from wells in the central plain, the sugar- producing area, contain less than 900 parts per million of chloride. Scofield has further given the limits of a permissible sodium to calcium and magnesium ratio. In terms of equivalents per million (not parts per million) it may be expressed as follows: Na X 100 = percent sodium Ca + Mg + Na When the percentage of sodium is 60 percent or more, the water is considered unsafe for irrigation. When the ratios for typical waters obtained from wells in central St. Croix are calculated, it is evident that nearly all waters must be classed as poor or harmful if used for irrigation. The per- centage of sodium, according to the formula given above, for waters from the well at Anguilla is about 75 percent; from the spring at Envy and the Bethlehem "salt well," about 80 percent; from the Bonne Esperance well, 96 percent and 92 percent (1940 and 1919 samples, respectively) ; and from Sion Farm well, 96 percent. Thus, it seems improbable that the Tertiary limestone will anywhere yield a good irrigation water. Wells drawing from deeper strata in the alluvium yield water within the same range. Both the developed well f.nd the older drilled well at Fair Plain yield water containing r.bout 81 per- cent sodium, and water from the New Works well has about 91 percent sodium. Water from the drilled well at Anna's Hope contains about 90 percent sodium. Shallow wells and springs in the central part of the island yield water containing smaller proportions of sodiim. The water from the Golden Grove dug well yields water containing more than 70 percent sodium in the Scofield formula, but water from the sump north of Castle Burke contains only about 38 percent sodium, an amount within the permissible limits by a wide margin. Water from the spring at Plessen is probably only slightly less favorable, as is water from the shallow well at Fredensborg. Water from the Kobel well north of Jealousy in the dioritic- rock area has a sodium content of about 38 percent in the Scofield formula and may be classed as a good water for irrigation. In summary then, it would appear that most ground waters of St. Croix are very poorly adapted to use as irrigation waters. Most of them contain excessive sodium in relation to total calcium and magnesium. Some waters also contain excessive quantities of SUITABILIY OF WATER 89 chlorides and sulfates, and a few may be classed as poor on the basis of total dissolved solids alone. Although this paper is primarily concerned with the ground- water supply, it may be pointed out that samples of water taken from streams at Two Friends, Jolly Hill, and Salt River at Con- cordia show that the sodium content according to the Scofield formula is about 33 percent, 23 percent, and 32 percent, respec- tively. These waters may, therefore, be classed as good irrigation waters if the proportion of sodium to calcium and magnesium does not greatly increase in some seasons. It seems logical that if irri- gation is ever to be carried out on a large scale in St. Croix, an effort should be made to impound surface water for the purpose rather than to develop ground water. However, it is reported that in times of severe drought very little surface water is avail- able, and at such times only ground water could be used. As practically all ground waters in St. Croix contain a mod- erately large amount of dissolved salts, which are harmful where concentrated in the soil, irrigation should be carried out by meth- ods insuring adequate natural or artificial drainage of the irri- gated tracts. Thorp74 has emphasized this point in discussing proper methods of irrigating certain acreage in St. Croix: In case a water supply of considerable volume is found, it will be necessary to take precautions to use the water carefully and to dispose of the waste water in such a manner as to prevent the excessive accumulation = of alkali and soluble salts. Alkali salts in small quantities were noted in many places. Most of these areas, except those close to the sea or to lagoons, contain only enough alkali at present to slightly affect the soil. Irrigation would tend to concentrate the salts in low places where subdrainage is poor and e.long areas where there is a change in slope from comparatively steep to mor? gentle. Thorp has further pointed out that soils affected by black alkali become puddled and impervious to water. Their naturally good granular structure breaks down, the soil runs together and it is impossible to make them productive without expensive chemical treatment combined with copious irrigation and carefully planned drainage. SUGAR AND BUM PRODUCTION It is probable that no ground water obtained on St. Croix should be used for boiler feed without treatment. Few of the waters can be made suitable for use merely by reduction of the hardness. Most of the waters contain undesirable amounts of sodium as bicarbonate and chloride, which would probably necessitate ex- pensive treatment to reduce corrosiveness, foaming, or other deleterious action. 74 Thorp, James, op. cit., p. 19. 90 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS Moderately mineralized ground waters are used as maceration waters. In the production of sugar the very small amount of salts which are precipitated with the sugar are not objectionable. In the production of rum practically all the salts are left after fermentation, and waters of high mineral content can be used. The water from well 8la containing about 1,900 parts per million of chloride and having a total dissolved solid content of 4,700 parts per million has been used successfully at Central Factory. SUMMARY OF GROUND-WATER RESOURCES WATER-BEARING FORMATIONS The existing wells on the island of St. Croix obtain water from the Mount Eagle volcanics, the diorite, the Kingshill marl, and the alluvium. The alluvium is the best water-bearing formation, and properly constructed drilled wells located in the larger alluvial valleys can supply moderate amounts of water. A modern gravel-packed well constructed during this investigation in such an alluvial valley produced from 50 to 80 gallons a minute in a 68-hour pumping test. Smaller amounts of water are also obtained by other wells developed in this formation. Wells in the alluvial valleys appear to be more dependable in times of drought than wells that obtain water from rocks. A smaller number of wells obtain water from the Kingshill marl. These wells are more likely to fail in times of drought and generally yield only a few gallons a minute under conditions of continuous pumping. Deep test drilling indicated thr.t Tertiary Kingshill strata are probably not underlain by water-bearing for- mations at depth. A few wells obtain water from the Mount Eagle rocks, and some of the areas underlain by these rocks are worthy of further exploration. There are two drilled wells in dioritic ro°.k, both of which furnish ample water to wind-powered pumps in times of drought. One of the wells yielded 20 gallons a minute with less than 15 feet of drawdown in a 71^-hour pumping test. Springs yield water in a few places. Small amounts of water are obtained from springs that issue from the alluvium not far from the base of the mountains in the central part of the island, and a somewhat stronger spring issues from the Kingshill marl near the seacoast. ADEQUACY OF PRESENT SUPPLY The existing wells, springs, streams, and cisterns on St. Croix are sufficient to supply most of the island's needs in periods of SUMMARY OF GROUND-WATER RESOURCES 91 normal rainfall. During periods of drought, however, streams go dry, cisterns become depleted, many wells that are poorly located or improperly constructed fail, and the lack of water may become acute. QUALITY OF WATER The quality of water obtained from wells varies within wide limits, but in general it may be said that the ground waters are hard and contain a moderate to high content of dissolved salts, which are present largely as sodium and calcium bicarbonates and chlorides and lesser amounts of sulfates. The water from Mount Eagle rocks may have only a moderate mineral content and, as far as is known, waters from the diorite contain a very mod- erate amount of dissolved solids. The Kingshill marl generally yields water containing a moderately high to high mineral con- tent, and the mineral content of waters from the alluvium varies from moderate to high. The mineral content of well water is thought to be derived from rock-forming minerals, from alkali in the soil, and from salt spray. In a few places contamination by sea water has apparently taken place. The water from only a few wells may be suitable for irrigation, but most well water is suitable for domestic ures and for cattle. Most ground waters would need treatment before they could be used for boiler feed, but existing supplies can and are used as maceration waters and in further diluting molasses prior to fermentation in rum manufacture. RECORDS OF WELLS AND SPRINGS Wells and springs in St. Croix which are listed in table 11 were visited by the writer in 1938-39 with the exception of wells 47 and 64. These are listed by Vaughan75 in an unpublished manu- script and are included here because chemical analyses wer^> made of their waters. RECOMMENDATIONS Although water is present in all the rocks of the island, the deposits that will yield water to wells are not as widely distrib- uted. It has been found from a study of old wells drilled by the islanders and from more recent drilling by Government agencies that moderately large quantities of water may be obtained from sandy beds filling old alluvial valleys. Several of these valleys may yield several hundreds of gallons of water a minute, in the aggre- 78 Vaughan, T. W., Ground waters of St. Croix, V. I.: Unpublished manuscript, U. S. Geol. Survey, 1919. 92 GROUND WATER OP ST. CROIX, VIRGIN ISLANDS gate, to properly constructed and located wells. Probably enough water to care for a large part of the needs of the island could be developed in these valleys. Discussion of whether the costs of construction of efficient modern wells, installation of pumps, pipe line to points of use, and maintainence can be met or justified by either private citizens or governmental agencies does not fall within the scope or purpose of this report. It is merely pointed out that in certain areas water in reasonable quantity can be ob- tained from wells that are properly located with respect to geo- logic conditions and constructed by an experienced driller using modern equipment. The locations favorable for the development of a few hundreds of gallons of water a minute are in the lower courses of larger streams in the western and southern parts of the island, as fol- lows: (a) North of Frederiksted, (b) from Fair Plair northward and northwestward, (c) north of Hope and Blessing, (d) north of Jerusalem, and (e) northeast of Jerusalem. Other valleys, wholly or partly filled with alluvium, may be productive of lesser quantities of water. These are: (f) Valleys west of Spratt Hall, (g) Salt River Valley, (h) the upper reaches of the valleys mentioned in the preceding paragraph, (i) the short valleys near Rustoptwist and LaValle on the North Side, and (j) the short valleys in the East End area. At areas (a) to (e) the largest yields will be obtained, but less mineralized water will be obtained from areas (f) to (j) and particularly (h), the upper reaches of the major valleys. In almost all these favorable areas exploratory drilling may have to be done before the optimum sites for wells are known. Most of the exploratory holes will be less than 100 feet deep and need be only very small in diameter. They need not cost much to drill, and drilling them would be good insurance against the possibility of installing expensive equipment and obtaining less than a maximum amount of water. Further, by exploring several areas it may be possible to make savings in pipe line or to get a better quality of water than if only one or two areas wore explored. The maximum amount of water obtainable from wells drilled into the dioritic rock or into the Mount Eagle volc^.nics is not known. The diorite appears promising as a source of moderate amounts of water, but only a few shallow holes have been drilled in it and confirmatory data are lacking. Further, this rock is limited in areal extent. It is thought that streams in the areas of the Mourt Eagle vol- canics may be localized along fracture or fault zoner. and hence RECOMMENDATIONS 93 along stream valleys wells that have been drilled from 100 to 200 feet in these rocks may strike crevices which will yield more than the minimum quantities now pumped from shallow wells developed in these rocks. This is problematical, however, but might be most worth while to determine where there is a possibility of intercept- ing a fairly large underflow of only slightly mineralized w^ter, as in the valleys above Spratt Hall, Little Grange, Grove Place, and Little Fountain. The Tertiary limestones, by and large, are poor water-bearing strata. Only one well, at Paradise, is known to yield more than an extremely meager supply. In St. Croix, as in many other places, it is difficult to construct wells by hand that will yield water in times of drought. St. Croix is subject to severe and frequent droughts, and many inhabitants who depend upon shallow wells find themselves greatly incon- venienced at times. This recurrent situation cannot, except by chance, be eliminated by digging more shallow wells or even by constructing drilled wells unless they are properly located and finished. It must be emphasized that accumulated data on St. Croix ground-water conditions show that reasonable quantities of water from wells may be obtained in relatively few places; wells must be located in these relatively limited favorable areas in order to provide an ample supply of water for the island in times of severe drought. All wells proposed to furnish a maximum amount of water should be located by a trained ground-water geologist. Selection of each well site is an individual geologic problem on St. Croix; the location and interpretation of results of test holes, the location and details of construction of wells, and testing of wells, carried out with an understanding of the quantity and quality of water needed or sought, cannot be performed with maximum efficiency by an engineer or geologist lacking ground-water experience. TABLE 11. Records of wells and springs in St. Croix CO No. 1 2 3 4 5 6 7 8 9a 9b 9c 9d 9e 9f 9g 10 11 12 13 14 15 16 Estate or location M mile north of Butlers Bay--. Sprat Hall - _ .......... William-...-.........-- . Jolly HilL ................ Hill and Strand Streets, Fred- eriksted King Cross and Strand Streets, Frederiksted King and King Cross Streets, Frederiksted Queen and King Cross Streets, Frederiksted Hill and Princess Streets, Fred- eriksted New and Queen Cross Streets, Frederiksted 30 Hospital Street, Frederiksted Wheel of fortune _ ___ . ... Whim....-.. ............ Robe Hill... Topographic situation Se coast ... do - Valley.... ...... Coastal terrace ... ... do.... ..... ... Coastal terrace ... ... do... .... ..... ...do... ......... Slope toward sea ...do ........ ...do............ ... do ... .... .. do... .... .-... Broad valley... .. Rolling plain _ .. do .- ... do .... . do Coastal terrace ... Mountain valley Approxi- mate altitude above sea level (feet) 26 26 35 12 25 40 40 10 10 18 15 21 46 25(?) 30 40 55 125 95 14 240 Type of well' Dg Dg Dg Dg Dg Dr Dg Dg Dg Dg Dg Dg Dg Dg Dg Dg Dr Dg Dg Dg Dg Dr Depth of well (feet) 24 40 35 12 25 15 12 11 15 16 18 43 26 12 90 15 16 7 62 Principal water-bearing material Character of material Weathered volcanics -.do ... do ...do... ..... Weathered volcanics do. ... do... ..... ... - ... do... ... ... ... ...do.... .... ... do ...do...... ... do ... ... ...do. ... do... ... ..... ... . do Marly alluvium ..... Weathered volcanics Geologic horizon Upper Cretaceous do. . ...do ... . ... do........ . Upper Cretaceous ...do.._. ..... Mid-Tertiary do.... .... . ...do......... do......... do ... do. .... Pleistocene... ... do- .... Mid-Tertiary _ do ....... ... do .. .... do. ..... Pleistocene.. . Upper Cretaceous Water level Feet below surface 13J* 26 15 3M 23(7) 8J4 9 9 13 15 17H 41H '24 3Ji 21J-2 48 4}i 5 025H Date of measurement Dec. 21, 1938 do .. ... do do Jan. 11, 1939 Dec. 20, 1938 ...do... ...... do......... ... do ._ do ...do... ...... June 6, 1919 Jan. 12, 1939 Dec. 20, 1938 ...do......... do .. ...do......... ...do ..... Dec. 19, 1938 Method of lift2 M M W W W* G G G« B B B B B B B W W W W W W Use of water* DS DS Ss Ss Ind. Ind. PS PS PS PS PS PS DS S Obs. S Ss s s Chloride content (parts per million)4 98 280 110 112 548 218 1,800 2,220 1,690 2,740 2,470 1,270 410 206 710 168 240 470 138 Remarks According to owner, well yields 50 gallons a minute for limited periods. Goes dry in severe drought. According to owner, well yields 200 gallons a min- ute for limited periods. Water level varies from H£ to 9H feet from surface. Yields an "ample" supply. Water level 2 feet below surface June 5, 1919.7 Water level about 7 feet be- low immediately adjacent stream bed. RECOMMENDATIONS 95 3-1*!!!! S .*S»".3: H*° ^ " fe° o cSz E^"^ > -i -i it? ib 11 11 Q ^H ^^H sill" i i : ! air^'e^'^Z- s*~"' © " ' o> llll Illl" c c c -c -c -c -" »' iO-^ COC^ i i d g a 5§ ^ "§ 5 .3 cLr^.SJ C S a^ 53*^ .2 1 > T3 IJ 1 II 1" * -i c O^N^O »^ ^.oo > tj) t»D t»D tn i* (_i tj)tj) bC t»D Q QQQQQ QQ QQ c? c1^ w oa -^f co ws os KO bD fe ^ «^ m*^ «^ M co ! oa s >> a I *^ ^"a & ! f .0 rt a J "S B 'S 4 § |4 S «! i O 2So s^ § I ii u^ 0000 00-3- II Y CO h if t! S t E^ .SE^ 3|.i » i oo oo S § 06" oT 1 1 \* ^r 1 1 -i o> aj § S § ill -^ 4i C o -c c t i O OO O U)U> bD bD tn QQ Q Q Q 1OO O O W5 j|I Ala ill -§ McoO i i* 'I 3'is-sj >< o o . sss * ? i I > a a ' 1 * < dl Q .M fli J ! i * % s § s CO' gj N 8 -g -g 8 Q &H &H Q ^ r W3 O W3 O i-H *-. i-H -i 4 § 1 b fe"S a H ° &5 g -d ^ ~ &^^ O u t3 W S .2 o > > 3 1 ^ > &B | T3 T3 a? a | g § J H "i 2 04 o oo ws co o w: QQQ Q Q Q 0*0*0 o o c i li"H (M J3 Is QP 0 01 a -1 ^1 J bD a * § c e 'J i^ o-w -w | _S| § S ^i K COCO O 3 O 1-5 ,» >-S 'S *a 'S c, ° « | i 1 1 H III 1 1 3ih ^Q ja oo oa»-«Neo i^t' u3?ot*ooooosoT-tc^co -^ -^ ws?ot co o>o*-'*> ^ i 1 i IC3 n. Dg Dg Dg Dg Depth of well (feet) 245 58 225 58 11 170 90 67 42 145 25 145 150 1«6 94 99 135 Principal water-bearing material Character of material Marl and clay _ _ . - do ------- do Alluvium. .-...-. ... ...do....... - - .-do-.--... Limestone.. -------- Weathered volcanics Limestone. - ... do do. . do do ... ...do .......... do... .... ........ Geologic horizon Mid-Tertiary do do do. Mid-Tertiary .do .. . do Pleistocene . Mid-Tertiary Upper Cretaceous Mid-Tertiary do do......... do do......... do .... . ...do .... Water level Feet below surface 28 17 14M 9 110 89 '40 36}3 53 y2 '60 9314 B3} 2 W'4. 94J£ 115^ Date of measurement 1936 . Mar. 23, 1939 Mar. 2 , 1939 Dec. 8, 1938 June 4, 1919 Dec. 8, 1938 June 5, 1919 Dec. 9, 1938 Dec. 8, 1938 June 4, 1919 Dec. 9, 1938 do do ..... Dec. 8, 1938 ...do. . Method of lift' W G W W» W W W» W5 W6 w« W6 w« W W Use of water3 Obs. D Ind. S S S S DS S Obs. DS DS Obs. Ohs. DS S DS SD Chloride content (parts per million)4 340 700 384 585 235 830 418 380 904 600± 254 145 214 250 Remarks Yield small. Water highly mineralized. Reported to yield more than 20 gallons a minute. Yields 70 gallons a minute with 15% feet of draw- down after pumping 52 hours. Yields about 10 gallons a minute with 5 feet of drawdown. Water level 9 L£ feet below surface June 4, 1919.7 Supply does not fai. in driest seasons. Water reported to be highly mineralized. Water level 55 feet below surface June 4, 1919.' Chloride content increased slightly after pumping 2 hours. Yield estimated to be about 7 gallons a minute. Water level 56 feet below surface June 3, 1919. 7 Yields from 3 to 5 gallons a minute. Supplied 50 people and 50 animals with water in severe drought. Water level 74 feet below surface June 4, 1919.' Water level 49 feet below surface June 3, 1919. 7 CO W H O S O RECOMMENDATIONS 97 "SS-gS ' >> s >.<=> >>=, 2 o .l-as-ess g JioJTS^ a's"' fe S "^ *2 0 . +* S~^ ^* "~ M* § 5 "5 JO O OO 5 oa ^ tf5 co H »O CO T » CO IN O2 OQQQ X H Q OQ QmraQ C = & * *»»* Ol Oi Oi t OC CO 1-H 1-H I Ci s «" -*"S » -= g | I"! & N \« \M 1 \ ^f\ TH\ M "1 C3 IO IO t-1 CO It ^H c< c T: c ^ c c c -i 4 l.| S-1§ 4 SSS £^ III Jl.i -^ *** O OOO OO O OO OO O OQ O 1** M5 OO M IO Oi -^ i-H 00 OO CO -^ CM CO t- **< i-H 00 *-* O »-l l>- IO -IN-' '-"-' ""* 02 02 .2.0 02 02 O2"2 "2 "2 02 Q O2 QO2O2 OO Q QO2 O2O2 Qw WO2O2 QI-H Q O3 * 6 S&& OOss CO M CO 05 "C 2 Q OO U3 CC M c -§ a § .1 5 i 4 1 -§ c T: c T: ^ i OO ^ CO 1 r-H O CN - ! > 1 i« 1 |'|| | -§ .-( lO W O5 ^p i Oi t W OM fct fcl] t- Ml 1- Ml M UlMbMMMb MMMMMb bbMMb b M Q Q QQQQ Q Q Q QQQ QQ Q QQ QQ QQ QQQ QQ Q Q M5 O MOO °SS s £ & S e= 'ti gi O5 cBo! Jo -§| §=3 o>- 3 OS Ui ! O =s "o S 1 &.M fe a g 5^ ^2 II 5 S -s 1^ of '"a fe I 111 S 1"8 -S s co^ 10 co r* oo 1 CO Tf ift J GO COCO oo oo TABLE 11. Records of wells and springs in St. Croix Continued No. 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 Estate or location Water Gut and West Streets, Christiansted Water Gut and MarketStreets, Christiansted No. 1, Strand Street, Chris- tiansted King and Princess Streets, Christiansted ...do... .... .............. ... do... ..... ... .... ... Darby Hill ............... Cotton Valley- .. __ ... 14 mile north of Great Pond-.. Topographic situation ... do. ... ..... do ... Coastal terrace ... ...do. ...... ...do ........ Valley. ...... Edge of coastal terrace Valley... ....... Coastal terrace _ do. ... Broad lowland... ...do ..... . ... do... .... ..... Approxi- mate altitude above sea level (feet) 60 20(?) 10(7) 20 30 30 75 9 12 8 35 90 45 9 11 25 30 10 20 30 60 50 Type of well1 Dg Dg Dg Dg Dr Dr Dg Dg Dg Dg Dg Dg Dr Dg Dg Dg Dg Dg Dg Dg Dr Dg Depth of well (feet) 14 9 82 80 33 18 12 10 80 10 10 12 30 10 24 40 76 35 Principal water-bearing material Character of material ... do- ... ... .. ...do ... ... do .... .... - ... do dO ... ... ... ... ... do ... .... ... do - ... do ... --- ... do . ... do ... ... ... - ... dO .... Weathered granitic rock? do. ... ... ..... do... ... ... ...... ... do... .... .... Sand Weathered volcanics ... do... .... ..... ... dO ..... ... .. Geologic horizon do do do ... do ... .. do do -.dO . ... do... ... ... do .... do ... do ..... Pleistocene. . . ... do ... do-. .... do Upper Cretaceous dO- ... - ... do .... - Water level Feet below surface 4 '5 10 19* 7 3M JflH 6* ne' 2 19* 6M 17 Date of measurement June 4,1919 1927 Jan. 25, 1939 Jan. 25, 1939 do June 3~, 1919 Jan. 25, 1939 do ... Jan. 25, 1939 Method of lift2 W5 B B B W W w« W6 W« W5 W5 W W5 W5 W W W5 W5 W5 Use of water5 D D D D S S Ss ss ss DS S s Obs. 0S Ss s s s Chloride content (parts per million)4 340 375 950 508 270 275 250 225 950 1,020 320 280 336 870 500 5,150 1,740 538 320 544 Remarks Water level reported by B. Nelthropp. Well may also develop wea- thered granitic bedrock. In nearby abandoned well more nearly in axial part of valley water level is 20 feet from surface. Furnished water to 300 cattle during a period of drought. Water level 5 feet below surface June 3, 1919. Supplies 2,500 gallons a day. Yield is small. Yields about 2 gallons a minute. Dried up in drought of 1924-25. CD OO § CQ H O » O M X 110 11? 1 1 Q 114 1 1 K I1fi 117 do- do -- - do- -- .- do do ... do ..... ... do ... ... ... Vallov 80 45 40 5 50 50 190 T> Dg F)r Dg F)r Dg Dr Dr 47 45 10 83 38 82 28 rock --do__--. --------- ... do do do... .......... .- --do.,.. . _-.dO ----- do. dO Cretaceous do. ... 39^ 4?4 38 723 Jan. 24, 1939 Jan. 24, 1939 do W W5 W5 M6 G« W W W S Ssss s DS S 455 1,280 645 830 608 780 250 Well reported to have flowed when drilled in dry season in 1930. 1 Dg, dug, Dr, drilled. 2 M, pump operated manually; W, wjndmill; G, pump powered by gasoline engine; B, bucket. 8 D, domestic; S, stock; Ind., industrial; PS, public supply; Obs., obsolete; T, test well. 4 As determined by field tests. 5 Well being pumped when visited. 6 Well not being pumped when visited. 7 Data from unpublished manuscript "Ground waters of St. Croix, Virgin Islands," by T. W. Vaughan. CO CO APPENDIX 101 NATIONAL PARK SERVICE DRILLING PROJECT INTRODUCTION A well-drilling project for the Virgin Islands was undertaken in the period December 1940 to July 1941 by the Civilian Con- servation Corps in response to a request made by Governor Law- rence W. Cramer, of the Virgin Islands, to Mr. Conrad L. Wirth, CCC and ERA Coordinator, National Park Service, United States Department of Interior. This request was prompted by severe local shortages of water that occurred as the result of drought condition in 1941, during which many existing wells and Creque Dam Reservoir went dry. It was subsequently arranged that the CCC would furnish a geologist to investigate the project, recommend the most favor- able drilling sites, and send its drilling crew and equipment to the islands to develop water supplies on Federal-owned or con- trolled land for Federal or Federal-sponsored projects. Benefiting agencies were required to furnish the well casing, screens, and production equipment. The project was carried out under the supervision of Dcnald C. Hazlett, a geologist of the National Park Service, working under Mr. H. E. Rothrock, of the same organization. The writer advised both Mr. Rothrock and Mr. Hazlett in detail on ground-water conditions in St. Croix before the project was initiated, recom- mended many sites for drilling, and made available copies of his geologic map and basic well data. The writer was not associated with the project, however, while it was being carried out. Subse- quently, largely through the efforts of Mr. Brent S. Drane, of the National Resources Planning Board, the basic data gained in the 103 104 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS CCC drilling project were made available to the writer and sub- mitted with the request that they be incorporated with the results of the writer's own studies and interpreted accordingly. The following section has been prepared mostly from a series of typewritten memoranda written by Donald C. Hazlett. Much of the data is particularly valuable in bringing out new geologic information. However, the lack of background detail concerning the various wells that were drilled, lack of data on quality of water obtained, and lack of information on the performance of wells after completion are keenly felt. For some wells basic data are incomplete. This report on the CCC project is not intended to be a full account of the various drilling or water-suppl;T activities engaged in by the CCC while in the Virgin Islands, but rather a discussion of drilling on St. Croix only. SELECTION OF WELL AND TEST-HOLE SITE" Wells were drilled in the Whim homesteads area southeast of Frederiksted, in the Colquhoun homesteads area at the south foot of Blue Mountain, and in various other locations in the central part of the limestone plain, particularly at Fair Plain and in the St. John Princess homesteads area northwest of Christiansted. A few wells were also drilled in the eastern part of the island. With few exceptions the location of well sites was restricted to the limited areas in which water was to be used; much drilling was done in areas known to be unfavorable, and the low yields obtained from many wells and failure of others wer^ to be ex- pected. By and large, however, the light rotary rig used seemed to be adapted to the drilling conditions encountered, and the small- yield wells that were developed in these areas seem tc have filled a real need. Wells of somewhat larger yield, which were sunk in a few places, were very worth while and confirmed the writer's conclusions regarding the occurrence of ground water in the island. A few of the "dry" wells or test holes showing negative results require attention because it is believed that some of these were not properly located, or, although located at a logical spot from a consideration of the geology and existing records, drilling showed that that particular location was unfavorable and that exploration for probable water-bearing strata should have been carried on elsewhere, in some instances only a few hundred feet away. SUMMARY OF RESULTS Of the wells discussed in this report on the CCC drill'ng project, five wells provided 3 gallons per minute or less, two others yielded WELLS AND TEST HOLES 105 respectively 4 and 5 gallons a minute, two yielded 8 gallons a minute, two yielded 15 gallons a minute, and two yielded 40 gal- lons a minute. Six wells were "dry" and two provided very small quantities of brackish water. The two wells of highest yield were developed in the alluvium- filled valley near Fair Plain. Fifteen gallons a minute was ob- tained from wells in alluvial fill at Golden Grove and near Chris- tiansted. A well that obtains water from the Tertiary limestones at Paradise yields 15 gallons a minute. A well at Whim that yielded 8 gallons a minute may obtain water from Tertiary lime- stones, and a well of similar capacity at St. John may also obtain water from these limestones. The data on the successful wells, together with informatics on wells of low yield and the dry holes, confirm the conclusions pre- viously drawn by the writer; namely, the alluvial-filled channels yield fairly good supplies to properly constructed wells, ard in places the Tertiary limestones yield moderate amounts of water. In many places where the alluvium is reasonably thick, as north- west of Christiansted, permeable water-bearing formations such as coarse clean sands are lacking. Apparently good water-bearing strata in the alluvium are confined to restricted channels leading back to well-defined gorges in the mountains or hills. Where broad areas are overlain by a mantle of alluvium, as in the vicinity of Bethlehem and Colquhoun, the present streams do not neces- sarily mark the course of the deep alluvium-filled valleys and a number of shallow test holes may have to be drilled before the ancient valleys are located. It is also shown conclusively that solution cavities in limestone, in St. Croix as elsewhere, are very irregular in occurrence, and in some places these rocks yieM a reasonable amount of water but at other nearby localities the fame strata may yield very little or no water to wells. DESCRIPTION OF WELLS AND TEST HOLES WHIM HOMESTEADS Four wells were drilled on Federal homestead lands north of Hope, about 2y% miles southeast of Frederiksted, for domestic consumption and livestock. Their locations, C-l, C-2, C-3, and C-4, are shown on the map (pi. 1). These wells are 114, 136, 123, and 90 feet deep, respectively, and presumably they penetrated the Kingshill marl. C-l and C-2 each yielded 1 gallon a minute of brackish water, C-4 yielded 2 gallons a minute of water, and C-3, 8 gallons a minute. Water from C-3 and C-4 was probably fresh. The available data are so meager that no comments or dis- cussions of the results can be made. 106 GROUND WATER OF ST. CROIX, VIRGIN ISLANDS PARADISE Well C-5 was drilled to supply livestock at Paradise, in the narrow valley just east of well 29. The Kingshill mr.rl was struck at 34 feet, and strata of "very porous" white limestone were present from 40 to 45 and from 60 to 71 feet. Perforated 6-inch casing was set from 45 to 71 feet; upon bailing, th^ well yielded 15 gallons a minute. The length of the test or draw-down are not known. This well is presumably considerably better than well 29, at Paradise, which is dug to a depth of 37 feet and which is reported to yield about 100 gallons per hour. It is at a slightly higher eleva- tion than well C-5 and obviously does not reach the, porous lime- stone strata encountered in C-5. Further, water from well 29 contains 945 parts per million of chloride, according to the field test, whereas water from C-5 is reported to contain only 210 parts of chloride. It would appear that well 29 obtains water seeping in from the immediate sur- rounding area through material highly loaded with salts originat- ing as salt spray or otherwise, whereas C-5 reaches a stratum through which ground water has been moving relatively rapidly, perhaps fed from the alluvium to the north or east, in which the dissolved salt content is low. In connection with the relatively high yield of well C-5, atten- tion may be called to the limestone spring at Envy, about 1 mile to the southeast. This spring may issue fnom the strata encoun- tered in well C-5, or from similar strata. PROPOSED UNDERFLOW DAM AT FAIR PLAIN Tests were made at Fair Plain to determine the feasibility of creating an underground dam by grouting the water-bearing strata with cement through 2-inch wells. Test hole No. 1 (C-6), on the east side of the valley near the contact of alluvium with the Tertiary limestones, encountered clayey fill to a depth of 79 feet (see log, table 12). This area was considered undesirable for a grouting test. Test hole No. 2 (C-7), on the east bank of the present stream, struck permeable sands and gravels at 22 to 29 fe^t, at 48 to 51 feet, and at 72 to 78 feet. Two hundred gallons of cement grout was poured in the hole through 2-inch pipe, which was perforated opposite the water-bearing horizons. After 24 days a hole was drilled 2.6 feet north of the grouted hole in an effort to determine the effectiveness of the grout. Sands proved to be still open, and drilling mud was lost constantly and attempts to secure cores were unsuccessful. When cuttings were examined it was found WELLS AND TEST HOLES 107 that even in the most gravelly beds "evidence of penetration of the grout was so slight as to be negligible. It was estimated that the particles of grout found in these cuttings would constitute less than a teaspoonful. No grout was found adhering to the pebbles." It was concluded therefore that an underground dam could not be created by this method. The writer is of the opinion that an underground dam at this point, even if it were possible to build one, might be of doubtful value. It is thought that the largest amount of underflow passing Fair Plain and other restricted alluvial channels (at Hope, Bless- ing, Jerusalem, and other places) could be captured by pumping wells heavily enough to create a cone of depression in th^ channel- mouth area. Such pumping could be controlled to prevent back- flow from the sea, yet be great enough to withdraw from the ground most of the normal seaward underflow. FAIR PLAIN At the request of United States Army officials, test holes and wells were drilled to supply water to the airdrome located in the Manning Bay area. Test holes were drilled in the deer* alluvial- filled valley 14 to V& mile northwest of the Fair Plain vrell (45a). The results of drilling were about what would be expected from the record of the Fair Plain well; in two test wells (C-8 and C-9) three water-bearing sands were penetrated (see logs, table 13). The second test hole was converted into a 10-inch gravel-pack well, and when pumped with air it furnished about 45 gallons a minute. The chloride content is said to have "decreased from 450 parts per million to 245 parts per million after several months' pumping. To avoid overdraft on one well, two other wells (C-10 and C-ll) were drilled, 100 and 200 feet west of C-9, respectively. Data on these are not available except a statement that they yield about as much water as the first well. It was planned to pump the three wells by suction from a central point using only one pump. It was further anticipated that 30 gallons a minute w »O kO O O O C C^l (M CO M7 OO c 1 - OS 1 i «*" c .2 -3 o r3 S B^ T3 5§ % T: £ r 15 "rt fc e 1 1 g > SO-'S sji c a ~ *= j3 O «> _c c 1 < c .re in in us CM g C T1 - gbe T3 ||o OM c c c r c T Ct 03 00 O c c c c c a 1 bC O OO OO O O O CO OO 0^1 «0 £« s S . -c gO 0,3^ « 2 O* OJ '3 «2 °jtS ° "? -" = ^ ;|§ S C3feCG^C3 <"5 r?^ '^3 "^- o oo °o c &H fc< -i > X fc r c -i r -c c 'r c c c £ c T: c c r-j =0 *c o c T c > =1 * 1 i ^s i-s o "5 -*- O SCQ g tS'c ~o c S 00° O T t C^ CO 6 66 666 666 66 66 P 14 C-15 C-17 P 1 B P 10 p on P 91 0-?? 0-?,3 C-24 P oc P oc fl 0*7 P 9B P 90 east of Bethlehem southeast of Col- quhoun southeast of Col- quhoun south of Col- quhoun Mt. Pleasant south of St. John of St. John St. John southeast of La Grande Princess Central Factory of Christiansted homesteads do......... do......... do......... Princess homesteads do......... do....... . ...do.. ....... .-do......... .-do......... VICO do do.... ...... .-do.......... Valley.. ... ...do ....... ...do... .... do ...... ...do ....... Valley........ ...do... .... do 80 135 140 155 190 300 100 50 25 55 65 150 30 47^ 54.8 111 53 91 85 55 100 46 44 80 106 59 93 30 8 6 6 8 do Marl........... do ... ...do........... ...do........... do. ... ... .. ... do (?) ...do.... ............. ... do..... ...... ...... do. . ...do.. ............ 0 89 60 80 0 108 45 44 51.2 28 1 Dry 1 Dry 4 8 Dry 3 Dry 3 5 H 15 2 Ab Ab D Ab D D D Ab D Ab D D D T of C-13. See log. formation encoun- tered at 77 feet. 101 feet 6-inch casing and strainer replaced by 3-inch casing. Lower end perforated. pump test "slight." Casing perforated from 41 to 59 feet. 70 tn 80 fppt. SPP W. See log. 1 Ab, abandoned; D, domestic; S, stock; T, test well. * Chloride content in parts per million. INDEX Page Acknowledgments .................... 3 Agriculture in the area....... 7-9 ; pi. 2B Alluvium, distribution of........... 27-28 Analyses of water 71, 72, 74, 79, 80, 83, 84 Beaches, emerged ................... 38 Buried valleys................ 38 ; pi. 3 A Coral reef, emerged........... 38 ; pi. 4B submerged ....................36-37 Cushman, J. A., fossils identified by 22, 23 Drainage pattern ................. 35-36 Drilled wells, construction of 41-42 ; pi. 6B Dug wells, construction of............ 41 Erosion cycles, early Tertiary ...... 34-35 ; pi. 2A Flora and fauna of the area........... 9 Fossil lists. ............. 22-23, 24, 25-26 Geologic history, summary of...... 14-15 Ground water, general condition of. 45-47 in alluvium. ........... 56-64 ; pi. 3.4. in beach sand ................... 64 in diorite..................... 50-51 in Kingshill marl ............. 51-55 in Mount Eagle volcanics .... 47-50; pi. 5B previous investigations of...... 43-45 Henbest, F. G., fossils identified by. 25, 26 Industry in the area.................. 9 Jealousy formation ........ 19-20 ; 26-27 Kingshill marl, deposition of....... 26-27 description and distribution of. 20-22 geologic age of............... 22-26 Land forms ......... Location of the area. 36 3-4 Mount Eagle volcanics, description of ....... 16-17 diorite intruded into..... 17-19 ; pi. 1 Population of the area............. Previous investigators, work of.... Purpose and scope of investigation., Quality of water, contamination by sea water ..... contamination from organic pollution ......... ... 7 12-14 . 2-3 69-73 73-74 Page Quality of Water Continued effect of alkali soils........... 75-77 effect of base exchange on .... 77-78 effect of salt spray ............. 77 in alluvium .................. 82-85 in beach sand ................... 86 in dioritic rocks ................. 79 in Kingshill marl ............. 79-81 in Mount Eagle volcanics ...... 78-79 saline minerals and Tertiary limestone ........... 74-7.3 surface water compared with ground water ..... 69 Quin, J. T., fossils collected by. ....... 22 Rain water, use of................ 38-39 Rainfall, accumulated departure from normal of ........ 12 relation to sugar production. ..... 12 yearly and monthly averages. . . 9-12 Recommendations ............... 91-93 Springs, use of water from........... 40 Stream flow, use of.................. 39 Structure in the area. ....... 29-34 ; pi. 1 Suitability of water, for cattle. ....... 87 for humans .................. 86-87 for irrigation................. 87-89 for sugar and rum production. . 89-90 Terraces, coastal. ......... 37-38 ; pi. 4A Test-drilling program........ 64-68 ; pi. 1 Topography and drainage. . . . 4-7 ; pi. 2A Vaughan, T. W., fossils collected l^ 22, 25 fossils identified by. ....... 24, 25-26 Water supplies, at Bethlehem Factory ............ 40-41 at Central Factory .............. 40 Well logs................... 68, 112-114 Well records ................... 115-116 Wells, use of Avater from. . . 39-4C ; pi. 5B Wells and springs, records of...... 94-99 Wells and test holes, at Central Factory .............. 110 at Christiansted ............... Ill at Colquhoun-Mount Pleasant homesteads ....... 108-109 at Cotton Valley .............. Ill at Cottongarden ............... Ill at Fair Plain ................. 107 at Fredensborg ................ 108 at Golden Grove ........... 107-108 at Paradise ................... 106 at Rattan ..................... 109 at St. John-Princess homesteads ....... 109-110 at Whim homesteads .......... 105 Woodring, W. P., fossils identified by.. 24 117 &U.S. GOVERNMENT PRINTING OFFICE: 195O 842683 o cj 55 & s- s- o 8" Co - s. CO I