Water-supply and Irrigation Paper No. 168 Series {J| J^gri^wltTrs, 50 G B DEPARTMENT OF THE INTERIOR 1 0^3 5 UNITED STATES GEOLOGICAL SURVEY CHARLES D. WALCOTT, Dirkctor K2S6J THE UNDERFLOW IN ARKANSAS VALLEY IN WESTERN KANSAS BY CHARLES S. SLIGHTER WASHINGTON GOVERNMENT PRINTING OFFICE 1906 Pass Gi£> 10:^5 Digitized by the Internet Archive in 2011 with funding from The Library of Congress http://www.archive.org/details/underflowinarlonOOslic If Water-Supply aud Irrigation Paper No. 153 „ . IK, Pumpinf!; Water. 12 \0, Underground Waters, 50 DKI'AHTMI'A'r (>K TIIK IXTKRIOR UN]Tl<:i) STA'IKS (iKOLOGJCAL SURVEY CUAKl.i:S 1>. WAl.CO'l'l', DIRECTOK ¥C^Y THE UNDERFLOW IN ARKANSAS VALLEY IN WESTERN KANSAS CHARLES S. SLIGHTER WASH1NGT?)N GOVERNMENT PRINTING OFFICE 19 6 CONTEXTS Pago. Tntro to 12 miles from the river channel, and of observations ))y the electrical method of the rate of movement of the underflow. The ground-watei- levels were obtained b}' observing the water levels in private wells in the neighborhood of the river and in a few wells which were sunk especially for this purpose. The slope of the water plane was found to l)e between 7 and 8 feet to a mile in a general easterly direction, and from 2 to 3 feet to a mile toward the river channel from the country immediately to the north and south. The southern margin of the river valley is bordered for 5 to 10 miles to the south by sand hills, which are only partiall}' covered with natural vegetation. These sand hills extend from east of Dodge, Kans., to beyond the Colorado line. The river valley proper varies in width from 1 to 5 miles. Near the river channel there is a strip known as '"first bottoms,'' which is only a few feet above the river level. The principal cultivated portion of the valle}^ lies from 3 to 8 feet higher than first bottoms, and is locallv known as ''second bottoms." North of the river valley the ground I'ises rather abruptl}" to the high plains with their well-known level topography and compact sod of native g-rasses. The slope of the water plane toward the channel of the river from the north is, as has been stated, about 2^ feet to a mile, but 10 to 14 miles to the north of the valle}' the slope of the water plane changes from southerl}" to northerly, and the land at the same time gently dips to the north toward the valley of White Woman Creek. The easterly slope of 7i to 8 feet to the mile is maintained, however, quite constantly throughout all of this region. Fig. 1 shows the results of the determination of the water plane. MEASUREMENTS 2 MILES WEST OF GARDEN, KANS. (CAMP i). The measurements showed a rate of movement nuich greater than had been anticipated. The first set of underflow stations were estab- lished at a point about 2 miles west of Garden (camp 1), as shown on the map (fig. 1). The stations were in a north-south line, which was 7 UNDERFLOW IN ARKANSAS VALLEY^ WESTERN KANSAS. about li miles in length. At this south clirectioQ, and borders closelv point the river flows in an east by on the north margin of the sand T. 23 S S 22 hills, leaving but little bottom land on the south side of the river. The channel of the river where the observations were made is about 1,000 feet wide. On the north side is a strip of low land, or first bottoms, MKASrHKMKXTS Ol' Till'. T \ I )Kl{Frj)W 9 altout l.ioi) t'cct wide, wliicli is only a I'm iiiclics ahoNc tlir o'ciici'al hottoiii of tli(> ri\i'r Itcd. This low hoKoiu has sevci'al sl()nchte^f^"-p '^S P/jnt Kiii. ■-?.— Map .ihowiiig- location (if unilerflow stiitioiis and lust wells at camp 1, 2 miles west of Garden. Kans. The velocity and direction of flow is shown by the length and direction of the arrows at the various stations. The depth is indicated in figures at each location. to rise graduall}^ for severalmiles farther north, this slope constituting- the cultivated portion of the A^alley — the so-called second bottoms. The measurements at this point were made at stations that lay. in general, in a straight line across the valle}^ (hg. 2). Most of the meas- urements were made in the river channel itself, or on the low ground 10 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. to the north. One test was made on the south side of the river at the foot of the sand hills and another 1 mile to the north. The velocities were determined at depths ranging from 11 to 65 feet. The results of the measurements at this location are given in Table 1. Table 1. — Underflow measurements at carnjj I, 2 miles v;est of Garden, Kans. Date of test. No. of station. Depth of well. Velocity of ground water. Direction of flow, east of north. Location and remarks. 1904. 9 1 3 2 5 5 4 8 10 12 6 40 Feet. 16 14 31 15 31 29 17 28 17 11 65 25 Ft. per day. 5.3 4.8 10.3 9.6 8.0 8.0 9.0 9.6 8.2 4.0 1.75 1.3 90 101 71 103 65 77 55 121 121 120 101 104 June 22 1,100 feet north of river. Do Do June 21 June 24 Do. Do. June 25 July 6 July! Do July 9 September 6 September 8 1,100 feet north of river. NW. corner SW. i sec. 2, T. 23 S., R. 33 W., 8i miles north of river. Average 6.6 94 Mean direction of river channel, 100° east of north. Of the stations for which data are given in this table, No. 9 was located on the second bottoms 1 mile north of the river, No. -10 was located on the uplands 8^ miles north of the river, and No. 12 was in the sand hills south of the river. The other stations were either in the first bottoms or in the channel. Station No, 6 reached so-called "second water," or the water beneath a layer of silt which- seemed quite impervious to the flow of water. The mean of all of the observed velocities was 6.6 feet a day. The average direction of the motion was 94° east of. north, which maj^ be compared to the average direction of the river valle}" at this point, which we have estimated to be approx- imately 100'^ east of north. On the cross section through the river channel and the first bottoms (fig. 3) are shown the depth of a number of the test wells near the river channel and the velocity of the under- flow. Except for occasional la^^ers of silt, the gravels were veiy uniform in size and character of grain ; a large percentage of an}^ one sample consisted of grains larger than grains of wheat. The gravel was also found to be very uniform in lateral extent, but showed a tendency to become coarser with the depth until 32 feet was reached. At about 32 feet fine sand and silt was encountered, which seemed, as nearly as could be determined from the wells sunk in a comparatively small radius, to be horizontal in extent. Fine material was encountered at a higher level at only one place, which was near the center of the river at a depth of about 18 feet, but 50 feet upstream it was entirely absent. MEASUREMENTS OF THE UNDERFLOW 11 A well was put clow ii at station No. 1 1 in oihUt to secure a sani[)le of this line material. It was found at the same h'vel as at stations No. (J and No. S, and consisted of al'out the same kind of material, (except that it coiitaiiHHl a consich'rahh' amount of *,'-ypsum nii.\e(l witli sand. This tint" sand must l)e more of less impervious, for no watei' could he drawn hy means of a hand pump from a well driviMi in the sand, and a hole waslu>d out S feet below the casino- ri'inained for a considei-able time untilled with sand. £1 zs3e Tesf tfe// A/a / <^ WA ^ Arkdnsms ffiver ^^ n. es36.9 <^>^ \M<^ (sT) \JM1 <%> F/ne sartt/ ofi^ s//f <^ Ve/ocify of uncferf/ow in feet per efay rpl To fa/ so/ic/s in parfs per loo. ooo (fij) Chlorine m parfs per loo. ooo Horizontal scale o ^ ^ 200 ^ 400 600 ^ 800 lop o feet Vertical scale 2 4 6 8 10 feet Fig. 3.— Cross section near camp 1, 2 miles west of Garden, Kans. Tlie total solids dissolved in the ground water at various depths are shown, in parts per 100,000, by the numbers inclosed in rectan- gles. The numbers inclosed in circles express the amount, in parts per 100,000, of chlorine found at the position at which the circles are placed. The velocities above this hiyer of silt are ver}- uniform, ranging from 4.8 feet a day to 10.3 feet a day, with an average for ten tests of T.68 feet a day, with the direction varying from 55'- east of north to 121° east of north. The direction of motion at these various stations, as has been stated, was in general toward the east, but several exceptions Avere noted from time to time. At tiie time iield work was begun the channel of the Arkansas River was dry, as is very usual in the months from ,Iiute to October. The summer of 190-1, however, proved to be an exceptional one, and high Hoods were of constant occurrence throughout the sea- son. One of these floods came dow^n the river soon after the first underflow stations were established near the l)ank of the river. This oti'ered an excellent opportunity of determining the influence of the 12 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. river waters upon the underflow.. At one underflow station, situated near tiie north bank of the channel of the river, 2 miles west of Gar- den, the direction of the flow of the ground waters was very greatl}^ changed by the flood in the river. It was therefore possible to meas- ure the rate at which the river contributed to the ground waters at this point. It was found that the water during the early stages of the flood flowed away from the river at the rate of 6 to 8 feet per twent}"- four hours. This point can be established by consulting the record for stations No. 2 and No. 5, as given in Table 1. These stations are located at the same point. The velocity at station No. 2 on June 21, 1904, before a rain on the night of June 21, and before a flood which came down the river at 3 p. m. June 22, was 9.6 fe^et per twenty-four hours in a direction 103° east of north, which is substantially the direc- tion of the river channel. After the flood the velocity at the same place (at a greater depth, however) was found to be 8 feet per twenty-fouj" hours, in a direction 65° east of north, or at an angle of 35° awaj^ from the river channel, the flood having therefore changed the former direction of flow b}^ about 38°. On June 26, when the flood had still further receded, a second determination of velocit}^ showed the same rate as before, but the direction had shifted to 77° east of north, or at an angle of about 23° with the river channel. It- was not only possible to actually determine this rate of loss of water from the river by the use of the electric underflow meter, but the northerly progress of the water from the river into the gravels could be noted b}" observation of the changes in the temperature of the ground water as it flowed north. The river water was much warmer than the natural ground water, and the increased temperature could be followed away from the river bank. These facts are shown b}^ the temperatures of the water recorded in Table 11. In that table will be found the following entries: Temperature of ivater of river and test wells, June ZO, 1904- °F. River 71 Test well No. 3, 360 feet north of river -. 62. 5 Test well No. 1, 1,100 feet north of river 59 The water taken from the other wells had a somewhat more uniform temperature, excepting in two cases — that taken from the wells at sta- tion No. 10 and station No. 8. At station No. 10, at a depth of 18 feet, the temperature was 51°; at station No. 8, 28 feet below the bottom of the river, the temperature was 18°, which was the coldest water found at any point. At these two stations the direction of the underflow was the most southerly of any found, being in each case 121° east of north. It was also possible to partially trace inward moving ground water originating in the river b}^ the change in the chemical composition of the water. Apparatus was at hand for determining the alkalinity, MK.\srnKi\iKN'i"s oi' Till', rM)i;i!i'"i.()\v. 13 haiTlncss, cliloriiK'. mid the total solids dissohcd in the water; and this apparatus was nsod to sccun^ the results just stated. A further veri- ticatioii of" the inwai'dly nioxinii; r. The water plane sloped awav from the ri\ cr aitout 8 feet to the mile durinjj;- the first sta»;es of high water, and coriesponded (juite accui-atidy a\ ith the ob.scrved velocities of the water. Fio-. 8 shows the slo])e of the wat(M- ])lane on June '23 and -luly ;>. Several g-radients correspondino- to other dates are ui\en in Tat)le 1. MEASUREMENTS AT SHERLOCK, KANS. (CAMP 2). SeA'eral underflow' measurements were taken at camp 2, which was situated at Sherlock, Kans., 7 miles west of Garden. The results diliered little from those found at the first set of stations at camp 1, except that more sorting of the g-ravels had taken place at the latter point, giving greater variety to the rate of movement. The location of the various test wells and underflow stations is marked in tig. -t. The same stations are shown in cross section in fig. 5. The details of the results are printed in Table 2. From this table it will be observed that the average velocit}" of the underflow for all of the stations was S.l:) feet per twenty-four hours. The mean direction of the motion was 93.5-' east of north, which may be compared w^ith the mean direction of the river valley at this point, w^hich w^as computed to be 105^ east of north. There was some water in the river throughout all of the time during which the tests were made, and on Jul}" 27 a heavy Hood swept down the river. Table 2. — Vnderflow measurements at camp 2, SJierlock, Kans. Date of test. No. of station. Depth Velocity of of ground wells. water. Direction of flow, ea.st l^ocation and remarks, of north. | 1904. JulvKi . 13 21 22 14 1 Fed. i Ft. per dai/. 18 5.7 28 ' 22. 9 28 2.8 22 9.1 21 ' 16.0 36 1 3.0 22 16.7 26 2.2 04.0 04.0 101.0 75. 101.0 July 30 July 31 Do. 1,700 feet north of river July 17... . July 23 IS 17 15 20 July 22 July 18 182.0 T)(). July 29 1>'>. Ju]v22 16 1 i LS 1 "0 79 1 Average - .... -. 8. 9 93. 5 Ilean direction of river cliannel, 105° east of north. By studying the results of the measurements it will be observed that station No. 22 was on the border of the second bottoms, 1,700 feet north of the north bank of the river. The velocity at this station was 2.8 feet per day, and the direction of flow was substantially 14 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. the same as the direction of the river valle3^ This result is impor- tant, as the measurement was made on Jul}'^ 31, at a time when the see. /3, Te4S.,ff.34 w. Testnel/ No 6 0^'^' Scale 500 1000 feet St3 16 Fig. 4.— Map showing location of underflow stations and test wells at Sherlock, Kans., 7 miles west of Garden. The velocity and direction of flow of the ground water are shown by the length and direc- tion of the arrows at the various stations. The depth is indicated in figures at ea'ch station. flood of July 2Y should have shown some influence upon the direction of flow, if it had any at all. The direction of motion at this station was in marked contrast to the direction of flow observed at stations MEASUREMKNTS OF Till'. I ' N I H'.Kl'i ,()W 15 No. l;')aiul No. i!l, lofiitcd in llic lirst holloins. T-lc of 41' northeast of the o-eneral direction of the river \aH(>v. These stations were within the innnediat(> inthience of the Ihictnations of the hejo-ht of the Avater in the river." Of the stations established in the channel of the river itself, it is interestino- to note tliat a station located north of the center of the <^ Velocity ofunderr/ow in feet per day EH ratal solids j parts per 10 a, OOO @ Chlorine i parts per 100,000 The variatior) oF total solids is, st?own by the contour lines Horizontal scaJe ^ 500 1000 1500 2000 feet Vertical scale 15 feet Fig. .5. — Cross section at camp 2, near Sherlock, Kans. The total solids dissolved in the ground water at various depths are shown in parts per 100,000 by the numbers inclosed in rectangles. The numbers inclosed in circles express the amount, in parts per 100,000, of chlorine found at the position where the circles are placed. The contour lines show the position of water of the same strength. The contribution of soft water from the sand hills is very apparent. channel (station IS) showed a component of velocit}^ northerl}^ to the general trend of the valley, while a station south of the channel (sta- tion 15) showed a component of velocity southerly to the direction of the vallev. At station No. 17, in the channel at the same point as sta- tion No. 15, but at a greater depth, the direction of the liow corre- sponded closely with the direction of the valle}^, indicating- that the influence of flowing water in the river did not extend so deep. Station No. 20 was located on the first bottoms, 200 feet south of the south hank of the river. The motion at this point showed a southerly com- ponent, the direction of flow making an angle of IT with the direction of the vallev. The measurement was taken while the river was in a This fac /will be further illustrated at a later place in thfs report. 16 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. flood. Station No. 16 was located in the border of the sand hills, nearl}" a half mile south of the river. The direction of flow was toward the river and away from the sand hills, as should be expected on account of the excellent collecting area offered by the sand hills to the rainfall. The fact that the influence of the rivev only extends to very shallow depths and that a considerable portion of the ground water origi- nates in the sand hills is shown by the cross section (fig. 5). The con- tour lines in this figure correspond to equal amounts of total solid's dissolved in the ground water. The soft water from the sand hills can be observed to be crowding the strong water of the underflow to the north of the valle}^ MEASUREMENTS AT DEERFIELD, KANS. (CAMP 3). Camp 3 was established near the Deerfield bridge, 14 miles west of Garden. The valley at this point lies mostly south of the channel. All of the south-side lands, to the edge of the sand hills, would proba- bly be classed as "'first bottoms." The surface of the ground on these lands is only a few feet above the river bed and the soil is unusually sand}^ The topography of the sand hills south of the bottom lands is unusually well adapted for collecting the rainfall, there being several level stretches inclosed or hemmed in b}^ the hills. A short distance south of station No. 23 there are found the remains of a former river bank, indicating that an ancient channel extended as far south as sta- tion No. 23 (see fig. 6). On the north side of the channel the river sweeps a high bank from 6 to 10 feet above the river bed for a distance of about 3 miles. The uplands begin not more than 1 mile north of the river. Since the channel here borders the extreme north margin of the valley the underflow measurements^ were made south of the river or in the channel. The results are printed in Table 3. Table 3. — Underflow measurements at camp 3, Deerfield, Kans. Date of test. No. of station. Depth of wells. Velocity of ground water. Direction- of flow east of north. Location and remarks. 1904. 25 24 23 26 27 28 29 - 32 Feet. 16 21 24 36 24 21 17 31 Ft. per day. 6.3 12.5 19.2 9.2 14.8 1.25 1.6 2.2 o 66.0 67.0 111.0 111.0 129. 74.0 56.0 63.0 In channel at center. Do In channel 400 feet south of center. 600 feet south of river. August 8 ... Do. August 9 1,050 feet south of river August 12. . . 1,800 feet south of river. September 22 August 17 1.8 miles south of river. 1,800 feet south of river. Average 8.4 84.6 Mean direction of river channel, 70° east of north. MEASTHKIMKNTS OF TllK T NDKHKLOW. 17 Tho iiverage volocity of tlio ground water, S.4 feet per twenty-four liours, compares accurately with the average velocities found for stations siuiilarly located at previous camps. The mean direction does not correspond as accuratidy witli tiic giMicral trend of the river W^^^^^'^ ■■■■■■■■■- •>%3%^, Fig. 6.— Map showing the location of underflow stations at camp 3, near Deertleld, Kaiis. The velocity and direction of flow of the ground water is shown by the length and direction of the arrows. The depth is indicated in figures at each station. channel as at other stations, probably in part owing to the fact that the river has at this point a very northerly course. It will be observed that the direction of flow at stations Nos. 23, 26, and 27, which are, respectiveh", 500, 500, and 1,050 feet south of the river, had a strong southerly component, the resultant direction of IRR 153—06 2 18 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. motion making angles in the three cases of 41'-', 4-V^, and 59°, respec- tively, away from the river. These are to be contrasted with the direction of motion nearer the sand hills, at stations Nos. 29 and 32, where the direction of flow was away from the sand hills and toward the river, the direction of flow in the two cases making angles of 11° and 7°, respectively, toward the channel of the river. MEASUREMENTS AT CLEAR LAKE, NEAR HARTLAND, KANS. (CAMP 4). About 2i miles southeast of Hartland, Kans., in section 13, T. 25 S., R. 37 W. , there is situated a small body of water called Clear Lake. This pond is nearly circular, 320 feet in length and 280 feet across at the narrowest point. The pond is located within 500 feet of the south- side ditch, and the owners of the canal have had under serious consid- eration the erection of a pumping plant to take water from the pond to suppl}^ the ditch with water for irrigation. It was expected by the promoters of this scheme that the lake would act as an enormous well and would furnish a large amount of water when its level was lowered by means of large centrifugal pumps. There have been the usual rumors current among the settlers to the efl'ect that the pond was very deep, and that its elevation was independ- ent of the amount of rainfall or the fluctuations in the river, which at this point is about 1 mile northwest of the pond. Investigations showed that the water in the lake was 11 feet below the water in south- side ditch. The location of the lake with reference to the ditch and the topography near it is shown on the map, fig. 7. This is a 5-foot contour map of the district surrounding the lake, made from the level of the water in the pond as datum. Mr. H. E. Hedge, engineer of the south-side ditch, furnished the field partj^ much assistance, and especially aided them in the construction of a raft from which to take soundings, so as to make a h^^drographic map of the bottom of the lake. The shores slope at an angle of about 35° to a depth of 16 feet, where there is practically a flat level floor of mud. At this depth the diameter of the lake is about 100 feet. From this it can be com- puted that the total volume of the lake is 483,000 cubic feet, or that the lake contains about 11 acre-feet of water. The bottom of the lake consists of an accumulation of black muck, which is ver}'- soft. A test well was sunk in the center of the lake from the raft for the purpose of determining the character of the material at the bottom, so as to settle, as far as practicable, the question of whether the lake could be used as a large well from which to secure a suppl}^ of water. In sink- ing a 2-inch pipe for this purpose it was found that it would sink of its own weight to a depth of 30 feet. The pipe was then forced down without driving to a depth of 40 feet, after which it was easily jetted and driven to a depth of 62 feet below the water, or 46 feet under the bottom of the lake. In clearing the material from the 2-inch MEASUREMENTS OF THE UNDERFLOW. 19 pipe Ta foot of wasli pipo was usod, so that samples wcro washed up from a depth of about i'2 feet below the bottom of the 2-inch well. The material washed out consisted of black nuid and v\n\. with some ([uicksand. Contour interval 5 feet Datum is surface of lake water @ Elevation of nater plane compared with surface oflaife Fig. 7.— Map showing location of underflow stations and test wells near Clear Lake, Kansas. A line of levels was run from Clear Lake to Arkansas River as nearly as practicable at right angles to the direction of the river channel. The result of this leveling showed that the river was at least 8 feet higher than the lake.^' a Field notes show that the river was quite high at the time of the observation on August 20, 1904. 20 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. This result was somewhat surprising, so that a second line of levels was run to the river along the east line of section 13 until this line intersected the river. This line of levels intersected the river at a point three-fourths of a mile below the former point. The river at this point was found to be 3 feet higher than the surface of Clear Lake. Since the river slopes about 7h; feet to the mile, this checks the former measurement that the river opposite the pond is 8 feet higher than the water in the latter. The above observations seem to indicate that the small pond known as Clear Lake is one of the many circular depressions which are found throughout the western plains, and which have been fully described by Mr. Willard D. Johnson. « This small pond is of especial interest because it is in line with the dry channel of a plains stream called Bear Creek. This stream rises in Colorado, and near the western border of Kansas has a well-marked vallej^, eroded to a depth of nearly 100 feet, but as it approaches Arkansas River, near the north edge of Grant County, it loses this, and its waters spread out on the plains and sink. The ordinary flow of this stream is very small, but during times of heavy rain in eastern Colorado and western Kansas it may carry a large quantity of water, which it pours out upon the high plains of northern Grant County and into the sand hills along the south side of Arkansas River, On some occasions the freshets in this stream have been so severe that the waters have nearly reached the Arkansas. There is a slight elongated depression extending through the sand hills in line with Clear Lake, which makes it possible to believe that the waters of Bear Creek have on some occasions in the past extended to the Arkansas, but so far as known there is no settler who can testify to having actually observed such an event. It can easily be believed, from the rather remarkable character of Bear Creek, that settlers would naturally associate Clear Lake with the disappearing waters of Bear Creek, so that the story would become current that Clear Lake was merely an evidence or indication of the existence of an underground stream extending from the sand hills to Arkansas Valley itself. On this account belief in the adaptability of the lake for a supply of a large quantity of water for irrigation has been prevalent, so that an investigation of the conditions surrounding the lake has importance. There are several streams of the same type as Bear Creek in western Kansas. Underflow stations Nos. 33, 35, and 36 were established, as shown on the map (fig. 7), for the purpose of determining the direction and mag- nitude of the velocity of the underground water. It was hoped to determine in this way whether or not there was any seepage at this point from the direction of Bear Creek toward Arkansas Valley. The a The High Plains and their utilization: Twenty-first Ann. Rept. U. S. Geol. Survey, pt. 4, 1900, pp. 609, 693-715. MEASUREMENTS OF THE UNDERFLOW. 21 (llroctioii uiul velocity of inox onuMit iir(> iiuliculiMl l»y tlic iirrows shown in ti«>'. 7, and the details of Ihc nioasiirenuMit arc oi\ en in 'lable 4. Station No. 33, 25 feet south of Clear Lake, gave a veU)citv of .5 feet a day ; the direetion was ahnost exactly across the dry channel of Bear C'reiMv and in the general direction of Arkansas Valley. Station No. 3«), located at the same place, but at a depth of 3(S feet, showed a velocity of 4.3 feet in the same direction. Station No. 35, 150 feet north- west of Clear Lake, showed a velocit}^ of 5 feet a day at a d(»pth of 30 feet. The velocities observed at this point may have been due in part to seepage from the south-side ditch, as the direction was almost directly away from this ditch and in the general direction of the slope of the ground. Even if this be the case, it nevertheless proves that there is no seepage nor movement of ground water extending down the so-called channel of Bear Creek, for if there had been such motion the resultant velocit}' found would at least have shown a component of motion in the direction of the flow in the channel of Bear Creek. It would be impossible for the seepage from south side ditch to disguise completelv a ground-water movement in another direction. Table 4. — Uiulertiow vwasurements at ccimp 4, Clear Lake, near Hartland, Kans. Date of test. No. of station. Depth Velocity of of ground wells. water. Direction of flow, east of north. Location and remarks. 190-4. August 19 33 35 36 Feet. 1 Ft. per day. 30 5.0 15 3.1 38 4.3 74 101 74 25 feet southwest of Clear Lake. August 20 150 feet northwest of Clear Lake. August 21 25 feet southwest of Clear Lake. An attempt was made to sink a set of wells at station No. 34, 230 feet south of Clear Lake. At this point wells were driven to a depth of 40 feet, but the material was so fine that no water could be pumped from the wells, except a very little at a depth of 16 feet. On this account no test was made. It can easily be concluded from the tests made above that it is not feasible to use Clear Lake as a well from which a large quantity of water can be pumped for irrigation purposes. While Clear, Lake undoubtedly has direct connection with the surrounding ground water and shows the level of the ground water in its neighborhood, the evi- dence from the character of the material encountered in stations Nos. 33, 35. and 36, and the evidence from direct observation of the flow of the water and the material encountered in the deep well sunk in the middle of the lake, show that the pond is not favorably situated for use as a source of a large supph' of water for the south-side ditch. These observations also show that no ground water reaches either Clear Lake or Arkansas River from the lost waters of Bear Creek. Any seepage water approaching Arkansas Valley from Bear Creek must take up a generally easterly movement almost immediatelv upon entering the sand hills. ^ 22 UNDEEFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. MEASUREMENTS OF THE UNDERFLOW AT THE NARROWS OF ARKANSAS RIVER, NEAR HARTLAND, KANS. (CAMP 5). Two miles west of Hartland, Kans. , Arkansas Eiver flows between rock bluffs, the distance between which at the narrowest portion is 2,250 feet. The river channel occupies 900 feet of this distance, only a portion of which was utilized by flowing water on August 24, 1904. MEASUREMENTS OF THE ITNOKHKLOW 28 Test wolls A, 1>, aiul C w^^vo driven to shallow depths for pose of (l(>t(Mniiirm,u- the slope of the water plane throuoh the the pur- Narrows. Q IISM iSSJ_ g£puei£seis £0N II3M js^i ^V y'S QNV 1 K HDlta NOZVI^V I/3M Sjs^uiiujag I S 5 si u = .a In addition to these test wells, the elevation of the water was taken at Demling-er\s well and in the w^ells of station No. 38, and in test wells 24 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. driven for the purpose of testing , for rock. These wells form a line about a mile long-, as indicated on the map (fig. 8). The gradient of the water plane in the first portion of this line was 7.5 feet per mifte; in the next portion it was 6.4 feet per mile, and in the next 9.2 feet per mile. Just above the Narrows the gradient was found to be 11.4 feet per mile, and in the last portion, in the Narrows itself, the slope of the water plane was 8.5 feet per mile. A profile showing these gradients is given at the bottom of fig. 9. Test wells Nos. 1 and 2 (shown in fig. 8) were driven for the pur- pose of testing for bed rock. What is believed to be rock was struck at test well No. 1, at elevation 3,011.7, or 37 feet below tlie water plane, and at test well No. 2 rock was reached at elevation 3,009.8, or 39.3 feet below the water plane. Rock was also struck at station No. 38 at 38.75 feet below the water plane. As a diamond drill was not at hand, the evidence that bed rock was reached is, of course, not conclusive. The only test that could be applied was the evidence supplied by the drill on the wash pipe and by the way in which the 2-inch casing acted when an attempt was made to drive it. Two measurements were made of the rate of movement of the underflow near the center of the Narrows at stations Nos. 37 and 38. The velocities determined were 9.6 feet per twenty -four hours at a depth of 16 feet and 3.4 feet per twenty-four hours at a depth of 25 feet. Table 5. — Underfloiv measurements at camp 5, Narrows of Arkansas River, near Hart- land, Kans. Date of test. No. of station. Depth of wells. Velocity of ground water. Direction of flow, cast of north. Location and remarks. 1904. August 23 37 38 Feet. 16 25 Ft. per day. 9.6 3.4 o 77 77 Center of channel. August 26 Do. From the cross section of the Narrows (shown in fig. 9) an estimate can be made of the amount of water which flows through the Narrows. The total cross section of the sands, assuming the above test borings as indicating the true position of bed rock, is 75,000 square feet. Assuming one-third as the porosity of the sands and 10 feet per da}'^ as the average velocity of the ground water, the total flow through the Narrows would be 250,000 cubic feet per day, or 2.9 cubic feet per second. The actual average velocity of the underflow is undoubtedly much less than 10 feet per day, so that the above result represents the maximum that can be claimed in a high estimate. (MI A VTVAi II. FLFC ri'ATIONS OF <; IiOUNl)-WA IKR T^FVFT.. INP^LUENCE OF RAINFALL AND OF HEIGHT OF WATER IN ARKANSAS RIVER ON THE GROUND-WATER LEVEL. Duiiiij;- the liold work of the suiuiiier .several o})portunitie,s were found to observe the iiiHuence of a change of level of the water in tli(» river upon the water plane in the adjacent bottom lands. The summer of i\H)4: was especially favorable for observations of this kind, as the season was an ej^ceptional one, both in respect to the rainfall and as to the ({uantity of water flowing in the river. There was water in Arkansas Kiver, in western Kansas, during nearly all of the time from the mid- dle of June to the middle of September, and on several occasions floods of marked suddenness and great severity passed down the river. The rainfall during the same period was above the average. The record of rainfall from May 1 to October 1, as observed by the volunteer sta- tion of the United States Weather Bureau at Garden, Kans., is given in Tal)lo C. Table 6. — Da'thj precipitaHon, Garden, Kans., May 1 to September SO, 1904. May. 0.58 Trace. 1.82 .75 .0 .0 .0 .20 .0 .0 .0 Trace. .0 .0 Trace .0 .0 .0 .0 .30 .0 .0 Trace. .0 .19 .0 .0 .0 .0 .0 .0 . 03 Trace. .85 .94 .0 .0 .0 .0 .0 .0 .0 .0 "Much less at Sherlock, Kans. July. August. September. 0.0 .25 .04 .0 .0 .0 .0 .0 .72 0.0 .0 .0 .95 Trace. .12 .55 1.10 .0 .05 .0 1.32 .OS .0 .0 .0 .0 .0 .0 Trace. Trace. nl.42 .0 .0 .06 .0 0.03 .0 .0 .28 .0 .45 .0 .0 Trace. .0 .0 .0 .0 .0 .0 .0 .0 .04 . 32 .0 .0 .0 .0 .0 .11 .0 0.0 .24 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 25 26 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. Table 6. — Daily jjrecipitation, Oarden, Kans., May 1 io September 30, 1904 — Continued. Date. May. June. 1 July. August. September. 27 0.03 .04 .0 .0 .0 1 0.0 0.0 .20 ! .0 .0 .0 .0 -0 0.0 .0 .0 .09 Trace. 10 28 1 85 29 1 10 30 10 31 Trace. Total 4.30 2.64 5.65 1.32 3 39 Total for Ave months, 17.30. Observations of the water plane were made very systematically during the various stages of the water in the river by Mr. Wolff, who was in charge of the party making the field observations. The results of these observations are given in the accompanying diagrams, which Mr. Wolff' has constructed from the field notes. The first underflow determinations were made at the camp located about 2 miles west of Garden, Kans., on the ranch of Mrs. M. Richter, which is referred to in the text as camp 1. At this camp a number of shallow test wells were put in place for the special purpose of observing the position of the water plane. These test wells are shown on the map (fig. 2), from which it will be observed that test wells Nos. 1 and 2 were located north of the river bank at a distance of about 1,070 feet; test' well No. 3 was closer to the river, at a distance of about 360 feet from the north bank. A large well located on the ranch of Mrs. Richter, and used for irrigation, was also used for the purpose of keeping track of the fluctuations of the water plane. The location of this well is shown on the map (fig. 2) near the quarter-section corner in the upper right-hand corner of the map. As will be observed, this well is situated a considerable distance upstream from test wells Nos. 1, 2, and 3; hence the water in it stood much higher than that in the test wells, since the water plane slopes eastward at the rate of about 7i feet per mile. The land in which test wells Nos. 1, 2, and 3 are situated is what is commonly called in that locality "first bottoms." Immediately north of, test wells Nos. 1 and 2 the "second bottoms" begin, the land here being some 3 to 6 feet higher than in the " first bottoms." Two sloughs shown on the map were grass covered, but contained more or less water either during high stages of the river or after heavy rains. In fig. 10 the elevations of water in Arkansas River from June 16 to July 11, 1901, and the elevations in test wells Nos. 1, 2, and 3 and in Mrs. Richter's well are represented graphic- ally. The elevations are expressed in feet above mean sea level, as determined from the United States Geological Survey permanent bench marks in the valley. The detailed observations at these stations are printed in Table 7, in which the elevations are given in feet above mean sea level. The observation of the height of the river was made from a gage rod set up in the river and observed from the bank with KLrcTr.vnoNs of ohounh-watkh [.kvkj.. 27 a level. Observations were made mornini^' and e\riiiiin- diiriiiy the period covei'ed h\' tht> tal>le. TliM-e were occasional omissions of o!)sei'vation ol" ri\er heij»'ht, due to the absence of the level from camp. 'r.\i«i,r, 7. — Ehratioii of (/round water in the Arkansas Rirer ami /»/ lei^t ireJh near cainii 1, 2 miles trest of (larden, Kans. [Wells Niis. 1 iUKl J .ue 1,070 feet north of river; well No. 3 is 360 feet nortli of river. Datiuii is 2, sou feet above mean sea level.] Date. Tinn". Eleva- tion of water ill wrllXo.l. Hydrau- lic'^radi- ent per milefroni well No. 2 to well No. 1. Eleva- tion of water in well No. 2. Hydrau- lic gradi- ent per mile from well No. 1 to well No. 3. Elevation of water in well No. 3. Eleva- tion of water in river. Barometric pressure in inches of mercury. 1904. 12m ... 6 p. m a. ui V 111 Fed. 33. 97 33.86 33.90 33. 87 Feet. Feet. Feet. Feel. Feet. 36.7 Tnclies. 28.60 Do 26.54 26. 62 Do 26. 65 Do a. lu Op. m 6 p. m 6 a. m 33.98 33.76 33.75 33.89 26 63 Do June 19 8.1 33. 53 33. 58 33. 60 6.3 5.9 5.3 34.61 34.55 34.61 34.41 34.47 34.33 34.38 35. 02 36.2 26.50 20. 45 June 20 Do 36.1 36.0 36.0 June 21 t) a . m 12 111 33. 82 33.59 33.77 34.45 4.8 5.5 4.5 4.2 26.47 Do 6.4 7.2 8.9 33.36 33. 51 34.13 Do () p. ni (') a. m 12 m June 2'' Do .. Do 6 p. m 6 a. m 6 p. m 6 a. m 6 p. m 33.94 34.05 33.87 34.00 33.77 33. 93 33.93 33.77 33.93 8.8 7.6 8.0 7.1 6.8 5.1 3.6 35.12 35.07 34. 95 34.95 34.69 34.61 34. 42 37.7 36.9 36.9 36. 5 36.3 36.2 3.5.9 June 23 Do 6.7 33.81 26. 27 June 24 Do June 25 6.9 6.7 8.1 33. 75 33. 53 33.64 26. 35 June 26 Do 12 m 6.1 7.7 33.55 33. 67 June 27 3.9 34. 45 35.9 35.9 3.5.9 35.8 35.8 35.7 35.7 35.7 3.5.7 35.7 35. 6 35.6 35.6 35.6 35.6 3.5.6 35.7 3.5.7 3.5.6 35.8 Do June 28 June 29 Do Julv 1 6 a.m Do Julv 2 Do 6 p.m Julys Do 6 a. m 6 p. m 33.19 6.7 32.95 7. 2 34.16 Julv 4 6 a. m July 5 6 a. m Julv 6 a. m July 7 6 a. m 33.99 7.2 33.73 2.72 34.35 Do 6 p.m Julys 6 a.m 34.53 33.88 7.2 8.1 34.27 33.-59 3.64 1.71 35.02 34.11 Julv 9 6 a. m Julv 10 6 a.m Julv 11 6 a. m^ 33.68 8.1 33.39 3.15 34.10 28 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. From the morning of June 21 until noon of June 22, which are left blank in the table, there was no material change in the height of the river. The water in the river slowl}'^ sank during the period covered from noon of June 16 to noon of June 22. The record shown in fig. 10 begins on June 16. The levels in the various wells remained substan- tially stationary from that date until June 22. During the night of June 21 a heavy rain fell, which is given on the official record at Grarden as 0.94 of an inch. The test wells on the morning of June 22 showed marked changes in the elevation of the ground water, due to the rain of the previous night. Well No. 1 rose 0.68 of a foot; well No. 2 rose 0.62 of a foot; well No. 3 rose 0.64 of a foot, while the Richter well rose 0.05 of a foot before noon of June 22, and by the morning of June 24 had risen 0.10 of a foot. The river remained sta- tionary until 3 p. m. of June 22, when a flood consisting of an abrupt wave swept down the river, causing a rise of 1.7 feet. Notwith- standing this rise in the river, the water in test wells Nos. 1 and 2, 1,070 feet from the river, fell during the interval between the morn- ing and evening of June 22, while test well No. 3, which was situated within 360 feet of the river bank, was only 0.1 higher at 6 p. m. of June 22 than it was at 6 a. m. on the same da3^ These results show that the heavy rain of the night of June 21 raised the water in all of the test wells, but that the flood of the afternoon of June 22 raised the water only in the well nearest the river. The river gradually receded from the high-water mark reached on the afternoon of June 22, and all of the test wells gradually fell. There was no rain until July 4, except a slight shower on June 28. Test wells Nos. 1, 2, and 8 showed a tendency to fall, although the water in the river was from 2 to 3 feet higher than the water in the wells during all of this period. The rise in the water plane from 6 p. m. of June 21 to 6 a. m. of June 22, amounting to a rise of 0.68 foot in test well No. 1 and 0.62 foot in test well No. 2, was due, as stated above, to a heavy rain which fell dur- ing the night. From the data at hand it is possible to express the magnitude of the contribution to the underflow as so manj^ cubic feet of water for each mile of the river valley. If this contribution be sup- posed to extend uniformly over a given period of time, then the addi- tion to the ground water may be expressed as a continuous flow of so many cubic feet of water per second for each linear mile of the river valley. Thus, in the present case, if we suppose that the rainfall of the night of June 21 fell uniformly during the twelve hours from 6 p. m. to 6 a. m., we can readily compute that the observed increased amount of ground water was equivalent for each mile of valley along the river to a continuous flow of water amounting to 23.8 cubic feet per second. To put this in other words, we can say that if the sands of the valley had contributed to the river by seepage all of the water which the rain added to these same sands, the seepage would amount to a continuous FLUCTUATIONS OF (JROUND-WATKR LKA'EL. 29 How into each mile of the river of 2P>.8 cubic feet i)er socond, main- tained for twelve hours. ClevHion abovr sea level 'n reel CC -1 f. £3, tc ii; o .1 1 1 5! 21 i.»l — 51 ■ ' " t: J J ^ ^ \ -4— - -^ JC i5. ■ - -1^ " t rr ^!^ X- l F" §-.^ 7 -- -? i- ^^ i\ - 1 J ^^ ^4 _f V JI ^ ^v 1^^ ^- 4"!; IT ^_ii ^ -4 4 )l i J' .' ^ t ^ _^ .^ '5f^ "\~ ^ 77 5^° "^u — 4^ % T 5 ^'' 6' ^ r -F'^'' ^i 4 ^ --^ 7") 2 ~ ^ Pdmped ' ^ '^ C ^T . l\ _,^ ^ 4 iZ- -1^ .±^^? L i^, t • V v\ - V '- \ Xr^Pu 'SP£d_ V5 " Ji: f -1 -^ ~--i ^ - Y- ^,; i- si ^ rjii t '-- t j"^' i; . t <)f\> \ ■iU) "^ ^ .- ^ t . "^^ t " ^^--. ^ --«; r- r* ^ ^r^ped _ \ ^ \ 5" <^ ( Pumped 1 '^v i Is: i' ^. ^j^ "^^ ^ ^^ ~^^? ^ V ^^>^^''" i i^z-f^ ^ 7 r^' t :: -. f ; "" ii r: ° L7 ' V / ^^ r ^ -A -jL In a similar way, if the water contributed to the g-round by the iiood in the river fron/s p. ra. to 6 p. m. of June 22 be considered as spread uniformly over twelve hours, it can readily be computed that the gain by the ground due to this cause represents a seepage loss for each mile 30 UNDEEFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. of the river of 6.4 cubic feet of water per second. It can readily be seen, therefore, that the rainfall contributed a much greater volume of water to the underflow than was contributed by the flood in the river. The average rise of the ground water during the night of June 21 was such that it would require a rainfall, without run-off, of 2.2 inches to fully account for it. The rainfall recorded at Garden for the night of June 21 was 0.94 inch. The difference between the measured rise of the ground water and the rainfall is explained by the fact that there is almost no run-off' from the level lands of the river valley, so that nearly all of the drainage is underground by means of the deposits of sands and gravels. The seepage of this drainage is in part toward the low- water plane along and near the river channel. At such a place the amount of rise in the ground water would naturally be higher than could be accounted for by the localized rainfall. After the high water of June 22 the river gradually fell until, on the morning of June 2Y, it had reached an elevation of 2,835.9 feet, which was 0.1 foot lower than its elevation on the morning of June 22. The water in the test wells gradually fell during the same period, the corresponding loss of ground water being given in Table 8 as a con- tinuous flow of water expressed in cubic feet per second for 1 mile of river valley. By the morning of June 27 nearl}^ all of the water con- tributed to the sands of the valley by the rain of June 21 and the flood of June 22 had disappeared. The gain and loss can be expressed as follows, in the form of a balance sheet: Table 8. — Loss and gain of ground water per mile of river valley, 1904- I.— FROM RIVER TO WELL NO. 1, 1,070 FEET NORTH OF RIVER, GARDEN, KANS. Time. Gain in ground water per mile of river val- ley. Remarks. Sec. feet. - 0.93 23.8 7.3 - 5.4 - 3.] - 2.7 2.3 22.4 -14.6 - 1.0 No change in elevation of river water, and only slight change in elevation of water in well No. 1 until June 22. Due to rainfall of 0.94 inch. Due to rise in river. Due to rain. No change in elevation of river water. Due to rain night of July 7. No change in elevation of river water. Rate of loss during 24 hours after pre- cipitation of lr2 inches of night of July 7. FLUCTL'ATIONS OF tUiOU NJD-WATKK LKVKL. 31 Tahi.k S. -Liisk (ukI I/dill "J (jrniiiid ir and l);ir()nu>tor there are diurnal periodic changes in the position of the water plane, and these tiuetuations are such that it is always more satisfactory to compare ol)servations taken at corresponding times of the day. unless the intermediate changes are veiy violent. The mornint'' l(>v(d of th(> li'round water is normailv higher than the Fset 6.d I 6.1 2866.0 v N N. Aneroid barometer, uncorrected Elevation of water in test well No.l N % / / ..^ X 7 / / / ^ a< \ \ \ / / / / V / \ \ ^ / / V * "^ y fe' '■V ^ ^^ % \ ^ r% / J' 26.33 26.28 26.23 26.4-7 ^ 26.42 ^ 26.62 26.57 26.52 4 6 8 10 12 2 ^ 6 Q 10 12 A.M. * P.M. Fig. 11. — Curves of barometric pressure and height of water plane, showing correspondence between the fluctuations of the barometer and the water plane as observed on several dates at Sherlock, Kans. The dotted lines give the diurnal variations in the barometric pressure; the full lines show the elevations of the water in test well No. 1. evening level, the fluctuations in the wells discussed above being indi- cated very clearly by some of the lines in fig. 10, especially those show- ing the June fluctuations in test wells Nos. 1, 2, and 3. Some results showing the correspondence between the barometric pressure and the ground-water elevation were sought for at camp 1, lER 153—06 3 34 UNDEEFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. near Sherlock, Kans. The data obtained are depicted graphically in fig. 11. The results were not what were expected, as the influence of the barometric pressure should be to raise the ground water as the barometer falls. '^ This indicates that the low position of the ground water in the afternoon of each day is probably a temperature effect, due to the decrease in the capillarity of the water with the tempera- ture. The ground water at test well No. 1, Sherlock, and in test wells Nos. 1, 2, and 3, Garden, was within 3 feet of the surface of the ground and the difference in temperature of day and night was very great. In fig. 10 the level of water in the Richter well, 2,500 feet north of the river, is compared for a period of about thirty days with the ele- vation of the water in Arkansas River. The total variation of the water plane, as shown by the levels observed in the well twice daily during the thirty-day interval, did not exceed 2 inches. This shows that the influence of the river upon the ground water dies out to prac- tically nothing in a distance of one-half mile. The influence of the rainfall upon the water in the well is traceable by a comparison of the rainfall record and the well curve, but it is uncertain whether any connection can be detected between the elevation of the river and the well curve. The influence of occasional pumping upon the ground- water level is quite pronounced. The observations given above indicate the following conclusions: 1. The level of the ground water shows a marked tendency to remain at a level lower than the channel of the river at a point about one- fourth mile north of the river channel. 2. The elevation of the water plane is very sensitive to the amount of rainfall, the rise in the water plane (due to a rain) in the first bot- toms being greater than can be accounted for by the localized pre- cipitation. 3. High water in the river has much less effect upon the level of the ground water than the rainfall, its influence being confined to a dis- tance of a few hundred feet from the river channel. 4. The water plane falls at a very rapid rate after its elevation has been increased by rainfall or b}^ a flood in the river. 5. The fact that the water plane lies for a considerable distance at a level lower than the river channel, even when there is water in the river for an extended length of time, and the rapid way in which the ground water sinks after its rise due to heavy rain, establishes the fact that the underground drainage through the sands and gravels beneath the river valley is more than sufiicient to carry ofl' all of the rainfall without run-off into the river channel. aSlichter, C. S., Motions of anderground waters: Water-Sup. and Irr. Paper No. 67, U, S. Geol. Survey, 1902, p. 73. FLUCTUATIONS OF GKOUND-WATER LEVEL. 35 FLUCTUATION OF GROUND-WATER LEVEL AT SHERLOCK, KANS. Observations of ('luini4('s of level of >;routul watcM- near Sheilock, Kans., were mado during the period extending from fluly 15 to August 8, 1904. For tiiis purpose a nunilier of test wells were driven, the location of which is shown in fig. 4. Of these test wells, No. 2 was 900 feet and No. 3 was 400 feet north of the river; No. 5 was 550 feet and No. 6 was 2,500 feet south of the river. The complete record of observations taken in the field is given in Table 9. llie principal results presented ])y this ta])le are shown graphically in tig. 12. As shown by this diagram, Arkansas River gradually fell from July 15 until July 27. At this time the water in the river had reached a very low stage, the flowing water occup\'ing a width in the channel of about a rod and a depth of about 6 inches. 36 UNDEEFLOW IN AKKANSAS VALLEY, WESTERN KANSAS. s Barom- eter pres- sure of mercury, uncor- rected. Inches. 26.34 26.32 ; S g S § §■ g' S ^ ^' S' S S" S" § S S ^ S s" g Hydrau- lic gradi- ent per milefrom well No. 5 to well No. 4. 1 ■' oi in • tP Tf ^ • 1-^ r-: i> Eleva- tion of water in well No. 4; 1,000 feet south of river. f^ • c- o . o o 1 to CO 66.97 66.95 66. 95 Hydrau- lic gradi ent per milefrom well No. 6 to well No. 5. t^ • C^ X i^ i> t^ Eleva- tion of water in well No. 6; 2,500 feet south of river. f^ C 00 ci to Oi O^ Cs Oi OS Ci Oi Eleva- tion of water in well No.l. Feet. 66.36 66.28 66.26 66.24 66.21 66.20 66.19 66.23 66. 23 66.21 66.16 66.12 66.10 66.09 66.08 66.08 66.13 66.08 66. 06 66.01 66.00 66.01 66.04 66.04 66.04 d a C32 P 03 in o CO p 03 o CO c p o r-l £ o CO P I- £ CO £ p cS m o g o CO p S CO p o CO in H P o3 CO cj 00 £ cj o T-i P CO a & in £ p in CO E ft in r-- £ p in a d in d Tj5 ; t- !0 >- S o o Q o O o Q o d fl 00 r-i > i-J o , fl O fl 6 fl o fl o fl O: s o fl d fl d fl d fl C fl o fl o fl o fl FLUCTUATIONS OF GROUND-WATER LEVEL. 37 S3 3? ^ g 1-1 • 00 I '• 00 a> >-i o .-1 • 1~ x x x • OI 1/3 1-- ; 66. 96 06. 82 CD 00 to to to to 5 i ^ g : ■4 - t~ to vC iC ■* -f -J' CO I CO • -^ to t~ ci l^tOtOCOXCOiCi/S i-I"^"^c4o»cJiC-^' oi 65.18 04. 97 64.88 64.95 64.88 64.84 64.83 64.89 65.29 65.68 , 65. 88 65. 95 05. 88 65.56 to OS '■ ■X' . CO (M lO iC Xt^^^rXCOr-*i-< - .-■ ^- to to .C ■d C<5 05 • c-i 'v !- 65.09 64.91 .- .- - 64.85 64.93 64.83 04. 79 64.75 64.73 64.87 65. 07 65. 24 65.45 65. 40 65.27 : g 9.5 10.0 o o o o 00 Oi iC O O O oi GC 00 00 oi oi ■ OJ 65. 93 65. 86 65. 87 '• a> M ■ t~ X • lO I/: • to tc If: to o to 05. 80 65.91 06. 04 66.21 06. 25 60. 12 • o ■ to • to c 3 : M ^ 3 £ i c 3 i 3 3 3 3 5 r 3 ^ 1 1/ 3 c 5 C 3 ^ i ° i 3 i i i 3 5 U 3 i 1 3 i 3 X Q c o ; 1- G t. 3 I 1- 3 u il 3 h ^ ^ c 3 ■" ■5 *- 3 u 3 5 C 5 f M H 5 c 3 5 ^ I c \ C '■\ : 5 f 5 = Ot c ^ c^ c 3 3' :■ 3 10 3 . < - 5 ' 3 5 38 UNDERFLOW IK ARKANSAS VALLEY, WESTERN KANSAS. During this same per except on July 22 and a ■iod of fall in the river there was no rainfall very light rain on July 25. The rain of July 1 ^ = - _ ^ - ..s / r <-§, / 't: / i , r ^ 1 ^ r >o'~ j n f\ r \ ^ ^ - \ ' "^ \ :: 1 : /, 4 - / 4 - i / J 4 - ^■/ -V t / / / t t - / ^/ / ^ L _ ^ / k ' . ^; ^ ■/ tl-iL- \, ' ^ i / s t^ i :u V) ^\tt #'" 2 -t <3 i% ^hX %\ ttx f ztx 7 U -t ^ / 4-^i >/ , tit \ ttt I \ / ■^ / / < < J" " I / "^f ^ "- v~ ^ ) r_ __ . ¥" V. k' ^' .s'l 1 1 1 1 kl -t; fe ^^ ^i i P 5> IS I ^/^ . ^^ ^- 4: t ^ f' ? ^ X o 1^ 7 ' 1 / - ' -^ ,1 I ^ 'O H -O ^ 3 -S"^ o g he ■ rt E '*-' S ^ I! « C PI (D oj ^^ " b S " /?a/ ^// uoi^eAgij 22 was measured at Garden by the volunteer observer of the United States Weather Bureau as 1.42 inches, but the rainfall at Sherlock was very much less. During this period of fall of level of the water FLUCTUATIONS OF GROUND-WATER LEVEL. 3V> in the river the test wells north of the river fell at corresponding- rates. The total fall in the river amounted to 0.95 of a foot; the fall in test well No. 3, 400 feet north of the river, during the same period was 0.9 of a foot; in test well No. 2, 900 feet north of the river, 0.77 of a foot; in test well No. 5, 550 feet south of the river, 0.5 of a foot; and in test well No. G, 2,500 feet south of the river, 0.3 of a foot. On July 27, between 11 a. m. and 5 p. m., the river rose 1.6 of a foot, restoring- the level of water in the river to the height of July 15 plus 0.6 of a foot. This sudden rise in the river was not accompanied by rainfall in the neigh ])orh()od of Sherlock. Its influence upon the various test wells is shown by fig. 12. The immediate efl'ect upon test wells Nos. 2 and 3, north of the river, was very apparent. Between 11 a. m. and 7 p. m. test well No. 3, 400 feet north of the river, rose 1.05 feet, and test well No. 2, 900 feet north of the river, rose 0.49 of a foot. By the next morning at 6 a. m. the river had fallen 0.25 of a foot; test well No. 3, 400 feet north of the river, had risen about 0.1 of a foot, and test well No. 2, 900 feet north of the river, had risen 0.23 of a foot. The river continued to fall very slowly, on the morn- ing of July 29 having fallen onl}' about one-half of 0.1 of a foot from its elevation on July 28; the water in test wells Nos. 2 and 3 had dropped about the same amount, and on August 1, at 8 a. m., when the river had fallen 0.6 of a foot below its elevation of July 29, test wells Nos. 3 and 2 had dropped 3.6 and 1.8 feet, respectively. During this same period of time the water plane south of the river acted very differently from that observed on the north side of the river. The water in test well No. 6, 2,500 feet south of the river, fell continuously from July 18 to August 1, notwithstanding- the flood of July 27; and that in test well No. 5, 550 feet south of the river, fell from July 18 until July 27, the total fall amounting- to 0.47 of a foot. No observation was made at this test well on July 28, but by the morning- of Jul}- 29 the water had risen 0.45 of a foot. On August 1 it had fallen 0.2 of a foot below its level on the morning of Jul}- 29, in sjniipathy with the general fall of the water in the river. It can be seen from this that the elevation of the water in the various test wells showed all varieties of change during the flood in the river. The wells within 900 feet of the river fluctuated quite accurately with the changing level in the river itself, while the water in the test well one-half mile from the river seemed to show no effect of the flood in the river during the period of observation. In explanation of the gradual fall in the test wells from July 18 to July 27, it must be remembered that the position of the water, as found on July 18, was high on account of the heavy rains which fell during the first twelve days of July. From July 4 to July 13, inclu- sive, 3.27 inches of rain were caught at the rain gage at Garden, Kans. ; the rainfall at Sherlock, Kans., was probably as great, so it is very likely that the level of the water found in the test wells on July 15 40 UN"DERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. and 18 was high owing- to the previous rains. In fig. 13 the results of the flood of Jul}" 27 are shown in greater detail than in the previous diagram. A photograph of a cardboard model showing the changing positions of the water plane at Sherlock is reproduced in Pis. II and 111. The top of each cardboard corresponds to a cross section of the water plane taken across the valley on a certain date, the right side of each card corresponding" with the north side of the valley, the left side corre- FiG. 13. — Elevation of water in Arkansas River and in two test wells near Sherlock, Kans., for various hours during the flood of July 27, 1904. The vanishing influence of the flood with increasing distance from the river is clearly brought out by the diagram. Test well No. 2 is 900 feet north of north bank of river; test well No. 3 is 400 feet north of north bank of river. spending with the south side of the valley. The location of each test well is shown by a vertical line, and the position of the channel of the Arkansas is indicated by the level segment of each card near the mid- dle of each section. The model shows to the eye the way in which the river and the water in all of the test wells gradually fell from July 13 to July 27, and it also illustrates the influence of the flood of July 27 upon the wells near the river. It also shows that the level of water in well No. 6, one-half mile south of the river, was not influenced by the flood in the river, but continued to fall during the entire period. The decreasing influence of the river on the water plane with the dis- tance from the river is brought out clearly by the diagram (fig. 13). It is apparent from this model, as well as from the one shown for camp 1, that there is a marked tendency for the ground water near the river, especially on the north side, to remain at a lower level than FLUCTUATIONS OK (iHOU ND-AVATER T.KVEL. 41 the water in the river itself. At th(> titue the (lata presented by the iiKxiel were obtained. thei-(^ had l)e(>n water in tiie i-iver for six or seven \V(Hds.s and the amount of ruinfall had been abo\-e tiie average. These facts indicate that the uiuh'ro-round drainage tii rough t\\o sands and gravels is more than sullicient to drain otl' the })recipitati()n, without return seepage into surface streams and without run-oft" from the sui-- face of the ground. The \arious amounts of ground water gained or lost ])y each mile of the valley along i\w river at Sherlock from fFuly 15 to August 1, 1 ;H)4, is expressed in Sections 1 1, III, and IV of Table S (p. 81). For the purpose of making the results as definite as possible the gain or loss for each mile of valley is given as a continuous flow of water expressed in cubic feet per second. Thus, according to the table, the strip of ground between the river bank and test well No. 2, 900 feet north of the river, extending along the stream for a distance of a mile, lost water from July 15 to to July 20 at a rate equivalent to a steady flow of water equal to 2 cubic feet per second. During the flood on July 27 this same strip of countr}- absorbed water from the river during •the first two hours of iiood at the rate of 5-1 cubic feet per second. The rate of gain during the three following periods of two hours each was 72, 65, and 32.4 second-feet, respectively. During the eleven hours from 7 p. m., July 27, to 6 a. m., July 28, the rate of gain fell to 1.5 second-feet, after which the ground lost water. These results, and similar results for the south side of the river, are given in the table. Putting all of these results together we can compute the amount of water furnished to the sands by the flood in the river as follows, the computation appl3dng to 1 mile of the river valley only: Water furnis^hed to sanch near Slieduck, Kcms., hy Jtood of Arkansas River. North of river: Cubic feet. July 27, 11 a. m. to 1 p. m., 2 hours, at 54 cubic feet per second . . . 389, 000 July 27, 1 p. in. to 8 p. ni., 2 hours, at 72 cubic feet per second 525, 000 July 27, o p. m. to 5 p. m., 2 hours, at 65 cubic feet per second 467, 000 July 27, 5 p. m. to 7 p. in., 2 hours, at 32.4 cubic feet per second . . 234, 000 July 27, 7 p. m. to 6 a. m. July 28, 11 hours, at 1.5 cubic feet per second 59, 500 Total gain « 1, 674, 500 « ____^;^__^^ South of river: July 27, 11a. m. to 1 p. m., 2 hours, at 63.8 c .bic feet per second . 459, 000 July 27, 1 p. jn. to 3 p. ni., 2 hours, at 28.9 cubic feet per second .. 208, 000 July 27, 3 p. m. to 5 p. ni., 2 hours, at 13.4 cubic feet per second .. 96, 500 July 27, 5 p. m. to 7 p. m., 2 hours, at 1.34 cu])ic feet per second . . 9, 650 Total gain 773, 150 July 27, 7 p. ni. to 8 a. ni. July 28, loss at 0.22 cubic foot per second. 10, 296 Net gain b 762, 854 Total gain both sides of river '•2, 437,354 a Equals 38.4 acre-feet. t> Equals 17.6 aore-feet. '• Equals 56 acre-feet. 42 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. The gain of 56 acre-feet took place on land having an area of 175 acres. The above results show the gain between test well No. 2, 900 feet north of the river, and test well No. 5, 550 feet south of the river. There was some gain in ground water in the lands north and south of these boundaries, but the data are not at hand for the computation. The susceptibility of the adjoining lands in receiving seepage water from the river was greater on the north side than on the south side of the river. FLUCTUATION OF GROUND-WATER LEVEL AT DEERFIELD, KANS. Observation of the ground-water level was made at camp 3, near Deerfield, in three test wells. The location of these test wells appears on the map, fig. 6. The water in the river occupied but a small part of the river channel during most of the time during which these obser- vations were made, and therefore the distances of the test wells from the edge of the flowing water are given in fig. 19, in preference to the distances from the river bank. Test well No. 1 was 1,100 feet, and well No. 2, 1,730 feet south of water in the river. Test well No. 3 was 1,100 feet south of the river, but 1,000 feet upstream from test well No. 2. 6 Table 10. — Elevation of water in river and test ivells at Deerfield, Kans. Date. Time. Elevation of water in well No. 1, 1,100 feet from river. Hy- draulic gradi- ent, per mile, from well No. 1 to well No. 2. Elevation of water in well No. 2, 1,730 feet from river. Hy- draulic gradi- ent, per mile, from well No. 2 to well No. 3. Elevation of water in well No. 3, 1,100 feet from river. Hy- draulic gradi- ent, per mile, from river to well No. 3. Elevation of water in river. 1904. August 4 August 5 9 a. ml do Feet. 2, 923. 02 2, 923. 14 2, 923. 21 2,923.23 2, 923. 23 2, 923. 23 2, 923. 23 2, 923. 27 2, 923. 29 2, 923. 32 Feet. 0.25 .17 .50 .50 .34 .42 >.25 .17 .08 .00 Feet. 2, 922. 99 2, 923. 12 2, 923. 27 2,923.29 2, 923. 27 2,923.28 2, 923. 26 2, 923. 29 2, 923. 28 2, 923. 32 Feet. 8.7 9.0 8.8 8.4 8.5 8.5 8.7 8.8 Feet. 2,924.57 2,924.75 2,924.87 2, 924. 82 2,924.83 2,924.83 2, 924. 84 2,924.89 Feet. -1.10 1.44 2.98 -1.58 Feet. 2, 924 80 2, 924. 46 August 6 do 2, 924. 25 Augusts Do... 7.30 a. m 10 a. m 12m 2,925.10 Do Do 4.30 p. m 9 a. m 2.30 p. m 7.30 a. m August 9 Do -1.49 2, 925. 20 August 10 8.7 2, 924. 91 1.73 2, 924. 55 The chart given in fig. 14 shows that a flood on August 7 in the river had no influence upon the water level in any of the wells, although frequent observations were made to detect such influence. The diagram likewise shows the efl^ect of the rain in raising the ground water as shown b}^ all of the wells from August 4 to August 7. Dur- FLUCTUATIONS OF GROUND-WATEK LEVEL. 43 ing" this same interval the river was falling, while the ground water was rising. The rainfall was measured at camp l)y catching i-ain in a tin bucket and correcting for difference in area between top and bot- tom of ])ueket. The ol)served i-ainfall on August 4 and August 5 amounted to about l.Tf) inches. The water in the various test wells rose b}' the following amounts between August 4 and August 6: Test well No. 1, 0.17 foot, or 2.02 inches; test well No. 2, 0.29 foot, or 3.48 inches; and test well No. 3, 0.30 foot, or 3.()0 inches. If we assume that the soil had a porosity of 33i per cent, these observed changes in the level of the water plane are equivalent to actual increments of 0.7, 1.16, and 1.2 inches, respectively. These amounts will average 1 r N/ A r^ wcr ^ L V -^ \ \ U r N — 1 tk¥e) k i \ N.^ ■~<\^ / \ V V ^ J V y /^ ■^ V. -o // UJ =0- o=— -o r^ Jl>s f /Vo./Tj^ ^ St tviU No.2\ fa/n 1 AUGUST Fig. 14. — Elevation of water in Arlcansas River and te.st wells at Deerfleld, Kans., August 4 to 14, 1904. Test well No. 1 is 1,100 feet south of stream. Test well No. 2 is 1,730 feet south of stream. Test well No. 3 is 1,100 feet from stream and 1,000 feet from test well No. 2. almost exactly 60 per cent of the rainfall for the two days, August 4 and August 5, 1904. This result gives very direct proof of the excel- lent quality of the catchment area furnished by the sand}' bottom lands on the south side of the river at Deerfield. EVAPORATION EXPERIMENTS NEAR DEERFIELD. The table of meteorological data below has value in showing that a considerable amount of stored ground water is lost in the first bottoms of Arkansas River bv" evaporation. Although these measurements extend over only a very brief period, they are sufficient to establish 44 UNDERFLOW IN ARKANSAS VALLEY^ WESTERN KANSAS. the fact that the loss of ground water by evaporation is about ten times as great where the water is within 1 foot of the surface of the ground as it is where the water lies at a depth of 8 feet. The pump- ing plants that materiallj^ lower the ground water in the bottom lands will thus save a considerable amount of water that now goes to waste in evaporation and in supplying the rank growth of wild grasses that flourish in the first bottom lands. It is safe to say that this savable loss amounts on the average to a foot of water for each acre of first bottoms for the months of July and August alone. The following is a record of observations of evaporation from three tanks filled with natural soil in which the water plane was kept at a constant depth, compared with the evaporation from a tank of open water. The tanks were located in the bottom lands of Arkansas Valley, near the head gates of the Farmers' ditch. The soil is a sandy loam changing to coarse sand at a depth of about 3 feet. Meteorological records at Deerfield, Kans., from July 3 to September 8, 1905. Week of- July3-9a July 9-16 July 16-23 July 23-30 July 30-Aug. 6. Aug. 6-13 Aug. 13-20 Aug. 20-27 Aug. 27-Sept. A Sept. 3-8 a Per cent Rain- Vapor of fall in pres- relative inches. sure. humid- ity. 0.11 0.0 .440 47.3 0.08 .482 .50.2 1.24 .560 61.2 1.50 .568 63.9 0.38 .478 54.8 0.05 .530 57.3 0.0 .520 49.3 0.03 .395 41.4 0.71 .489 60.4 Evaporation in inches. Velocity of wind 16.50 15.89 16.13 19.78 12.05 13. 62 13, 26 19.58 17.19 14.54 Open water. 2.53 2.39 1.80 2.45 2.22 3.04 3 19 1.21 1 foot to water, soil culti- vated. 1.48 1.34 1.14 0.92 0.70 1.73 1.21 1.38 1.51 0.87 2 feet to water. 0.65 0.60 0.49 0.49 0.60 3 feet to water. 0.13 0.13 0.23 1.40 0.05 0.43 0.17 0.08 0.12 0.04 a Week incomplete. (Ml APTKH III. CIIKMICAT^ COMPOSITION OF TIIK AVATKRS OF THE UNDERFI^OAV. Cheniiciil tests of the oTound waters were made wherever possible during the process of the woriv. Portable t::ld apparatus was at hand which could be used in inakino- a few simple tests. The determina- tions made included titrations for chlorine, alkalinity, and hardness. Total solids were determined by means of the Whitney electrolytic bridge. The curve of total solids used in this case was obtained by evaporating a sample of water containing U5.!» parts per loO,00() total solids. The results of the test are brought together in Table 11, and the curve used for the determination of the total solids is printed as tig. 1.5 (p. 47). Tablk 11. — Analyses of ground water in the Arkansas Valley, irestern Kansas. WEST OF GARDEN, KANS. Date. Chlorine (parts per 100,000). Alkalin- ity as CaCOa (parts per 100,000). Degree of hard- ness (parts per 100,000). Total solids (parts per 100,000). Temper- ature. Depth of well. Location. 1904. June 16 4.61 14.0 21.35 °F. Feet. June 20 71.0 Do. 5.31 5.85 21. 79 8.10 8.51 11.00 8.51 31.0 13. 75 22.9 16.5 19.0 22.0 17.0 24.5 15.0 20.0 16.1 11.4 12.9 11.9 14.7 20.4 23.0 19.5 22.0 20.5 17 5 30.9 49.1 25. 3 93 105 49 33 119 127 113 67.0 88.0 55. 58.0 48.0 51.0 58.5 Do. Do. July 11 July ti July 5 10 9 28 17 17 15 15 32 32 58 48 56 Windmill .south of riyer. Station 12. June 28 Do 34.7 37.6 33.9 38.65 43.20 40.51 39.5 13.9 21.6 14.6 33.5 38.8 53.5 47.9 Station 8. Station 10. June 21 June 20 Station 4. Station 2. Do 6.72 4.96 6.00 3.05 2.70 1.67 4.59 5.42 13. .50 12.80 9.60 21.30 6.72 52.0 Station 1. June 18 . Station 3. Julys 121 37 32 36 105 114 52.0 52. 55.0 .53.0 55.0 62.0 Station 6, \vell A. Do Do. July 9 Station 6, well B. Do Dn July tj 30 Stfltinn n Do 16 5 5 3-4 3-4 3-4 3-4 12 June 16 camp. Do. June 16 Do. Do 48.0 48.0 36.0 39.1 39.5 Test well No. 1. June 20 59.0 62.5 60.0 52.0 Do. Do Test well No. 3. June 23 June 28 13.12 j 24.0 11.00 1 20.0 121 126 Do. New well (camp). 45 46 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. Table 11. — Analyses of ground water in the Arkansas Valley, western Kansas — Cont'd. WEST OF GARDEN, KAN8.— Continued. Date. Chlorine (parts per 100,000). Alkalin- ity as CaCOa (parts per 100,000). Degree of hard- ness (parts per 100,000). Total solids (parts per 100,000). Temper- ature. Depth of well. Location. 1904. June 16 8. 88 10. 62 .78 .67 2.06 4.2 2.1 5.1 4.1 3.4 11.4 17.6 1.2 19.5 22,5 13.1 13.6 19.9 19.2 11.4 18.0 20.5 18.1 19.2 22.7 18.5 39.9 "43. 7 10.7 11.2 2.5.6 53.3 31.1 82.0 31.2 27.9 39.3 45.9 21.3 °F. Feet. 12 14 New well (camp). Do Jnlv 7 6 6 35 86 57 119 102 68 76 150 26 65.0 57.0 Sand hills sec 36 T 24 Do S., R. 34 W. Do September 22 . . 16 25 20 40 36 a 13 115 35 a 30 Sec. 2 T. 23 S R 33 W 1905. January 24 Do Shultz. Do L. C. Working. Do Do Do Eaye. Do Do Frank Kolbus. GARDEN, KANS. 1904. September 22 . . Do Do 1905. January 24 0.92 .85 3.96 1.6 14.1 15.9 20.3 18.8 25.6 30.0 69.2 29.5 16 16 80 42 130 110 16-f40 78 Atchison, Topeka and Santa Fe R. R. well. Carter's well. City waterworks well. S. L. Leonard. SHERLOCK, KANS. 1904. July 16 July 22 July 16 July 19 July 16 July26 July 18 July21 July 30 July 22 Do July 23 July 16 Do July 15 July 27... Do July 19 Do September 22 . . 4.04 13.20 13.90 27.70 37.90 73.0 74.0 71.0 73.0 3 85 .89 21.20 13.09 27.0 63.0 8 .60 17.50 4.64 30.0 58.5 18 .58 21.50 2.38 56.0 60.0 8 1.10 17.85 26.20 42.0 56.0 26 3.62 16.75 27.30 36.0 57.0 22 2.46 21.30 ■ 28.10 55.0 56.2 36 4.61 19.45 44.70 83.0 65.0 10 4.58 15.90 40.60 80.0 .56.0 22 4.05 15.90 42.90 78.0 57.0 14 5.20 17.45 46.30 104.0 53.0 20 3.47 14.65 30.00 93.0 57.7 18 5.10 15.75 31.10 97.0 65.0 22 5.18 15.50 15.26 107.0 96.0 54.0 66.6 IS 28 4.97 48.6 4.90 16.25 50.6 97.0 67.5 28 .96 16.86 19.00 21.30 20.0 26.9 29.9 21.0 37.0 44.0 .17 2.24 40 River water. Do. Test well No. 6. Station 16. Test well No. 4. Station 20. Station 16. Station 17. Near station 17. Station 18. Do. Station 19. Station 14. Do. Station 13. Station 21. Station 22. Sec.30,T.24S.,R.34W. Sec.20,T.24S.,R.34 W. Sec.30,T.22S.,R.33 W. a To water. CHEMKWTi rc^M POSITION ol' THp; WATERS. 47 Table W.—AiKih/neK of (jrunnd mttcr in tlir Arhtnxus Vdlli'ij, rveMerti. A'a».ws— Cont'd. DEEKl'IKlJi, K.VNS. Total solids Alkalin- Degree Chlorine ' ity as of hard- 1 parts per CaCO:, I ness /p„rtspcr 100,0CX1). (parts per (parts per ^'^, f^. 100,000). 100,000) ^^''^'> 1904. September 22 . . 1.49 August 6 ! • 2.60 A\igust 10 1 2.45 August 9 j 5.00 August 4 7.60 Do 6.64 August 8 j 5.11 August 4 1 8.61 August 5 '' 5.32 August 6 .^. 39 15.1 14.7 17.7 15.7 16.2 15.3 16.0 17.7 l.'S. 5 16.7 31.2 2S. 9 32.7 51.2 .55. 2 .57. 48.4 65.9 44.4 48.3 74. 95, 117 114 90, 117 108 106, Temper- ature. °F. Oti. 60.0 .56.0 .59. 5 .58.0 57.0 59.5 59.0 57.0 Depth of well. Feet. Location. NE. quarter sec. 26, T.24 S., R. 35W. S\V. quarter .sec. 24, T. 24 S., K. 35 W. Near station 28. Station 27. Well at camii. Station 23. Station 26. Test well No.l. Station 24. Station 16. \\ \\\ \\\ \ w \\ \v \ \\ v \ \ V\ \\ \N \^ l\ \ s.^ i^- ^ N N <^ V ^ ■^ ] V "^ b^ ^ ~~^ =^ rr: 'ZZZ :::--; ~ ■ IC l(. W 3C *0 St O 6i 70 O BC O 9<3 o la. ■>0 II 70 12 OO 13 OO /* 00 li OO 16 00 '7 00 /0 ZOO Fig. 15.— Curve for Whitney electrolytic bridge used in converting resistance in ohms into total solids for ground waters of Arkansas Valley. A comparison of the results of the tests at various stations shows a marked decrease in the (juantity of dissolved solids in the water w ith the depth at which the sample was taken. In forcing down test wells at almost an_y point in the bottom lands of Arkansas River the increas ing softness of the water can be noted almost from foot to foot. At a considerable depth, say from 60 to 100 feet or more, there are found waters which are popularh^ called in this region "second" or "third'' waters, which are verv much softer than the water obtained from 48 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. shallow wells. At points located in the sand hills south of the river there are places where shallow wells furnish water much softer than the so-called second or third waters found in the vicinity of Garden. The total solids in the ground water determined at wells in the first camp, 2 miles west of Garden, varied from 121 parts per 100,000 for water taken ■! feet below the water plane to 103 parts per 100,000 for water taken at 6 feet, and 80 parts per 100,000 for water taken at 14 feet. Water taken from the railroad well,. 130 feet deep, at Garden, showed total solids of 16 parts per 100,000. Water in the sand hills south of the river at a depth of 9 feet showed 33 parts per 100,000 total solids, and another well, deeper, but of unknown depth, showed 6 parts per 100,000 total solids. The tendency of the ground water near the surface in the bottom lands of the river to run high in solids seems to indicate that this increased hardness is due to the loss of the ground water hj evaporation. The water plane in these bottom lands lies close to the surface of the ground and is subject to frequent fluc- tuations due to rain and changes of conditions in the river itself. These changes are sufficient to account for a large excess of dissolved solids in the surface waters, and it is believed that no other explana- tion is necessary. As the ground water moves downstream, the vari- ous filaments of moving water must thread themselves around the grains of sand and gravel, continually dividing and subdividing the water as it moves through the capillary pores. The effect of this action is to slowly work the concentrated water near the surface down to greater depths, forming a ground water of graduated strength. Every layer of silt, cla}^, or other impervious material which possesses a considerable area acts as a partition, separating the. moving ground water into layers which do not mix, except where the impervious strata give out. This results in layers of water of distinct difference in total solids, which are locall}^ known as "first," "second," and "third" water, etc. In the following table (Table 12) the various samples of ground water are classified by depth of the wells, and the averages of the dif- ferent determinations are tabulated. From this arrangement a com- parison is possible between the waters of different depths, in which the errors due to special peculiarities of particular wells are partly eliminated. Some of the well water taken from stock or domestic wells showed marked pollution, but all such samples have been included in the table. CHEMU'AI. ('OMl'OSIIION OF lilK WATKRS. 49 T.vm.E 12. — (Jititliti/ of i/roi(iitl initrr III ArLniisdN Rinr VaUi'ii, &£ [\ r\ \ 1 ^ \ y X!2! ■ /itt/i ^Curi V ent- clear waU r v_ v 5 lb 17 IS 19 ZO Zl ZZ Z3 Z4 Z5 Z6 Z7 ZB Z9 30 31 I Z 3 ♦ .JULY 1 AUG \ Inch z es V IN FA lL a- ■ GAR 0£N IS 16 17 IB 19 20 21 ZZ Z3 Z* Z5 26 Z7 ZS Z9 30 31 \ 1 Z 3 * JULY 1 AU6 1 Fig. 16. — Elevation of water surface of Arkansas River at Sherlock Bridge, compared with rainfall record at Garden, Kans. of the water in flowing torrents. In consequence the rainfall is cora- pletel}^ taken care of by absorption into the ground and by evaporation and use b}^ the vegetation. Eastward from the High Plains region rainfall is greater, and the sod is not able to prevent the formation of rills and eroded channels, so that much of the water runs off' into sur- face streams. Westward from the High Plains district, as Colorado is approached, the rainfall decreases and in consequence vegetation becomes so scant that it is not able to protect the surface of the ground from erosion even from a diminished rainfall. Hence it is that both to the east and west of the High Plains there is a marked run-off, but in the plains district proper the rains are disposed of by absorption. ORIGIN AND EXTENT OF UNDERFLOW. 53 The above facts ar(> well shown hy the i-esults previously diseiissetl ill this pai)ei-. The summer of l!»(>4 was one of unusually ample rain- fall in the plains, ami many Moods eanic down the river. The river was carefully watched by the tield party and its elevation noted. Fitfs. 1(5 and IT show the elevation of the river at Siierlock and Deer- tield bridges, respectivelv, compared with the rainfall at Garden. A simihir diagram for cimp 1, near Garden, is given in fig. 10. A study of these diagrams shows practically no inflnenc - of the rainfall upon 2526.0 ZSZS.O Z9B4.S Fig. 17. — Elcviitiiiii of wiittT surl'ac ni Arkansas River at Deertield Bridge, compared with raiulall record at Garden, Kans. the stream. Man}^ of these rains extended into Colorado, where they were the cause of floods that showed themselves at the camps in Kansas many hours after the rain. Thus we have ample evidence of no run-ofl' from the country between Garden and Deertield, and at the same time have proof of a considerable run-otf from the watershed toward the western limit of Kansas and in Colorado. The few instances in which small surface streams are formed near the Colorado line — like the plains streams known as Bear Creek and White Woman Creek — are no exception to the statement above that 54 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. there is no run-off into the Arkansas in the High Plains district, for these streams entirely disappear as surface streams before the Arkansas is reached. Their waters, less the evaporation, are ultimately joined to the underflow. The situation may be summarized in the following- words: The underground drainage in this region is so enormous, and the water passes through the gravel so freely, that there is no surplus water left to form surface streams, or to form a perennial supply for Arkansas River. If the gravels of the plains near Garden were less deep, it is entirely conceivable that the Arkansas River would be a perennial spring-fed stream at this point. The large contribution to the underflow, which is made by the rainfall upon the sand hills south of the river, is clearly demonstrated by the course of the contours in fig. 5. In this diagram the soft water from the south side of the river can be observed to be pressing the hard water of the first bottoms northward toward the left side of the river valley. Annual precipitation at Dodge and Garden, Kans. Year. Dodge. Garden. Year. Dodge. Garden. 1875 10.78 15.40 27.89 17.96 15.43 18.12 33.55 13.14 28.50 30.36 23.71 19.35 15.71 22. 94 19.17 1890 . 11.72 32.34 19.66 10.12 12.60 20.31 19.87 21.58 31.46 28.45 20.76 16.06 17.70 15. 27 17.19 1876 1891 . 27.21 1877 1892 . 1878 1893 1879 1894 11.45 1880 1895 1896 ., 1897 . . . 1881 1882 1883 1898 . 28.75 1884 1899 . . . 20.58 1885 1900 ; 1901 1902 19.29 1886 18.34 1887 19.65 1888 1903 20.64 1889. . 1904 21.05 NORTH AND SOUTH LIMITATIONS. A noteworthy feature of the underflow is the lack of any natural north or south limitation to the easterly moving stream. There are important changes from place to place in the north and south slope of the water plane, but none are of sufficient consequence to materially modify the dominant influence of the easterly gradient of T to 8 feet to the' mile. The velocities found at the edge of the sand hills to the south of the river, and at a distance as high as 9 miles from the chan- nel of the river, are about the same as those found near the bed of the river in similar material. There is nothing surprising in this except that the stratification of the sand and gravel on the High Plains is such that there is no natural north or south limitation to the eastward- moving ground waters. CHAPTER V. SUMMARY OF TESTS OF SMALIj PUMPI:N^G PLANTS I:N^ THE ARKANSAS VAT. LEY. GENERAL RESULTS. Table li shows the results of tests of a number of pumping plants used for irrigation in Arkansas Valley between Garden and Lakin, Kans. Most of the entries in the table explain themselves. The fuel used in most of the plants is gasoline, the current price of which during the summer of 1904 was 22 cents a gallon, a cost that is almost prohibitive, even when pumping water from the most excellent wells found in the valley. Table 14. — Tests of small pumping plants, Arkansas Valley, Kansas. 1 2 3 4 5 6 - 7 Owner of plant. Location. Kind of pump. Horse- power of en- gine. Fuel used. Price of fuel per gallon. Total lift. D.H.Logan Garden, Kans. do No. 3 centrifugal 6 10 1^ 7 2a 14 Gasoline.. do.... do.... do.... do.... SO. 22 .20 Feet. 22.1 Mrs. M. Richter lii S C.E. Sexton do 2 vertical 6 by 16 cyl- inder. Chain and bucket do .22 15.0fi Nathan Fulmer Lakin, Kans . . do .21 .22 17.0 J. M. Root 15.8 King Bros Garden, Kans. do No. 4 centrifugal 63.0 Waterworks I. L. Diesem do No. 4 centrifugal No. 3 centrifugal No. 14 centrifugal 2 horizontal 6 by 5 cylinders. No. 4 centrifugal 10 6 80 3i 5 Gasoline.. do.... Coal Gasoline.. do.... .121 .12J 22.13 L. E. Smith do 17.fi0 H. B. Holcomb Sherlock, Kans Garden, Kans. do a 4. 00 23. H. S. Kipp .12^ 21.7 J: R. McKinney .12i 21.47 a Price per ton. 65 56 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. Table 14. — Tests of synall pumping plants, Arkansas Valley, Kansas — Continued. 1 8 9 10 11 12 13 14 Owner of plant. Distance water is lowered Yield of well per minute. Specific capacity of well per min- ute. Area of per- colating or strainer surface. Specific capacity per square foot of strainer per minute. Cost of fuel per acre- foot of water. Cost of fuel per 1,000 foot- gallons. Feet. 6.85 5.3 3.0 6.35 4.16 20.3 5.48 6.72 2.16 9.60 2.83 8.39 Gallons. 272 394 91 540 215 183 290 363 198 2,300 96 420 Gallons. \ Sq. feet. 42.2 107.0 Gallons. 0.394 .27 .53 .254 .246 .106 .31 .356 1.290 .128 .75 .42 $2.93 2.90 3.75 1.37 2. 78 Ce7its. 73.0 30.3 85.0 51.7 9.0 77.0 54.0 91.6 240.0 34.0 60.0 266.5 57.2 334.0 210. 85.0 247.0 151.0 70.7 . 1,876.0 45.3 116.0 ■ ,V C.E. Sexton Nathan Fulmer IS J. M. Root King Bros 2.10 1.67 a. 85 1.09 1.20 ^T L. E. Smith H. B. Holcomb H. S. Kipp ^5 J. R. McKinney "Including cost of labor and lubricating oil. SPECIFIC CAPACITY. The numbers in column 10 express the readiness with which the well furnishes water to the pump. The numbers in each case were found by dividing the numbers in column 9 by the corresponding numbers in column 8; these numbers, therefore, express the amount of water the well would furnish if the water level was lowered but 1 foot. These numbers constitute what the writer has called the "specific capacity" of the well, and are large in the case of a good well and small in the case of a poor well. The water-bearing gravels are usually from 9 to 15 feet below the surface of the ground, and good wells can bever}^ cheaply constructed. There is no quicksand or hardpan or other troublesome material above the water-bearing gravels. The well tubes or strainers are usually 12 to 20 inches in diameter, and are made of slotted galvanized iron. For the most part the wells are of the very best design and possess a remarkably high specific capacity; the writer knows of few places where better ones can be constructed. The usual construction consists of a dug well, 6 to 10 feet in diame- ter, excavated several feet below the level of ground water, with a num- ber of "feeders" or tubular wells penetrating the bottom of the well. No better construction can be suggested for small plants. The only modification in detail that seems likely to better the present excellent results would be the use of galvanized-iron strainers with . larger slots than are at present in use. This would be practicable at some of the wells. Heavy pumping would remove much of the fine material that now remains in contact with the present well strainers. » TKSTS OK SMALL PUMPING PLANTS. 57 III column I'J (here are <;iviMi tli(^ saiiic inatiiiituclos as arc expressed in colunin UK ihhIucihI in oach caso to I s(|iiarc foot of well strainer. Tlio nunit)ers in tliis cHtlunm express, tiierefore, the amount of water in gallons per minute furnished l)v 1 s(inare foot of well strainer under a head of 1 foot of water. Thcv arc a numerical expression of the degree of coarseness of the material in which the well is placed. These numbers are almost the same for all of the w-ell plants, when proper allowance is made for difference in construction. At the Kichter, Fulmer. and Root plants, there are large dug wells with sev- eral feeders in the bottom. The numerous feeders interfere with each other somewhat, keeping the specific capacity lower than it would otherwise be. At the Logan and Sexton plants the construction is different. The Logan well is constructed of 20-inch casing, through the bottom of w^hich are two 4-inch feeders extending 2() feet below the bottom of the 20-inch casing. The 20-inch casing is perforated for 10 feet at the bottom. At the Sexton plant there is a 12-inch W'Cll 22 feet deep, and a 10-incli well 31 feet deep, both perforated 10 feet from the bottom. COST OF PUMPING. While the cost of water at these various pumping plants may at first glance seem high, and the results not especially encouraging, yet a more careful inspection shows that the facts are really highh^ favor- able. It must be remembered that the cost of pumping is based upon a 22-cent price of gasoline. This price is almost prohibitive, but for- tunateh" there exist several possible wavs of cutting down very mate- riall}" the cost of power, and on this point the following suggestions are offered: In the first place, the cost of pumping can be reduced by the use of crude oil in place of the gasoline. Crude oil from Kansas fields should be laid down at Garden at from 3 to 4 cents a gallon. The crude oil recpiires a special device, which must be used in connec- tion with the gasoline engine, called a generator, in which the crude oil, or part of it, is converted into a gas before it is led into the engine cylinder. By the use of such a generator the cost of fuel can be lowered to a point about equivalent to a 5 cents a gallon price for gas- oline. The crude-oil generators will work best on engines of 12 to 30 horsepower. If plants of from 20 to 50 horsepower are constructed, as I believe will inevitably be the.case in the near future, the cheapest power will probably be found in the use of coal in small gas-producer plants in connection with gas engines.*^ These small gas-producer plants are largely automatic in action and can be operated b}^ anyone. With hard coal or coke or charcoal at $8 per ton, the cost of power would be less a See test of producer-gas plant, Chapter VI. 58 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. than one-half cent per horsepower for one hour, or only one-fifth of the cost of power from gasoline at 22 cents a gallon. The writer antici- pates no difficulty, therefore, in keeping the cost of water below 60 to 75 cents an acre-foot for fuel, or below $1.25 to $1.50 per acre-foot for total expense. Hundreds of such plants have been put in use in Eng- land during the past ten or more years, and they are in charge of unskilled labor. These gas-producer plants are used in England for a great variety of purposes, such as power for agricultural machinery, and for small electric-light plants for country estates, etc. They are used in as small units as 5 horsepower. In this country the producer-gas plants have been in use for several years, and at the present moment they are fast taking the place of steam power in new plants. The cost of a producer plant and gas engine is about the same as the cost of a steam engine and boiler of same size when everything is included, but the cost of power from the producer-gas plant is very much less than that obtained from small steam engines. In producer plants, ranging upward from 100 horsepower, a style of plant may be installed in which soft coal or lignite may be successfully used. This still further cuts down the cost of power. In fact, large plants of this type furnish the cheapest artificial power that has yet been devised. The saving is not only in fuel, but also in labor, as one man is capable of running a 300-horsepower plant. That part of the operating expense which is properly chargeable to fuel cost can be accurately determined. Column 13, Table 14, expresses the cost per acre-foot of water recovered. In column 14 is given the cost of fuel for lifting 1,000 gallons of water 1 foot. For the purpose of comparison, these results are expressed in fractional parts of a cent. It should be noted that the cost given in the table is based upon a 22-cent price for gasoline. There is no doubt but that producer-gas plants in moderate-sized units would enable irrigation by pumping in the bottom lands of Arkansas River to be highly profitable. No allowance has been made for interest, depreciation, and labor. These expenses, if included, would about double the cost per acre-foot. CHAPTER VI. D15TAII.S OF TESTS OF PUMPIT^G PT.A:NTS. TEST OF PUMPING PLANT OF D. H. LOGAN, GARDEN, KANS. This plant is located in the northeast corner of sec. 13, R. 33 W., T. 24 S. , and is in the northwest corner of the city of Garden. The outfit consists of a 6-horsepower Fairbanks, Morse & Co. horizontal gasoline engine connected by a ])elt to a No. 3 centrifugal pump. The well is constructed of 20-inch galvanized-iron casing- 32 feet long, per- forated 10 feet up from the bottom, inside of which are two i-inch feeders 28 feet long, perforated their entire length, and extending 26 feet below the bottom of the 20-incli casing, making a total depth of 58 feet. The pump has been in operation since April, 1902, and the engine since April, 1903. The water was measured b}^ the use of a fully contracted weir with a length of crest of 0.66 foot. The engine was started at 9 o'clock and the weir was ready for water at about 10.30. The water was turned on weir and the head read until it became constant at 1 p. m. In order to determine the expense of pumping, all of the gasoline was used out of the reservoir, then 1 gal- lon was poured in and the length of the run noted to be one hour and thirty-two minutes, or two-thirds gallon per hour. As the engine is a 6-horsepower one, this equals 0.111 gallon, or 0.'145 quart of gasoline per horsepower hour. The average corrected head on the weir was found to be 0.440 foot. Using weir formula « where /> = 0.66, whence c=:0.592, the discharge is found to be ^ = 0.6045 second-foot = 272 gallons per minute. Data of Logan pumping plant, Garden, Kans. Feet. Average depth to water while pumping 18. 6 Normal depth to water 11. 75 Amount lowered by pumping 6. 85 Elevation of well platform 2, 835. 26 Distance water was raised above platform 3. 5 Lift, *or total distance water was raised 22. 1 Total area of well strainer, 107 square feet. 59 60 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. The fuel cost of pumping was, therefore, 0.9 cent per 1,000 gallons of water recovered, or $2.93 per acre-foot. The cost of 1,000 foot- gallons (1,000 gallons raised 1 foot) was, therefore, 0.0406 cent, or one twenty -fifth cent. The specific capacity of the well is 42.2 gallons a minute, or 0.394 gallon for each square foot of well strainer. The engine ran at a speed of 350 revolutions a minute, exploding 143 times a minute. The diameter of engine pulley is 16 inches and of pump pulley 10 inches. This gives a speed of 560 revolutions a minute to the pump. The size of the pond was 40 feet by -60 feet, mostly covered with a green scum, which would prevent evaporation. As to seepage, the pond falls 8 inches in twelve hours at night. The pond being 2,400 square feet in area, the observed seepage represents a loss of 16.68 gallons per minute, which should be added to the capacity of pump and well, but not to the effective capacitj^ for Mr. Logan. There is a windmill at a well 20 feet north of the one pumped by the gasoline engine — a 12-foot airometer connected to a 10-inch pump of 12-inch stroke. After the weir measurements were completed the windmill was thrown into gear. There was a brisk wind from the south and the pump threw a good quantity of water, but no appreciable lowering of the water in the gasoline-engine' well 20 feet away was detected. The rise of the water in the well was obtained twice. Below are the two sets of observations: Rise of tvaler after cessation of pumping in Logan well, Garden, Kans. FIRST TRIAL— WINDMILL NOT RUNNING. Time. Depth to water. Time. Depth to water. 55 seconds Feet. a 18. 60 16.05 14.55 12.95 12.50 9 minntps and 8 sficnnds , , Feet. 12.35 1 minute and 5 seconds 2 minutes and 22 seconds 12.35 1 minute and 20 seconds 2 minutes and 33 seconds. 12.25 1 minute and 37 seconds . 2 minutes and 48 seconds. . ... 12.15 SECOND TRIAL— WINDMILL RUNNING. (a) 18.0 16.5 14.35 13.10 12.90 12.55 25 minutes and 48 seconds 12. 55 26 minutes 12.45 26 minutes and 23 seconds 12.25 26 minutes and 58 seconds 12.25 27 minutes and 15 seconds 12.25 27 minutes and 30 seconds 12.25 25 minutes and 38 seconds a Stopped pumping. DETAILS OK THSTS OK PITMPINO I'LAN'l'S. 61 Tlio curves showing- tlio rate of i-isc of water in the Loi^aii well after puiiipiiio- ceased are o-jven as curves I and 2 in li*;-. Is. C'urve 2 is the one wliicli ^svas produced when the windmill was punipin^- from a well 120 feet away. The comparison of this curve with curve 1. which was produced when the ncio"hl)orino- well was not used, is ver}' inter- esting-, showino-, as it does, a less rapid rise when the neighboring^ well was in use. To tind the specific capacity for the Logan well from these curves we nuist substitute the valiu\s of the various constants in the fornuila AH t'=l7.25 y log 7 ga-.ons per minute. /eve/ .5 4-5 f^M. -.30P.M ^ 'I'nc/m/// 1 ufjn/n^ y ^ fi/ / Fig. is.— Rising furve.s for Logan well. Curve 2 taken when neighboring well was being pumped by windmill. Curve 1 taken when windmill was shut off. The value of the area. A, of cross section of the Avell casing is 2.17 square feet, and H, the amount the water is lowered l)v the pump, is 6. 85 feet. The amount of depression, //, of the water level below the natural level at any time can then be selected from the curve, and the specific capacit}^ readil}- computed. If t be taken to be 40 seconds, or f of a minute, h will be found from the curve to be equal to 6.85-5.5 = 1.35 feet, hence c — 17.25 X I X 2.17 X log ( Y\k ) gallons per minute = 39.5 gallons per minute. The yield of the well for the maximum depression, 6.85 feet, must then be 6.85 X 39.5 = 270 gallons per minute. 62 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. The curve of rise of water forms one of the best methods of deter- mining- the yield of a well. Such curves can readily be obtained. Well data should always include measurements of the amount of lowering of the water surface by the pumps, and it is only necessar}^ to continue these measurements after the pumps have stopped to secure sufficient data to estimate the specific capacity and total yield of the well. This avoids the necessity of constructing- a weir or other method of meas- uring the water discharge. The accuracy is sufficient!}^ great for the purpose for which such data are used. The method can be used only in cases where an internal suction pipe extends into the well casing with sufficient room around it to permit a sounder to be lowered to the water surface. If there is no foot valve or other means for prevent- ing the water from returning to the well after pumping ceases, the rising curve may still be used for the determination of the specific capacity, provided that only the portion of the curve be used which was formed after the water had completely returned to the well from the pump. TEST OF THE RICHTER PUMPING PLANT, NEAR GARDEN, KANS. This plant is located in the northwest corner of SW. i sec. 14, R.. 33 W. , T. 24 S. The upper part of this well is cased with part of the old standpipe .from Garden. The casing is 10 feet in diameter and extends down 20 feet. In the bottom of this part of the well are placed four 8-inch galvanized-iron feeders, arranged symmetrically about the center; each feeder is 25 feet long, perforated its entire length, and extends about 2^ feet above the bottom of the large part of the well. The pump used is a Menge pump, which operates on the principle of a screw propeller of a steamship. It bores the water out and up a square wooden penstock or pump shaft. There are two of these pro- pellers mounted one above the other on vertical iron shaft inside the penstock. The top of the iron shaft carries the belt pulley and has a shoulder bearing which takes the thrust of the pump as a pull above. This pump is made in New Orleans. The pump is run by a 10-horsepower Otto gasoline engine, which runs at a speed of 300 revolutions per minute. The circumference of the drive pulley is 5.25 feet, and of the driven pulley 2.65 feet, mak- ing the pump run at 595 revolutions per minute. The screws are boxed up and under water when the pump is not in operation. A small pond was constructed at the end of the discharge trough and a fully contracted rectangular weir of length of crest of 1:2 feet was used to measure the discharge. The measurements for head were taken 6 feet away from the weir, and boards were interposed between DETAILS OK TKSrS oK ITMIMNC I'l.ANTS. 63 tlu' discharge trouiih iiiid weir to cut down (ho velocity, which iiiiohl tend to t^ivo erroneous results. The iiveraj^'e corrected head on the weir was 0.371 foot. Using the weir formula and taking c from Mei'i-iman's tables as 0.(i03, y = 0.87() second-foot — 894 gallons pei- minute. Using a small Price acoustic water meter in the discharge trough, by measuring the velocit}^ at different places and also by integrating, the discharge was found to be 0.76 second-foot, or 342 gallons per minute. The water in the Hume was so shallow that this determina- tion is of little value. By putting chips in the discharge trough and catching the time with a stop watch, the surface velocity was found to be 1.565 feet per second. This number multiplied by 0.8 gives an average velocity of 1.25 feet per second and a discharge of 0.884 second-foot, or 397 gallons per minute. An attempt was made to determine the amount of gasoline used. The reservoir y/as tilled full and the engine run for 1 hour and 36 min- utes, or 1.6 hours. All the gasoline we had, 9i quarts, did not then till the tank. This was at noon, July 6. On the morning of July 7, 9^ quarts were required to completely till the reservoir, a total of 18f quarts or 37i pints for the run of 1.6 hours for a 10-horsepower engine. . The makers claim their engines use one pint per horse- power hour. This would require in this case 16 pints, or less than half of what was actually measured, if the engine developed its full horsepower. A leak in the tank or feed pipe is clearly indicated, so this amount, while being of value to the owner of the plant, is value- less so far as comparative cost of pumping is concerned. Two observations of the rising curve were obtained which plot well together. The lower part of the curve is not accurate, because of the water in the penstock dropping back into the well when pumping ceases. Data of Richter pumpiny plant, near Garden, Kans. Feet. Elevation of the ground at well 2, 846. Average elevation of water in well 2, 836. 8 Average elevation of water in well when pumping 2, 831. 6 Elevation of discharge from penstock 2, 847. Lift 15. 5 Average amount water is lowered by the pump 5. 3 Number of explosions of engine, 126.5 per minute. Total area of surface of well strainers and all percolating surfaces, 226.5 square feet. 64 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. The curves of rise for this well were obtained on two different occasions and are shown as curves 1 and 2 in fig. 19. They plot together very well. To find the specific capacity of the well from the curve, we note the following values of the constants in the formula for specific capacity: A H c= 17.25 -,- log ~ir gallons per minute. The area, A, of cross section of the well casing, less the amount occupied by obstructions, is 76.79 square feet. The amount, H, that the water is lowered by the pump is 5.3 feet: The amount of depres- sion, A, of the water surface below the natural level at any time can be selected from the curve. From the curve, at the close of ten minutes, h equals 5.3 less 4, or 1.3 feet. t-oo 500 60O 700 eoo Time in seconds .Fig. 19.— Rising curves i'or Richter well, near Garden, Kans. Hence the specific capacity 76.79 5.3 c— 17.25 X .^ log :po = 81 gallons per minute. Multiplying by 5.3, the head under which pumping took place, the total yield of the wdl is 81x5.3=430 gallons per minute. The above determination of the specific capacity is inaccurate, since the first portion of the rising curve does not show the true rate of rise of water in the well. The penstock of the propeller pump holds 37.7 cubic feet of water, which immediately returns to the well when the pump is stopped. This amount of water is suflScient of itself to raise the level in the well bv 0.165 foot. For this reason, only that portion of the rising curve should be used which is not influenced by I DKI'AILS (»!■' I'KSl'S (»!-' I'l' iM I 'I .\ ( i I'KANTS. 65 the returning- wiitcr from the ])t'iist<)ck. 'I'lius, if we iiso tliiit part of tho curvt' from /^ 10(» sim-oihIs to /-{'){)() secoiids, we will eliminate the iiiaceurate poi'tion. iMakinj"' this modification, the data arc changed to 11 = 8.7;') feet; 7/ = 1.80 feet; t=)<^ minutes. Com]>utini4- th(> specific ca])acity on this basis, we obtain r=T;) i^ailoMs a minute. jMultiplyiuii- this by .").;>, th(> total (\stimated yield is 388 gallons a min- ute, w hich chcH'Us i-emaikably w ith 8!>4 gallons a minute obtained. TIk^ area of the strainer and bottom of the well is 2'iM).!) s(piare feet. The above specific capacity divided l)y !^«)6.5 gives (».84:1 gallon per minute as the specific capacity pt'r square foot of percolating surface. The engine ran at a speed of 30() revolutions and exploded 125 times per miiuite. This would indicate that it was working at about 83 per cent of its rated capacity. Assuming that such was the case, and that it would then use 83 per cent of the fuel necessar}' to run it at its full rated power (lo horsepower), Ave have 8.3 pints as the probable amount of gasoline used per hour ))v the engine during the test. This, at 20 cents per gallon, would make a cost of 21 cents per hour. This assumption makes the cost of water 0.89 cent per 1,000 gallons, $2.90 per acre-foot, and one-seventeenth cent per 1,000 foot-gallons. TEST OF PUMPING PLANT OF C. E. SEXTON, NEAR GARDEN, KANS. This plant is located at about the center of sec. 13, R. 33 W., T. 24: 8. , and is 1 mile west of Garden. It consists of two pumps of 16-inch stroke, with 6-inch pistons, connected to a walking beam and driven b}^ l|^-horsepower Fairbanks, Morse & Co. vertical gasoline engine. The east well has a 12-inch casing 22 feet deep, and the west well a 10-inch casing 31 feet deep, both casings being perforated for a distance of 10 feet up from the bottom. The pump rods are 2 by 4 timbers. The two pumps discharge into an artificial pond or reservoir, and the flow was measured with a weir at the outlet of the reservoir. The weir was fully contracted with a length of crest of 0.66 foot. The height of water on the weir was measured by placing a stick on the head of the nail and marking the water line on the stick with a pencil, then measuring with a pocket tape ; in the absence of a hook gage this was the best method that suggested itself. The weir heights taken as a measure of the discharge of the pump are those obtained after the water level in the reservoir had become stationary, as indicated by an absence of systematic variation of the IRR 153—06 5 66 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. weir heights. As evaporation would mal^e the results too small, the following- data are important: The size of reservoir is 50 feet by 90 feet, or 4,500 square feet; trees border the north and south sides, with high grass along the banks; brisk wind was blowing from southwest; temperature of air was 80"^, temperature of water, 52°; there was sunshine until about 3 p. m., when it became cloudy and the wind moderated. The east well threw a much smaller stream than the west well, prob- ably due to a leak in the suction pipe, and consequent pumping of air. No air was pumped by the west pump. Measurements to the water surface in the east well were made at five-minute intervals, but no soundings were obtained in the west well. The number of strokes of each pump averaged 24.5 per minute during the test; the number of explosions of the gasoline en- gine averaged 106.2 per minute. The battery used with the engine not working satisfactorily, a gasoline torch was used for ignition. Gage readings of distance to water in well were made downward from a point on the well platform whose elevation above sea level was 2,836.69. Data of Sexton pumping plant, near Qarden, Kans. Feet. Distance to water when level is normal 8.8 Distance to water when pumping 11. 86 Amount water level was lowered 3. 06 Elevation 2, 827. 9 Distance water was raised above point on platform 3. 2 Total distance water was raised (11.86-1-3. 2) 15. 06 Total area of well strainers, 57.2 square feet. The reservoir has been in use for some time and the seepage was probably quite small, a small enough per cent to be negligible. There was no leakage around the weir, or elsewhere. The gasoline tank was filled at the start, and when the run was com- pleted the amount needed to refill was measured, thus getting the amount used by the engine, which was 11 quarts for a run of 9 hours and 37 minutes, or 1.14 quarts per hour, making a trifle over three- fourths quart per horsepower hour. The average corrected weir height was 0.206 foot. Using the formula for a contracted weir and taking from Merriman's Hydraulics the value of the constant c for J = 0.66 and H= 0.206 as 0.611, we have for the discharge ^=0.202 second-foot, = 91 gallons per minute. With gasoline at 22 cents per gallon, or 5^ cents per quart, the expense of an hour's run, not counting gasoline used for ignition tube, is DETAILS OF TESTS OE I'UMPlNt: PLANTS. 67 $0.0r)25 per hour, or ^0,01 ir> per thousand o-jillons of Avator pumped, or $3.7r) per aen^-foot. 'Vho lift heitiu- 1 .").(!() feet, the cost ])er 1,000 foot -<»al Ions is ().()7<) cent, or ahout oiK^-tliirteeiitli cent ])er L, ()()() o-al- U)ns raised one foot. On July S the rise of \vat^^r in the east \V(dl was taken hy means of a thin jMne l)oard stuck down ]»etWeen the casing and pump. The intervals of time were measured with a stop watch. The pine strip was lowered into the well until the water was reached, after which the t)oard Avas drawn up, the wet line marked, the time recorded, and the l)oard replaced, the observations being repeated as fast as possible. The distances marked on the strip were measured later. Rise of water after cesi^ation of pump nig in Sexton veil, near Garden, Kans. Time. Rise. Feet. 0.46 2.44 2.92 3.01 3.02 3.03 The rising curve plotted from these data was of little use in deter- mining the specific capacity of the wells, both on account of an unknown amount of water returned to the well by leakage of the pump, and because of the unknown amount of lowering of the water in the west well. TEST OF PUMPING PLANT OF NATHAN FULMER, LAKIN, KANS. This plant is in the center of NE. i sec. 10, R. 3(3 W., T. 25 S., Kearney County, 3 miles south of Lakin, Kans. The well consists of a w^ooden casing, 6 feet in diameter and 10 feet deep, sunk with the top flush with the surface of the ground. Inside of this cylindrical casing and extending 9^ feet below the bottom of it is a tapered wooden curbing 10 feet long, -i feet in diameter at the top, and 5 feet in diameter at the bottom. This curb was given the tapering form in order to lessen the friction on the sides in sinking the well. The total depth of the two large curbs is 19i feet. Arranged in a circle in the bottom of the main well, about 5 inches from the edge, are 7 feeders. Four of these feeders are 7 inches and 3 are 8 inches in diameter. The length of each feeder is 23 feet 4 inches. The feeders extend down to within 3 or 4 inches of an underlying clay or silt and 8 inches above the bottom of the large well. The total depth of the well is 42 feet. The feeders are made of No. 20 galvanized sheet iron with three-eighths- inch perforations arranged in circles from three- fourths of an inch to 2 inches apart. 68 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. The material encountered in sinking- the well, according to Mr. Fulmer, was, first, 4 feet of cla}^, then sand, which became coarser with the depth. The bottom stratum consists of a mixture of fine sand and gravel, some of the latter being the size of a hen's egg. Water was found at a depth of 8 feet. A local make of chain and bucket pump, known as the Pittman pump, is used in this well. It consists of an upper shaft and sub- merged lower shaft around which run the two sprocket chains to which are attached the galvanized iron buckets, each with a capacity of 12.6 gallons. The buckets, 33 in number, are hung between the chains and are of such shape that when they come over at the top of the circuit they discharge the water readily into the discharge trough, allowing very little to run back into the well. To aid in starting the water down the trough a number of horizontal guide vanes are placed therein with a slope away from the descending buckets in such a way that the water is started down the trough with very little splashing back into the well. These pumps are of a recent design, and are made in Kearney County. There are three such pumps in operation, one run by a windmill near Garden, and one owned by Mr. Root, a test of which is described in this report (pp. 70-73). The power is supplied through the proper gearing by a Howe gasoline engine, built by the Middletown Machine Company, which develops about 7 horsepower at 285 revolutions per minute. The engine is cooled by water taken from the discharge trough. The supply of gasoline is put in a rectangular sheet-iron tank, 2.4 feet by 2.6 feet by 1 foot high, which is placed in the ground outside the engine house. The ratio of the gearing between the engine and bucket chain is such that 175i revolutions of the engine produce 1 revolution of the bucket chain, or 5^ revolutions of the engine to each bucket discharge. The discharge trough empties into a reservoir from which the seep- age is quite rapid. As there was, no chance to put a weir between the pump and the reservoir, and since one placed at the outfall of the reservoir would measure only a portion of the water entering the reservoir, the amount of water pumped was measured b}'^ counting the number of revolutions of the bucket chain and computing the capacity of several buckets to secure an average \alue. The average capacity was found to be 12.52 gallons. The computed discharge, obtained by counting the revolutions of the bucket chain and noting the time, was 561 gallons per minute. It was estimated that the buckets lacked about 0.05 foot of being full, this being about 4 per cient of the measured capacit}" of the buckets. Also, during the run, 22 buckets came up empty, caused by the failure of the valve in the bottom to work, which amounts to a loss of one -fourth of 1 per cent of the total discharge. Reducing the observed 561 gallons by 4 per DETAILS Ol'^ TKSTS OF PUMPING PLANTS. 69 cent ^ivos 540 oalloiis \)vy inimitc ;is \\\v corrcclcd (liscliiirjre of tlio well. 1'li(> wiitei" Knol was lowered ^\.ol^ feet below (he normal. The lift to tlH> (lischaro'c troii<>-h was 17 feet. The engine ran at 240 revo- lutions and averaged 64 explosions per minute. The amount of o-asoline used was determined l)y measuring" the depth of t^asoline in the tank at intin'vals and noting* the time at each measurement; then by plotting a curve the average rate per hour of lowering of the gasoline in tlu> taidc was obtained, and, th(^ horizontal cross section of the tank being known, the amount of gasoline used per hour was computed to be 0.(),5 gallon. The cost of gasoline was iJl cents per gallon in bai'rel lots, making the expense of running the reef 634- /vorma/ 6 Change t o C'dienie 'er S 4- / f Change to 4 'cfiamet ?/■ / 2 / / / 1 i -4 6 t 1 ■) 1 2 / » / 6 / a 20 Time m minutes Fig. 20. — Rising furve for tlie Fulmer well, Luliiii, Kans. engine 13.65 cents per hour. The cost of water per acre-foot was therefore $1.37. The cost of water per 1,000 gallons was 0.42 cent, and the cost per 1,000 foot-gallons was one-fortieth of a cent. A reservoir 100 feet wide by 240 feet long is used in connection with the plant. This reservoir was made b}- digging out the inside and using the material to form the banks. This produced a very porous bottom and much trouble has been experienced from seepage. To remedy this the bottom was puddled thoroughly by plowing and har- rowing, then putting in chatf and straw and herding cattle and horses in the bottom for several days, but the surface of the water still drops about 6 inches per day. One observation of the rising curve of this well was made (fig. 20). It will be noticed that there is an irregularity in the curve correspond- ing to a depth of about li feet below the normal water level, caused 70 TJKDERrLOW IK AEKAKSAS VALLEY, WESTERK KANSAS. by the sudden change in cross section from 12^ square feet to 28i square feet at the top of the lower casing. From the rising curve the specific capacity may be obtained from the following formula: c=lT.25-log5 t ^ h At a point 4 feet above lowest position of water level the average area A of the well from to this point is 17 square feet. t — l.Q min- utes; H = 6.35 feet; A=2.35 feet. Then c = 80 gallons per minute. This, multiplied by 6.35, the amount the water was lowered b}^ pumping, gives 508 gallons, which is within 6 per cent of the observed discharge. The total area of percolating surface, 7 feeders, and the bottom of the well, is 334 square feet. The above specific capacity divided by 834 gives 0.24 gallon per minute per square foot of percolating area. The amount of water recovered can not be increased without lower- ing the pump, as a glance at the diagram will show, the water level being now lowered slightly below the lower shaft. The Fulmer plant was installed in the spring of 1903 and has been in operation since April of that year. The cost of the entire plant is as follows: Cost of Fulmer plant, Lakin, Kans. Cost of material. Labor. Total cost. Time, in days. Cost. Well: 818. 50 a3 a 45 $6.00 90.00 $24. 50 Digging . . . 90.00 Seven feeders at $8.40 (24 feet each, at 35 cents a foot) Reservoir, man and team, at $3.50 a day 58.80 58.80 40 140. 00 140. 00 Pump, made bv Mr. Fulmer, market price about 260. 00 328. 50 18.62 35.00 260. 00 Engine: I 347. 12 Freight J a5 10.00 Shed, 8 by 22 by 7 feet 45.00 34.58 Total cost . . . 719. 42 246. 00 1, 000. 00 a Labor, $2 a day. Mr. Fulmer uses water from the south-side ditch, and only about 15 acres of cantaloupes and fruit trees are irrigated. The capacity of the plant is about 100 acres. TEST OF PUMPING PLANT OF J. M. ROOT, LAKIN, KANS. This plant is located at the southeast corner of northwest i sec. 4, R. 36 W., T. 25 S., Kearney County, 3 miles southwest of Lakin, Kans. The well consists of a wooden casing, 6 feet in diameter and 12 feet long, sunk with the top flush with the ground. Inside of and below DF-ypAiLS or TKsrs ok pumpincj plants. 71 this is !i lo-t'oot cusiny-, 4o tVct in (liiunctiT at the top iuid 5,1 I'oet at thv hottoin. sunk imtii tlio lop is 2 feet above tlic bottom of the- upper casiny,-, luakiiii;- the total (l<'])th of the iiiain woll 20 feet, hi the bot- tom of this main well arc sunk ."> feeders in a circle about 10 inches from the odo'c oi the lower casinii. The feeders are iS inches in diam- eter; two of them an' 24 feet lono- and thr(>e are IS feet lono;. The 2-1-foot feeders project 2 feet above the bottom, while the 18-foot feeders jiroject only 1 foot. These feeders are made of No. 20 gal- vanized iron, and the perforations are the same as in Fulmer\s well, previously desci'il)ed. The material encountered in sinking- the well was, first, about 1 foot of sand, then about IT feet of lilack tlirt, followed l)y 1 foot of 3'ellow clay and '2 feet of sandy cla}-. There is no record of the material encountered in sinkino- the feeders. The Pittman pump is used in this w^ell and is of the same pattern as that described in connection with the Fulmer plant. The buckets are smaller, having- a capacity of 0.3 g-allons, and the bucket chain has places for 40 l)uckets, 24 of which were in place at the time of the test. The vacant places were left at regular intervals around the chain, Init the effect was to give the chain a swinging- motion, which caused the slopping out of a great deal of water. The valves in the l)ottoms of the buckets also leaked excessively. Power is furnished by a vertical 2i-horsepower tw^o-cycle Weber gasoline engine witli throttle governor, built by the Weber Gas and Gasoline Engine Company, Kansas City, Mo. The engine is cooled ])y a small tank and exi)loded by an autosparker. The ratio of the gearing betv/een the engine and the bucket chain is such that 257 revolutions of the drive wheel produce 1 revolution of the bucket chain, or ♦5.4 revolutions of the engine to each bucket raised, if the buckets are all on the chain. There is no reservoir used with this plant. The discharge was measured with a fully contracted weir, with a length of crest of 1 foot. The average head observed was 0.2805 foot, giving the follow- ing discharge by the Francis fornuila: ,j = 3.33 (/^-0.2 H) H3 = 3.33 (1.0 -0.056) 0.2805* = 3.33 X 0.944 X 0.1485 = 0.4675 second-foot • = 210 gallons per minute. By the formula given by Merriman for fully contracted weir of length of crest of 1 foot the discharge is computed to be 218 gallons per minute. The following computations are based on a discharge of 215 gallons per minute: As the water level was lowered 4.16 feet the specific capacity is 51.7 gallons per minute. The lift was 15.8 feet. The 72 UNDEEFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. engine averaged 488 revolutions per minute, exploding- at everja-evo- lution. The amount of gasoline used for a three-hour run was exactly 6 quarts, or at the rate of 0.5 gallon per hour. This gasoline cost 22 cents per gallon, making the cost of fuel 11 cents per hour. The cost of water is 0.855 cent per 1,000 gallons, $2.78 per acre-foot, and one- nineteenth cent per 1,000 foot-gallons. The lack of economy in this plant is in the engine, which is old and in poor condition, and in the buckets, the valves of which leak badly. Also the water was low- ered so far that the buckets did not start up full, and the swinging motion of the chain spilled a great deal. The owner has never been able to keep the plant running for more than half an hour at a time, and it took as long to put the plant in order as it did to make the test. /^orm^f It vel B_^S-30 .\N'I'S. 73 ()I)s('I'\(mI (liscliiirj^c. 'IMu' price )latin!4' surface area of foodcM's plus hottoni of woll — is 210 square feet, and divi(liii*>' the specific capacity (leteniiined fi-om the disciiarge i)y 210 we j^'et 0.24(> gallon per minute as (lie specific capacity per sipiare foot of percolating area. This large error is prol)al)ly caused by the ste(>p slope given to the rising curve t)y the leakage of the water from the biuckets. The pump nuist Ite lo\\('r(Hl Itefoi'e a gi'eater (luantity of water can be recovered, as the water at present is lowered to the level of the lower shaft. This plant has not been utilized for irrigation as yet, but its usc^ is contemplated for irrigating about 20 acres of beets, cantaloupes, melons, and garden truck. The Root plant was installed in the spring of iDO-i, being completed in the latter part of May. Its total cost was as follows. Cost of Root pampuKj plant near Lakin., Kans. Cost of material. Labor. Tirae in days. Total cost. Well: Lumber Feeders Labor — Prospecting for location, digging big hole Making big curb Sinking big curb Sinking feeders Pump Engine Installing Shed: Lumber, nails, and window Paint and painting Labor $27 42 Total . 100 100 4 18 24 34 100 100 6 33 4 10 "Labor, 92 a day. TEST OF WELL AT KING BROTHERS' RANCH, GARDEN, KANS. This well is located near the west side of sec. 30, R. 83 \V., T. 22 S., about 12 miles northwest of Garden, Kans. The well consists of a shaft, about 5 feet square, sunk 41.4 feet, to within 1.2 feet of the water level. From the bottom of this shaft a 15-inch, perforated, galvanized-iron casing extends down to a depth of 40.5 feet from the normal surface of the ground water. It was put down by King Brothers to determine the amount of ground water which could be recovered at this point from a single well and its influence on other wells. 74 UNDERFLOW IN" ARKANSAS VALLEY^ WESTERN KANSAS. Fifteen feet from the first well a second well was sunk to a depth of 91 feet, 7.9 feet lower than the first well. This second well was put down for the purpose of determining the efi'ect on the water plane of lowering the water in the first well by pumping. A No. 4 Byron-Jackson centrifugal pump was placed at the bottom of the shaft of the first well and connected b}^ a long belt running over 2 idle wheels to a 14-horsepower thresher engine on the surface of the ground. The discharge was measured b}^ a fully contracted weir with a crest of 1 foot. The head at the time of maximum discharge, when the water in the well was as far down as the pump could lower it, was 0.25 foot, corresponding to a flow of 183 gallons per minute. This maximum rate was very difficult to maintain for any length of time, because of the temporary manner in which the machinery was installed. The belt was liable to slip and allow the water to rise sev- eral feet; also the idle wheels at the top of the shaft over which the belt ran were poorly mounted, and at times a stop was necessar}^ to cool off a hot box at that place. The above discharge was measured when the water level in the well was lowered 20.3 feet by the pump. Dividing the discharge b}^ the distance gives the specific capacity of the well, or the amount of water furnished for 1 foot of lowering, as 9 gallons per minute. The total percolating area of well strainer exposed to the water was 85 square feet. From this it appears that the specific capacity of the well strainer is 0.106 gallon per square foot per minute. As this was a test of the capacity of the well only, and not of the pumping plant, no indicator cards nor other device was used to get the efficiency of the plant, and no measure was made of the coal burned. The mechanical efficiency would undoubtedl}^ have been low, as there was a constant slipping of the belt, and the idle wheels were home- made, running in wooden bearings, which were smoking constantly. The maximum lowering of the water in the main well was 20.2 feet, and the corresponding depression of the water plane, 15 feet away, as indicated by the test well, was 3.5 feet. This shows the steep slope of the water plane and the comparatively small radius of the base of the cone of influence. Readings were taken of the water level in the main well and the test well, and the discharge was noted at intervals. The accompanying curve, fig. 22, shows rising curves for the main well and the test well plotted together. A study of the curve brings out several facts that might well be expected. The rise of the test well lags slig'htly behind that of the main well. The curve of the main well shows an irregu- larity due to the caving in of material around the strainer. King Brothers contemplate sinking 20 of these wells in a north and south line. They propose to connect them all with a tunnel just above the water plane and lay a main suction pipe in this tunnel, with nP.TAlLS OK TESTS OF PUMPING PLANTS. 75. branohos tap])iiii:' all the wolls. The iMuiips will h(> locatod in the shaft "already dug, aiul i-onnectecl by a belt to the power plant on the surface. The owners paid -iO eents a foot for sinking the wells and furnished one man. The price paid for the 15-inch, No. i(), iron casinj^* was ^l a foot. They contemplate usinii' wooden casino- iji the remainder of 21 _- — r0 .^iSJl^ ^""^^ ___^ -^ .^r^ 11^^— / / / 1 / / / / / { \ S /O IS , 20 ZS 20 Time in minufes Fig. 22. — Rising curves for main woU and test well, King Brothers' plant, Garden, Kans. the wells. This will be made of pine lumber 1 inch by 3 inches. It will take 16 such boards to make the circular casing, at a cost of $3.50 per hundred linear feet of lumber. One man, at $1.50, can make and perforate about 25 feet of this casing in a day. This would make the cost of wooden casing 62 cents per foot. 76 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. After the tunnels and wells are dug King Brothers purpose to con- tract for the installation of a compound Corliss engine and centrifugal pump at about $9,500. They expect the plant to raise 4,000 gallons of water per minute, with a 60-foot lift, this being at the rate of 2,000,000 foot-pounds per minute on 4,800 pounds of coal per twenty- four hours. If the coal contain 12,500 British thermal units, and if the boiler efficiency be assumed at 75 per cent, engine 13 per cent, and belt 90 per cent, the pump would be required to have an efficienc}" of 70 per cent to realize the above expectation. These figures require that the plant turn out 5.9 per cent of the energy in the fuel in the form of useful work. _ TEST OF CITY WATERWORKS WELL, GARDEN, KANS. The first test began at 4.25 a. m. June 28, 1904, when one pump was started. The second pump was started at 5.50 a. m. The hydrants used in flushing the sewers were opened at about 7.30 a. m. and closed at 10.35 a. m. The east pump was stopped at 11.15 a. m.; the west pump was operated constantly all day. On account of the flushing of the sewers an exceptionall}^ large amount of water was pumped during this test. ' Gage heights in the well were read every five minutes, and the num- ber of cycles of each pump was recorded every tenth minute for the ten preceding minutes. The pumping machinery consists of two compound steam duplex pumps, with cylinders 8 inches by 12 inches, which are very old and worn. The test was continued until 12.40 p. m. At 8 p. m, the test was again taken up, this being the time when the sprinkling of lawns is stopped. The cycles of the engine were counted and well heights taken as before. Pumping is stopped at 9 p. m. Sprinkling of lawns is allowed from 7 to 11 a. m. and from 4 to 8 p. m. Most of the rise of water in the well occurs before 9 p. m., when the pump is stopped. A plug was made for the feeder and inserted July 7, but it did not fit tight enough to stop the flow. The rising curve was taken July 7 in the evening and also July 8, when the plug was driven down so as to be water- tight. July 11 the rising curve was again taken when the water was lower. The well is 16.2 feet inside diameter and 20 feet deep. The bottom is about 8.9 feet below the normal level of the ground water. There is a 10-inch feeder in the bottom of the well, which extends to a depth of 42 feet below the ground and about 3 feet above the bottom of the large well. It is open at the bottom and perforated 10 inches up from the bottom. The water level is about 11 feet below the ground level. |)i:i\\ll.S n\' IKSI'S (II' I'UMIMNC IM-ANTS. 77 l>(thi of rill/ inilirirnrhs ii;ll, (luriliii, Kum^. Feet. Klevatiun nf top ct well r.x.f 2,837.26 Distance to tup of Distuncf, top to 0. 10. .35 13. 25 23. 60 I'll cvat ion, water iioniial : 8. 72 Normal elevation of water 2, 822. .38 round level. .. 2,832.( l-:ievatioii of hottolii of wcj L>,S 13.66 = of <:ay;e. reei 0.8 Norma/ /e ve/ = 8. 7f .7 .6 ■S JoJX^-- ^ -"^ ^ ^ -^^^ 3 \H^' 10- h^y^ Ji i^^ .1 ^ ^ y / eo 9 1 ^ ' y .6 .7 .6 V \'K y S 7.3 / / Time of day P.M. Fig. 23. — Rising curves for city waterworlis well, Garden, Kans. The rising- curve.s obtained for this well on June 28, Jul}' 7, 8, and 11 are reproduced in fig. 23. Fig. 24: gives the engine cycles and ele- vation of water in the well for several hours of heavy pumping on June 28, 1904, while the sewers were being flushed. The displace- ment in the two cylinders of one of the pumps amounts to 1.362 cubic feet. The curve in fig. 24 enumerates the cycles of pumps, so that the total discharge of the pumps can be obtained, if no allowance be 78 UNDERFLOW IN ARKANSAS VALLEY^ WESTERN KANSAS. made for slip, by multiplying the number of cycles b}^ 1.362. The discharge, computed in this way, amounts to 685 gallons a minute. The amount of slip is enormous. Using the rising curve for June 28, we may place H = 0.65, A = 0.46, and A = 204. 5 square feet in the formula for specific capacity. This gives a specific capacity c=53 gallons a minute. This result is much below the normal on account of the excessive amount of pumping on that day, due to the flushing of sewers. The maximum amount of lowering of the water in the well was 5.48 feet, V \ r\ \ / u^ \ / ^ \ \ ^ -^ \ 7 rJ ^ X •-( V yV'^fer /e\^ 1 1 kl 20 ^ \ I / 10 o Fig. 24.— Elevation of water in city well, Garden, Kans., and engine cycles of steam pump during heavy pumping while flushing sewers, June 28, 1904. which occurred at 10.30 a. m., when the pump cycles numbered 67 per minute. Multiplying 5.48 by 53, the total discharge at that time is found to be 290 gallons a minute. The slip of the pump must there- fore have amounted to 57 per cent at this time. Using the rising curve of July 7, the specific capacity of the well is found to be 67 gallons a minute. The following table shows the specific capacities computed for the several dates: Specific capacity of city luaterworks well, Garden, Kans., 1904- Rising curve. Specific ca- pacity per minute. June 28, 8.55 to 9.05 p. m July 7, 9 to 9.10 p. m July 8, 9 to 9.10 p. m July 11, 9.06 to 9.15 p. m DETAILS OV TKSTS OF I'UMlMNd PLANTS. 79 These results furnish intrrestino- coniparisou.s. I'he low specitic capacity on flunc 'iS was obtained after the prolonged and excessive puinpino- for tiushinn- of the sewers. There was a li<^ht I'ain on the niojit of .July 0, and a heavy I'ain on the i\ioht of Jul}' 7, which influ- enced both the consumption of watei- in the city, and in a slig'ht degree the amount of gi'ound wat(>r a\ailal>le. On -luly IS and 11 the feeder in the bottom of the well was plugged. The plug did not leak, l)ut the casing of the feeder must have leaked ])adly since no intlu(Mu-e upon the specific capacity of the well can be detected. In Water-Supply Paper No. 67," a rising curve for this same well is given, as observed b}- Johnson in 11)00. From that curve it is possi- ble to compute the specific capacity of the well in 19(»0. The follow- ing determinations are based upon various intervals after pumping has stopped, as indicated in the table. The specific capacity of a well always appears to be lower than its true value, if the very last portion of the rising curve be used, since at this period a large fraction of the water is being utilized in tilling up the ground around the well. Specific capacity of city waterworks well, Garden, Kans. , 1900. Intervals, Specific ca- pacity per minute. 0-10 minutes Gallons. 73.0 0-20 minutes 71.0 0-30 minutes €6.4 0-50 minutes 63 20-40 minutes 59.0 40-50 minutes 55.0 These results seem to be identical with those obtained in 1904. The average specific capacity (77 gallons a minute) as determined in 1904 indicates that the maximum yield of the well, if the water in the well be lowered 8 feet, is 615 gallons a minute. The pumps in use at present can not pump much more than half of this amount of Avater on account of the worn condition of pistons and cylinders. The total percolating surface of the well bottom and strainer of the feeder is 217 square feet. From this it can be deduced that the spe- cific capacity of the well is 0.31 gallon a minute per square foot of percolating surface. aSlichter, C. S., The motions of underground waters: Water-Sup. and Irr. Paper No. 67, U. S. Geol. Survey, 1902, p. 68. 80 UNDERFLOW IN ARKANSAS VALLEY, WESTERN KANSAS. TEST OF HOLCOMB'S PUMPING PLANT. A very important attempt to recover water from the underflow was begun in the 1904 season by the owners of the Riverside stock ranch, about 7 miles west of Garden, Kans. A well 200 feet long and 5 feet wide, excavated to a depth of 9 to 10 feet below the water plane was constructed of sheet piling, and 11 galvanized-iron feeders were inserted in the bottom of the wells to a depth of about 20 feet. A T5-horse- power Corliss engine with condenser, a 90- horsepower boiler, and a No. 15 Bj^ron-Jackson centrifugal pump were put in position at the north end of the well. Foundations for the engine and pump and buildings to cover the machinery were constructed in a very substan- tial manner. As soon as the engine and pump are in satisfactory working order it is purposed to sink a large immber of additional feeders in the bottom of the well in the expectation of increasing its capacity to 6,000 gallons per minute. The approximate cost of the plant is about $8,000 for machinery and $4,000 for the well. Trinidad slack coal is used for fuel at a cost of $4 to |4.50 per ton. The construction of this pumping plant has attracted very wide attention and if it proves to be a success it will mean a great deal for the progress of irrigation in the bottom" lands of Arkansas Valley. There is some question whether 6,000 gallons per minute can be obtained from the present well, even with a very large number of addi- tional feeders, but it will be entirely practicable to increase the length of the well without ver}^ much additional expense. The present well, with ten 20-foot feeders, 16 inches in diameter, would furnish about 6,000 gallons per minute, if we can rely upon a specific capacity of one-third gallon per minute for each square foot of strainer. This would require, however, the lowering of the natural level of the water to a distance of 10 feet, which is somewhat more than would be best for the most economical running of the plant. '^ Both suction and discharge pipe of the centrifugal pump are made of No. 16 galvanized iron, riveted and soldered. A 20-inch flap valve is placed at the upper end of the discharge pipe, dispensing with the use of a foot valve. The pump is primed before starting by opening a 1-inch valve in a lead pipe from the main pump to the air pump. When the proper vacuum is shown by the gage the 1-inch valve is closed and the engine started. A test run of the plant was made for five da3^s, from Jul}^ 18 to 23, 1905. The engine was started at the lowest speed at which it would work the pump satisfactorily. After running at this rate for fort}^- eight hours the speed was increased until nearly the full capacity of the well had been reached. The amount of water pumped during the test averaged about 2,300 gallons a minute, or 5 cubic feet of water a second. This is equivalent a Actual test of the plantshows that this amount of water can not be recovered without extending the well. DKI'AILS Ol' ll'isrs OF I'l'MI'lNC IMvANTS. 81 to ;i to supply tlu> i)unip and (Mioinc with all of the wat(M' that tlicy avo dcsio-ncd to handle; in fact, tlu^ i)unii) and engine are (•apal)le of haiuilini;- three times the amount of water at present availal)l(> for lono--eontinued runs. It is exp(>et well, and by eidaru-ino- the well, the capaeity of the i)lant will he greatly increased; but even at the present low rate of delivery, and consequent rather low^ efficiency of the machinery, the cost of watei- delivered is comparatively low. The a\erao-e amount of coal consumed was 2,450 pounds per twenty- four hours, or al)out 1} tons per day. At $-4 a ton the dail}' cost of coal was '^.^ per twenty -four hours. The cost of labor for the day and night man, each at iBl.25 per day, makes the cost for coal and labor $7.50 per twenty-four hours. The cost of lubricating oil and miscel- laneous supplies may be estimated at $1 a dav, luaking a total cost of §8.50 per twenty-four hours. At this rate the cost of water was 85 cents per acre-foot, not including interest on the plant nor any allow- ance for depreciation and repairs on the machinery and well. If these latter items be included, the cost of w^ater would be very materially increased. It seems, however, unfair to estimate these charges at the present time, as the expense of erecting the plant was incurred on the basis of securing a very consideral)ly larger amount of water than is at present delivered; for that reason the interest charges would be very high, if charged tigjiinst the present amount. It seems very probable that if the supply of water from the well is sufficientlv increased the plant will ultimately be capable of delivering water into the ditch at a cost not to exceed ii>l per acre-foot, including a moderate charge for interest and depreciation on machiner}', but not including any profit. The following tables show the fuel conssumed and the data obtained during the test. The well was not of sufficient size to supply the pump with water, and toward the end of the run difficult}' was expe- rienced in operating the plant. Occasionally the water became so low that air Avould be taken into the suction pipe, and the plant would have to be stopped to prime the pump. In order to secure proper returns, it will be necessary to enlarge the well to about three times its present capacity, otherwise the engine and pump will be entirely too large for the well. iRR 153—06 6 82 UNDERFLOW IN ARKANSAS VALLEY^ WESTERN KANSAS. Consumption of coal at test of Holcomh pumping plant. Pounds. July 20, 6 a. m. to 6.30 p. m 1, 386 July 20, 6.30 p. m. to July 21, 6.20 a. m 1, 386 July 21, 6.20 a. in. to 6 p. m 1, 134 July 21, 6 p. m. to July 22, 6 a. m. « 1, 260 July 22, 6 a. m. to 6.30 p. m 1, 134 July 22, 6.30 p. m. to July 23, 6.30 a. m. & 1, 134 July 23, 6.30 a. m. to 2.40 p. ni 832 8,266 Data of test of Holcomh jiumping plant. [From 7.54 a. m., July 18, to 2.40 p. m., July 23.] Date. Hour. Discharge of flume. Cubic feet per second. Gallons per min- ute. Depth of water below its initial position. Speed in revolu- tions per minute. Engine. Pump July 18.. July 18.. July 18.. July 18.. July 18.. July 18.. July 18.. July 18.. July 19.. July 19.. July 20.. July 20.. July 21.. July 21.. July 21.. July 22.. July 23.. 7.54a. m. 8 a. m... 8.45 a. m. 9.30 a.m. 11.20 a. m 2.25 p.m. 4.25 p.m. 5 p. m... 7.15 a.m. 9.30 a.m. 7a. m ... 7.45 a. m. 6 a. m ... 8 a. m ... 8.20 a.m. 7 a. m... do... 13.33 7.42 7.38 6.29 5.87 5.32 4.89 4.67 4.72 5.30 4.67 5.82 5.68 5.30 5.03 5.15 5,980 3,330 3,310 2,820 2,630 2,380 2,190 2,090 2,110 2,380 2, 090 2,610 2,540 2,380 2,250 2,310 Feet. 4.40 7.25 7.27 7.44 7.61 7.61 7.62 7.80 7.74 7.87 8.40 8.50 9.50 9.53 9.30 9.60 350 344 345 344 330 344 344 344 344 346 352 345 360 360 357 361 TEST OF PRODUCER-GAS PUMPING PLANT NEAR ROCKY FORD, COLO. The future of irrigation in the bottom lands of Arkansas Valley will be greatly influenced by the cost of power for pumping water. One of the possible ways of cutting down this cost is by the use of producer-gas in gas engines, as mentioned in Chapter V. A 35-horsepower producer-gas plant has been installed by Mr. A. W. Shelton, about 6 miles northeast of Rocky Ford, C^olo. It consists of a 40-horsepower Fintsch suction gas producer, a 35-horsepower single- cylinder gas engine, and a 12-inch Byron-Jackson vertical-shaft cen- trifugal pump. The water is pumped from a canal through 15-inch concrete tile (200 feet of intake and 300 feet of discharge) to an eleva- a stopped 35 minutes, & Stopped 36 minutes, DETAILS OF TESTS OF PUMPINO PLAN'I'S. 83 tion of 1(? foot ahovo tlio lovol of (ho wator in thocatial, callodtho tirst disoharg'o, and at aiiothor point to an olo\ati()n of 28 foet a})()vc the lovol of tho Avator in tlio canal, callod tho second discliarnc. A tost was niado of this plant, oxtondin*^' from Doconihor 'I to 5. Tho results of the test of tho engine and producer-gas appara- tus are given lierowith. Unfortunately the cement-discharge pipe gave out when tho plant was tirst started, so that water could not be pumped during the test, and the hvdraulic data for this plant are therefore not availal)lo. In connection with the generation of tho gas a vaporizer, scrubber, and purifier are used. Water is ovapt)rated at atmospheric pressure to generate the steam required in the producer. The vaporizer is located directly on top of tho producer. The scrubber is of the form in which a spray of wator trickles down through coke, the water run- ning out at the bottom of tho scrubber. After being scrubbed the gas passes through a purifier box, next through a gas governor, and then to the engine. The engine used was one of the Olds gasoline type, somewhat modi- tied for the use of producer-gas. The engine governor was not the one belonging to the engine. A belt was connected from a 30-inch pulle}- on the engine to a pulley on the vertical shaft of the centrifugal pump. A clutch at the engine shaft allowed the pump to be disconnected at will. A small pullo}', fastened to the shaft opposite the pulley end, carried a belt which drove a 3 by 5 inch "Baker'" feed- water pump. This pump, making about 45 revolutions a minute, drew Avater from the well and discharged it into a 3 b}^ 8 feet b}^ 30 inches storage tank near the roof of the building and above the producer. A pipe leading from the bottom of this tank furnished all the water used to operate the plant, viz, water for the engine jacket, steam, and scrubber. During the brake tests it also supplied cooling water for the brake. The w^ater, after being used, passed through the seals and then into the well from which it was drawn. Freliminary brake tests uf gas engine at pumping plant of A. W. Shelton, near Rockg Ford, Colo. , December 4, 1905. 1 Maxi- Maxi- Mean ef- mum mum O A • fective Mechan- pressure pressure Net revolu- tions per Explo- pressure, Brake Indicat- ical effi- of explo- of com- Test. load, in sions per in pounds horse- ed horse- ciency sion, in pression, pounds minute. per power. power. (per pounds in pounds square cent). per per inch. square inch. square inch. 1 151 169 199 200 99.5 100.0 44.0 45.6 26.7 30.0 34.0 35.4 78.5 84.9 226 235 140 2 133 3 174 199 99.5 50.0 80.7 38.7 79.4 235 148 84 UNDERFLOW IN ARKANSAS VALLEY^ WESTERN KANSAS. Diameter of piston 14 inches. Length of stroke 20 inches. Length of brake arm 56 inches. .„ , 2X7fX56 Brake constant= i^vSSOOO ~ 000888 .^ . 20X7rXl4Xl4 Engine constant = i2y4yS3000 ~ ' 007/8 Brake horsepower= Net load XSpeedX . 000888 Indicated horsepower = Mean effective pressure X Explosions X . 00778 ,^ , . , ^ . Brake horsepower Mechanical efficiency ^indicated horsepower" Indicator spring, pounds per square inch, =160 for Test No. 1, 250 for Test No. 2, and 250 for Test No. 3. Of the gas producer a test of three hours' duration was made. The gas governor was not in operation. The producer was filled with coal at the beginning, as well as at the end of the run. As the engine was not operating well the load had to be taken ofi^ for a time. Test of gas producer at pumping plant of A. W. Shelton, near Rocky Ford, Colo., December 4, 1905. Time. Net brake load. Revolu- tions per min- ute. Explo- sions per minute. Temperature °F. Coal. Ja Enter- ing. 3ket wat Leav- ing. er. Range. Engine room. Out- side air. P.m. 1.35 1.50 2.05 2.20 2.35 2.50 3.05 3.20 3.35 3.50 4.05 4.20 4.35 Av Total . . Pounds. 181 181 181 181 181 181 Load partly .off. 181 181 181 204 208 102 104 45 46 154 160 166 165 165 208 208 109 115 120 120' 163 163 60 60 60 64 64 . 64 50 50 49 48 48 47 46 . 46 46 44 43 42 41 Pounds. 21 24 17 18 21 206 206 206 204 202 103 103 103 102 101 45 45 45 45 45 213 220 117 110 45 45 45 125 207 105 45 175 132 6*2 46 "ioi"' Summary of test of gas producer at pumping plant of A. W. Shelton, near Rocky Ford, Colo., December 4, 1905. Duration of test 3 hours. Net brake load (maximum) 181 pounds. Net brake load ( average) 125 pounds. Revolutions per minute (average) 207. Explosions per minute (average) 105. Temperature of water entering jacket (average) 45° F. Temperature of water leaving jacket (average) 175° F. Bange of jacket-water temperature 132° F. DKTAILS OF TKSTS OF l'ri\llMN(; PLANTS. 85 Temperature of eiigiiu' mom (aveiajr*') 62° ¥. Tfiii[H>rature of dutsidi' air (avcramO 40° F. Total foal consnnu'd 104 imhiikIs. I'rof^sure niaxiimiiu explosion 258 pouiidH pi-r H(|uare inch. Pressure nuiximuin i-ompression 145 pounds per s(|uare inch. Pressure, suction at exit of producer 2 indies water. Pressure, suction at exit of scrubber 2.125 inches water. Pressiu'e, suction at exit of purifier 2.25 inches water. Pressure, mean effective , 51.6 pounds per square inch. Indicated horsepower (51. 6X105X. 00778)= 42.2. Brake horsepower ( maxinunn) = ( 181 X 207 X .000888) = . . .33.2. Brake horsepower (averai,'e) = ( 125X207 X.0008S8)= 22.9. 83.2 ^lechanical I'tliciency (maxinunn) j^\^ 78.9 per cent. 22.9 Mechanical efficiency (average) j^-n •"^'^■- Pf •" <"«nt. Pounds of coal per brake horsepower per hour, baseer cent. Kind of producer Pintsch. Producer rated horsepower 40. Temperature of water entering jacket 49.8° F. Temperature of water leaving jacket 165.8° F. Range of jacket- water temperature 1 16° F. Temperature of outside air 46.8° F. Temperature of engine room 72.0° F. Pressure of maxinuim explosion 260 pounds per square inch. Pressure of maximum compression 150 pounds per square inch. Pressure of maximum steam Atmospheric. Pressure of maximum suction at producer exit 2.2 inches water. Pressure of maximum suction at scrubber exit 2.2 inches water. Pressure of maximum suction at purifier exit 2.4 inches water. a The high mechanical efficiency i.s probably due to an error in the indicated horsepower. The mean effective pressure appears to be too low. The reducing motion used was made of wood and had become considerably worn when this test was made. 86 UKbEEFLOW IN AEKAiSTSAS VALLEY, WESTERN KANSAS. Data concerning coal used in test of producer-gas pumping plant of A. W. Shelf on, near Hocky Ford, Colo. Kind Colorado anthracite, Floresta mine. Cost at plant per ton $6. Size Pea. Total quantity fired 325 pounds. Total refuse (clinkers, ash, and unburned coal) . 66 pounds. Total clinkers 5 pounds. Total unburned coal ' 43 pounds. Total ash (siftings) 18 pounds. Calorific value of coal per pound _ 13,850 B. T. U. Pounds of coal per brake horsepower per hour, as fired and uncorrected for unburned coal in /" 325 \ , .o refuse [j^^^YaJ l-^^- Pounds of coal per brake horsepower per hour (corrected for unburned coal in refuse) 1. 24. Approximate analysis of coal used at producer-gas pumping plant of A. W. Shelton, near Rocky Ford, Colo. Per cent. Moisture 2. 2 Volatile matter _ 7.6 Fixed carbon 83. 8 Ash 6. 4 Water used per hour in producer-gas pumping plant of A. W. Shelton, near Rocky Ford, Colo. Pounds. By jacket 1, 200 By brake 930 By scrubber (approximately) 1, 300 By vaporizer 16 Efficiencies at various loads of producer gas pumping plant of A. W. Shelton, neetr Rocky Ford, Colo.; test of December 6, 1905. Time. Net brake load (lbs.). Revo- lutions per min- ute. Explo- sions per min- ute. Jacket water. Mean effective pressure (pounds per square inch). Indi- cated horse- power. Brake horse- power. Mechan- ical effi- ciency (per cent). Temperatures. Pounds per hour. Inlet. Out- let. Range. 9.50 10.18 10.42 11.06 11.30 12. 25 1.18 1.40 2.00 27 50 75 100 125 150 175 200 215 a 225 205 203 203 203 203 199 201 189 178 45.0 52.2 59.4 69.3 83.0 96.0 100.6 95.0 89.0 44 44 42 42 42 42 42 42 42 104 107 112 117 126 136 145 140 148 60 63 70 75 84 94 103 98 106 1,260 1,570 1,300 1,340 1,340 1,570 1,570 1,570 44.1 41.1 44.9 44.1 43.7 42.4 39.9 39.9 41.3 15.4 17.9 20.7 23.8 28.2 31.6 31.2 29.5 28.6 4.9 9.0 13.5 18.0 22.5 26.5 31.2 33.6 33.9 30.0 50.2 65.2 75. 6 79.8 83.8 a Engine would not carry load. I DETAILS OF TESTS OF PUMPING PLANTS. 87 AiHtlj/sis (if ijiii^ from (/((.s' producer ol puinp'niii phnil of .1. 11'. Sluilon, )ie(ir Rockfi Ford, Colo. Percent. B.T.r. per 100 cubic fcut oI'kiih at 60° F. COo CO 0114 H N f).3 1.0 0.0 IS.G 50. 9 7J)1H 0,107 100 13,fi20 It will 1)0 observed from the results obtained in the test that 1.24: pounds of coal per hour produced 1 brake horsepowei'. At $0 a ton the cost of fuel was therefore three-eighths of a cent per brake-horse- power hour. At this rate power was obtained at a cost for fuel equivalent to gasoline at 3 cents per gallon. One-half cent per brake- horsepower hour for labor and tive-eighths cent per brake-horse- power hour for supplies, depreciation, and repairs should cover all other charges. The total cost of power should not exceed, there- fore, li cents per brake-horsepower hour, or about $4.50 per da}^ of ten hours, for the present plant. In this length of time the plant should furnish about 8 acre-feet of water on the 16-foot lift, or at a cost of about 58 cents per acre-foot. The first cost of the pumping plant in round numbers, was $3,300 for the producer, engine, and pump; §2()(> for the building, and $1,500 for the intake and discharge pipe and tlumes. INDEX. Pago. Alkalinity, iiU'asurciiK'iit.s uf 45^7 Analyses of ground water ,. . . 45-47,49-50 Arkansas River, cross section of, figure showing 23 narrows of, measurements at 22-24 valley of, topography of 7 wells in, water of, quality of 49-50 water of, gain and loss of 31, 41-42 height of 28-13 figures showing 29, 32, 38, 43, 52 water of, temperature of 12 Barometric pressure, relations of water table and 32-34 relations of water tal)le and, figure showing 33 Bear Creek, character of 20 water from 21, 53-54 Catchment area, location of 5 Clear Lake, Kans,, character of 18-19 location of 18 map of vicinity of 19 underflow stations at, measurements at. 20-21 water supply from 18, 21 Colorado, coal from, use of, for fuel 6 origin of ground water in 51 Deerfleld, Kans., evaporation at 43^4 ground water at, fluctuations of. . . 31, 42^4, 54 rainfall at 44 river at, height of, figure showing . . .53 underflow at, analyses of 47 undei-flow stations at, location of, map showing 17 measurements at 16 Diesem, I. L., pumping plant of, data on. . . .5.5-65 Dodge, Kans., rainfall at 54 sand hills near 7 Evaporation, measurements of 43-44 Field work, character and extent of 5,7 Floods, influence of, on ground water ... 5, 11-12, 14, 28-34, 39-40 source of 53 Fulmer, Nathan, pumping plant of, cost of 70 pumping plant of, data on 55-56, 68 tests of 67-70 well of, character of 57, 67-68 rise of water in, figure showing 69 Garden, Kans., cross section near, figure .showing 11 gravel near, character of 10-1 1 groimd water at, fluctuations of 26-30, 54 rauifall at 30, 38-39, 54 figures showing 52, 53 I'age. Garden, Kans., rock under, depth to 51 run-oll at 6 underflow near 6 analyses of 45^6 solids in, figure showing 11 underflow stations near, location of, map showing 9 measurements at 7-13 waterworks of, pumping plant of, data on 55-56, 76 pumping plant of, tests of 76-79 well of, cliaracter of 76, 79 rise of water in, figures showing 77, 78 Gasoline, use of, for fuel ... 6, 55, 57, 63, 65-67, 69, 72 Gas-producer plant, test of 84-87 use of 6, 57-58 Gravels, character of 10-11, 13 depth of 51.56 Ground water. See Underflow. Hartland, Kans., underflow stations near, location of, map showing 22 underflow stations near, measurements at 20-21,24 Hedge, H. K., aid of 18 High Plains, character of 52 Holcomb, II. B., pumpingplant of, cost of.. . 80-81 pumping plant of, data on 55-56, 80 tests of 80-82 well of, character of 80 Johnson, W. D., on ponds on High Plains. . 20 Kansas oil, use of, for fuel 6, ,57 King Brothers, pumping plant of, cost of . . 76 pumping plant of. data on .5.5-56. 73 tests of 73-76 well of, character of 73-74 rise of water in, figure showing .... 75 Kipp, H. S,, pumping plant of, data on ... . 55-56 Lakin, Kans., pumping plant near 70 Logan, D. H., pumping plant of, data on '. 55-56, 59-60 pumping plant of. tests of 59-62 well of, character of .57, 61-62 rise of water in, figure showing .... 61 McKinney, J. R., pumping plant of, data on. .5.5-.56 Owen, Ray, work of 5 Producer gas. See Gas-producing engines. Pumping, cost of 57-58, 60, Si, 6.5-67, 69, 72, 87 Pumping plants, power for 6 tests of, details of .59-87 summary of 5.5-58 See also individual plants. Rainfall, amount of 2.5-26, 54 effect of 5, 28-.30, 34 89 90 INDEX. Richter, Mrs. M., pumping plant of, data on 55-56, 62, 65 pumping plant of, tests of 62-65 well of, character of 57-62 fluctuations in 26, 34 rise of water in, figure showing 64 Rocky Ford, Colo., producer-gas pumping plant near. See Shelton, A. W., pumping plant of. Root, J. N., pumping plant of, cost of 73 pumping plant of, data on 55-56, 70 tests of 70-73 well of, character of 57, 70-71 rise of water in, flgare showing 72 Run-off, absence of 6, 54 Sand hills, catchment area in 5, 16, 51, 54 location of 7 wells on, water of, quality of 49 Sexton, C. E., pumping plant of^ data on. 55,56,65 pumping plant o!, tests ot 65-67 well of, character oi 57, 65-67 Shelton, A. W., pumping plant of, data on 82-84 pumping plant o", gas from, analysis of 87 tests of 82-87 Sherlock, Kans., cross section near, figure showing 15 ground water at, fluctuations of. . 31,33, ,3.5-42 fluctuations of, figures showing 38, 40 rainfall at 38-39 underflow at, analyses of 46 underflow stations at, location of, map showing 14 measurements at 13 Page. Sherlock Bridge, wa.ter at, height of, figure showing 52 Silt, occurrence of 10 Smith, L. E., pumping plant of, data on . . . 5.5-56 Specific capacity of wells 56-57 Temperatures, relative, of ground and river water 12 Underflow, analyses of 45-47, 49-50 chemical composition of 45-50 variations hi 5, 12-13, 16 conclusions concerning 5-6 direction of 10, 11, 13, 1.5-18, 21, 24 extent of '5, 54 influence of floods on. . . 5, 11-12, 14, 28-34, 39-40 influence of rains on 5, 28-30, 34 level of, figure showing 29 fiuctuations of 25-44 solids in, amount of 5, 45-50 variation in 5,47-48 figure showing 47 source of 51-54 temperature of 12 velocity of 5, 10, 13, 16-17, 21, 24, .54 Water plane, map of, figure showing 8 slope of '. 5-7 figure showing 11 Wells, character of 56-57 specific capacity of 56-57 water of, quality of 49-50 yield of 6 See also individual pumping plants. White Woman Creek, flow of 53-54 Whitney electrolytic bridge, use of 45 use of, results of, figure showing 47 Wolff, H. C, work of 5,26 CLASSIFICATION OF THE PUBLICATIONS OF THE UNITED STATES GEOLOGICAL SURVEY. [\Vater-Sui)i)ly I'liper No. l.'iS.] The serial publications of the United States (Jeolo^ical Survey consist of (1) Annual Reports, (2) ]\Iono^raphs, (.'}) rrofessioiial Papers, (4) Bulletins, (5) Mineral Resources, (6) Water-Sup]ily and Irrigation Papers, (7) Topographic Atlas of United States — folios and separate sheets thereof, (8) (geologic Atlas of the United States — folios thereof. The classes numbered 2, 7, and 8 are sold at cost of publication; the others are distributed free. A circular giving complete lists may be had on application. Most of the above pul)lications ma}' be obtained or consulted in the following ways: 1. A limited number are delivered to the Director of the Survey, from whom they may be ol)tained, free of charge (except classes 2, 7, and 8), on application. 2. A certiiin numl)er are delivered to Senators and Representatives in Congress for distribution. 8. Other copies are deposited with the Superintendent of Documents, Washington, D. C, from whom they maj' be had at prices slightl}' above cost. 4. Copies of all Government publications are furnished to the principal public libraries in the large cities throughout the United States, where they may be con- sulted by those interested. The Professional Papers, Bulletins, and Water-Supply Papers treat of a variety of subjects, and the total number issued is large. They have therefore been classified into the following series: A, Economic geology; B, Descriptive geology; C, System- atic geology and paleontology; D, Petrography and mineralogy; E, Chemistry and physics; F, Geography; G, Miscellaneous; H, Forestry; I, Irrigation; J, Water stor- age; K, Pumping water; L, Quality of water; M, General hydrographic investiga- tions; N, Water power; 0, Underground waters; P, Hydrographic progress reports. This paper is the twelfth in Series K and the fiftieth in Series 0, the complete lists of which follow (PP=Professional Paper; B=Bulletin; WS=Water-Supply Paper); SERIES K, PUMPING WATER. WS 1. Pumping water for irrigation, by H. M. Wilson. 1896. 57 pp., 9pl.s. (Out of stock.) WS 8. Windmills for irrigiition, by E. C. Murphy. 1897. 49 pp., 8 pis. (Out of stock.) WS 14. New tests of certain pumps and water lifts used in irrigation, by O. P. Hood. 1898. 91 pp., Ipl. (Out of stock.) WS 20. Experiments with windmills, by T. O. Perry. 1899. 97 pp., 12 pis. (Out of stock.) WS 29. Wells and windmills in Nebraska, by E. H. Barbour. 1899. 8.5 pp., 27 pis. (Out of stock.) WS 41. The windmill; its efficiency and economic use, Pt. I, by E. C. Murphy. 1901. 72 pp., 14 pis. (Out of stock.) WS 42. The windmill, Pt. II (continuation of No. 41). 1901. 73-147 pp., I.t-16 pis. (Out of .stock.) WS 91. Natural features and economic development of Sandusky, Maumee, Muskingum, and Miami drainage areas in Ohio, by B. H. Flynn and M. S. Flynn. 1904. 130 pp. WS 117. The lignite of North Dakota and its relation to irrigation, by F. A. Wilder. 1905. 59 pp., 8 pis. WS 1.36. Underground waters of Salt River Valley, Arizona, by W. T. Lee. 1905. 196 pp., 23 pis. WS 141. Observations on the ground waters of the Rio Grande Valley, 1904, by C. S. Slichter. 1905. 83 pp., 5 pis. WS 153. The underflow in Arkansas Valley in western Kansas, by C. S. Slichter. 1906. 90 pp., 3 pis. SERIES O, UNDERGROUND WATERS. WS 4. A reconnaissance in southeastern Washington, by I. C. Russell. 1897. 96 pp., 7 pis. (Out of stock.) WS 6. Underground waters of southwestern Kansas, by Erasmus Haworth. 1897. 05 pp., 12 pis. (Out of stock.) WS 7. Seepage waters of northern Utah, by Samuel Fortier. 1897. 50 pp., 3 pis. (Out of stock.) I II " SEEIES LIST. WS 12. Underground waters of southeastern Nebraska, by N. H. Darton. 1898. 56 pp., 21 pis. (Out of stock.) WS 21. Wells of northern Indiana, by Frank Leverett. 1899. 82 pp., 2 pis. (Out of stock.) " WS 26. Wells of southern Indiana (continuation of No. 21), by Frank Leverett. 1899. 64 pp. (Out of stock.) WS 30. Water resources of the Lower Peninsula of Michigan, by A. C. Lane. 1899. 97 pp., 7 pis. (Out of stock.) WS 31. Lower Michigan mineral waters, by A. C. Lane. 1899. 97 pp., 4 pis. (Out of stock.) WS 34. Geology and water resources of a portion of southeastern South Dakota, by J. E. Todd. 1900. 34 pp., 19 pis. WS 53. Geology and water resources of Nez Perces County, Idaho, Pt. I, by I. 0. Russell. 1901. 86 pp., 10 pis. ((^ut of stock.) WS 54. Geology and water resources of Nez Perces County, Idaho, Pt. II, by I. C. Russell. 1901. 87-141 pp. (Out of stock.) WS 55. Geology and water resources of a portion of Yakima County, Wash., by G. O. Smith. 1901. 68 pp., 7 pis. (Out of stock.) WS 57. Preliminary list of deep borings in the United States, Pt. I, by N. H. Darton. 1902. 60 pp. (Out of stock.) WS 59. Development and application of water in southern California, Pt. I, by J. B. Lippincott, 1902. 95 pp., 11 pis. (Out of stock.) WS 60. Development and application of water in southern California, Pt. II, by J. B. Lippincott, 1902. 96-140 pp. (Out of stock.) WS 61. Preliminary list of deep borings in the United States, Pt. II, by N. H. Darton. 1902. 67 pp. (Out of stock.) WS 67. The motions of underground waters, by C. S. Slichter. 1902. 106 pp., 8 pis. (Out of stock.) B 199. Geology and water resources of the Snake River Plains of Idaho, by I. C. Russell. 1902. 192 pp. , 25 pis. WS 77. Water resources of Molokai, Hawaiian Islands, by W. Lindgren. 1903. 62 pp., 4 pis. WS 78. Preliminary report on artesian basins in southwestern Idaho and southeastern Oregon, by I. C. Russell. 1903. 53 pp., 2 pis. PP 17. Preliminary report on the geology and water resources of Nebraska west of the one hundred and third meridian, by N. H. Darton. 1903. 69 pp., 43 pis. WS 90. Geology and water resources of a part of the lower James River Valley, South Dakota, by J. E. Todd and CM. Hall. 1904. 47 pp., 23 pis. WS 101. Underground waters of southern Louisiana, by G. D. Harris, with discussions of their uses for water supplies and for rice irrigation, by M. L. Fuller. 1904. 98 pp., 11 pis. WS 102. Contributions to the hydrology of eastern United States, 1903, by M. L. Fuller. 1904. 522 pp. WS 104. Underground waters of Gila "Valley, Arizona, by W. T. Lee. 1904. 71 pp., 5 pis. WS 106. Water resources of the Philadelphia district, by Florence Bascom. 1904. 75 pp., 4 pis. WS 110. Contributions to the hydrology of eastern United States, 1904; M. L. Fuller, geologist in charge. 1904. 211 pp., 5 pis. PP 32. Geology and underground water resources of the central Great Plains, by N. H. Darton. 1905. 433 pp., 72 pis (Out of stock.) WS 111. Preliminary report on underground waters of Washington, by Henry Landes. 1904. 85 pp., ipl. WS 112. Underflow tests in the drainage basin of Los Angeles River, by Homer Hamlin. 1904. 55 pp., 7 pis. WS 114. Underground waters of eastern United States; M. L. Fuller, geologist in charge. 1904. 285 pp., 18 pis. WS 118. Geology and water resources of east-central Washington, by F. C. Calkins. 1905. 96 pp., 4 pis. B 252. Preliminary report on the geology and water resources of central Oregon, by I. C. Russell. 1905. 138 pp., 24 pis. WS 120. Bibliographic review and index of papers relating to underground waters published by the United States Geological Survey, 1879-1904, by M. L. Fuller. 1905. 128 pp. WS 122. Relation of the law to underground waters, by D. W. Johnson. 1905. 55 pp. WS 123. Geology and underground water conditions of the Jornada del Muerto, New Mexico, by C. R. Keyes. 1905. 42 pp., 9 pis. WS 136. Underground waters of Salt River Valley, Arizona, by W. T. Lee. 1905. 196 pp., 24 pis. B. 264. Record of deep-well drilling for 1904, by M. L. Fuller, E. F. Lines, and A. C. Veatch. 1905. 106 pp. PP 44. Underground water resources of Long Island, New York, by A. C. Veatch, C. S. Slichter, Isaiah Bowman, W. O. Crosby, and R. E. Horton. 1906. 394 pp., 34 pis. WS 137. Development of underground waters in the eastern coastal plain region of southern Cali- fornia, by W. 0. Mendenhall. 1905. 140 pp., 7 pis. WS 138. Development of underground waters in the central coastal plain region of southern Cali- fornia, by W. C. Mendenhall. 1905. 162 pp., 5 pis. SERIES LIST. Ill \VS I ;;'.!. I)c\i'l(i|iiiicni 111' unilcrt;r(itinil waters in llu' wesU-rii conslul iiliiiii rcKi"ti i.f suiiihcrn Ciili- foriiiii, l.y \V. C. M.iHlciilmll. 1 1 10. Fiohl moiisuroiiu'iits ni (lie into of inovcmciil of inind Ohio, by Frank Leverett, in Eighteenth .-Vnnual, Pt. IV; New developments in well boring and irrigation in eastern South Dakota, by N. H. Darton, in Eighteenth Annual, Pt. IV; Rock waters of Ohio, by Edward Orton, in Nineteenth Annual, Pt. IV; Artesian well prospects in the Atlantic coastal plain region, by N. H. Darton, Bulletin No. 138. Correspondence should be addressed to The Director, United States Geological Survey, Washington, D. C. May, 1906. o u N '06 ^ ^ .^.'