C B I02.S" Watei -Supply and Lrigalkm Paper No. 155 0, Underground DEPARTMENT OF THE INTERIOR UNITED STATES GEOLOGICAL SURVEY CHAR ES D. WALCOTT, Director FLUCTUATIONS OF THE WATER LEVEL IN WELLS. WITH SPECIAL REFERENCE TO LONG ISLAND, NEW YORK BY _A. C. YEATCH WASHINGTON' GOVERNMENT PRINTING OFFICE 1906 Glass \Z — Book. Digitized by the Internet Archive in 2011 with funding from The Library of Congress http://www.archive.org/details/fluctuationsofwaOOveat ifs* Water-Supply and Irrigation Paper No. 155 Series 0, Underground Waters, 52 DEPARTMENT OF THE INTERIOE UNITED STATES GEOLOGICAL SURVEY CHARLES 1>. VVALCOTT, Dieectok 76 FLUCTUATIONS OF THE WATER LEVEL IN WELLS, WITH SPECIAL REFERENCE TO LONG ISLAND, NEW YORK J±. C. YEATCH WASHINGTON GOVERNMENT PRINTING OFFICE 1 9 6 o "V ^> G- A. n Long Islam of the fluctuations of water in w such that it may be affirmed that the underground water is derived wholly from the rain which falls on the surface of the island, and the problems involved are, there- fore, not unduly complicated, as they are in many regions, by the possibility of the influx of water from other areas. In addition to this comparatively complete ground-water isolation, the is- land is of such a size — 120 miles long and 20 miles wide — that ground-water phenomena can attain a relatively complete de- velopment, and the geologic structure of the water-bearing beds, while not complicated, is sufficiently varied to produce several differing conditions. Topographically the western part of Long Island — the portion involved directly in this paper — may be said to consist of a single range of rolling hills, usually 150 to 250 feet high, though in one place attaining an elevation of over 400 feet. This hill range descends somewhat abruptly to the north shore, where it is cut by several reentrant bays occu- pying old valleys. On the south side is a very flat gravel plain, sloping gently to the ocean, along which a series of barrier beaches inclosing long marshes has been developed. To the east the hill range divides and produces two hilly peninsulas, each with a single ridge on the northern side. Geologically the island may be sand beds, containing irregular , New York, are particularly favorable for the study ells. The geologic and topographic conditions are -Sub O.ceanic springs -Cretaceous flowing weM North Shore flowing well dependent on local clay bed Springs Springs which supply Brooklyn Waterworks ponds regarded as a series of relatively porous gravel and and discontinuous clay masses, the whole limited 10 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. below by the peneplained surface of a mass of highly disturbed ana metamorphosed Paleozoic and pre-Paleozoic rocks, which have little water value except as a more or less complete barrier to do wn ward percolation ( fig. 1 ) . While these unconsolidated beds represent, in the geologic time scale, several of the divisions of the Upper Cretaceous and as many as five Pleistocene or glacial stages, and as a whole are stratified deposits dipping at very low angles south and southeastward, they are, under the island, essentially continuous from a water standpoint, and the rain falling on the surface is relatively free to pass to any part of the mass. The Pleistocene beds which form the surface are, however, as a rule, coarse and more porous than the underlying Creta- ceous and tend to increase the absorbing power of the island. As a result, the per- centage of rainfall which passes into the streams without first going through the ground is extremely small. « This percolating water has entirely saturated the porous strata above the bed rock, except a limited portion at the surface, and has driven out the salt water which filled these beds when they were first deposited and which reoccupied them, at least in part, during the several submergences to which this region has been subjected. The surface of this zone of complete saturation, or the main ground- water table, is coincident with the sea level at the shores and becomes more and more elevated in passing inland, though the rate of increase of elevation is less than that of the surface, of which it is but a subdued reflection (fig. 1).& This slope of the ground-water table permits the development of artesian wells at many points on the coast, at elevations which are commonly less than 10 feet above high tide. The head is, in all cases, due to the greater height of the ground water in the adjacent hill mass. In order that such a differential head may be developed, it is merely necessary that the water-bearing bed in question be coarser than the over- lying beds. A clay or other impervious cover is not essential and, indeed, is often absent. OBSERVATIONS OF THE UNITED STATES GEOLOGICAL SURVEY. Observations on the fluctuations of the water level in wells were made by the Geo- logical Survey near Huntington, Oyster Bay, Valley Stream, Millburn, Long Beach, and Douglaston, all villages on Long Island west of longitude 73° W., and between latitudes 40° 35' and 40° 55 / N. (PI. I. ) OBSERVATIONS WITH DIRECT-READING GAGES. OBSERVATIONS AT HUNTINGTON, N. Y. The Huntington observations, from which the other Survey observations developed, were undertaken to test the common report that the discharge of most of the artesian wells along the northern shore of Long Island fluctuated with the tide; in some cases the flow ranging from at low tide to over 100 gallons per minute at high tide. Nearly all of these wells were being pumped, or were utilized to run rams, but per- mission was obtained to gage a newly completed well belonging to the Huntington Light and Power Company, at Huntington Harbor, until it should be connected with the pumps — a period of three or four days. A direct-reading float gage of simple type was quickly constructed by Baker & Fox, Brooklyn, N. Y. This consisted of a 2-inch cylinder of brass carrying a £-inch alu- minum rod 6 feet long and graduated to hundredths of a foot, with the zero point just "Spear (Rept. New York City Commission on Additional Water Supply, 1904, p. 829) has estimated that 43 per cent of the total stream flow (or 14 per cent of the rainfall ) can be considered as flood flow or as not having passed through the ground. He bases this judgment on the relative heights of the stream and ground-water levels near the south shore, where, as explained on page 51, a correct judg- ment can not be formed. The average flood flow is believed to be much less than 5 per cent of the precipitation. b For details of the slope of the ground- water table see Prof. Paper U. S. Geol. Survey No. 44, 1906, Pis. XI, XII. OBSERVATIONS WITH DIREOT-RE ADING GAGES. 11 above the cylinder. For convenience in carrying, ;is well as to avoid the use <>f so long a mil except where absolutely necessary, the rod was divided into three parts and jointed. The cylinder was so constructed that it would just carry the total length of (i feet, and when used with only 2 or 4 feet of .rod, weights, balancing the effect of the part removed, were added to the bottom of the cylinder. Some trouble was experienced by the float tending to approach theside of the well and develop a thin capillary film between it and the pipe, which decreased the sen- sitiveness of the gage. It is suggested that when direct-reading floats are usee I in wells of small diameter they be kept, away from the walls of the well by means of Fig. 2.— Sketch map showing location of well of Huntington Light and Power Company at Hunt- ington Harbor, N. Y. slightly arched wires, as in the float devised by Professor King for the self-recording gages used in the Madison experiments and later on Long Island. The well of the Huntington Light and Power Company is situated on a dock at Huntington Harbor, near Halesite post-office (PI. I, fig. 2. ) The natural level of the surface at the point where the well is sunk is between high- and low-tide mark, but the ground has been built up by filling about 5 feet higher. The well is 75 feet deep and 4 inches in diameter, and the water rises in the pipe from 1 to 3 feet above the surface of the made ground. The well was piped above the limit of flow, so that all the fluctuations could be measured directly, rather than inferred from variations in the rate of discharge. 12 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. The geologic section reported by the driller, Mr. H. J. Dubois, is as follows: Section of well of Huntington Light and Power Company at Huntington Harbor, Halesite, N. Y. Feet. 1. Filled ground 0- 6 2. Swamp deposit 6-10 3. Dark sand and gravel 10-70 4. Blue clay 70-71 5. Light-yellow gravel containing artesian fresh' water 71-75 The artesian flow is due to the height of the ground water in the steep hill just east of the well, the head being transmitted through the coarse beds encountered in the bottom of the well. Water escapes in many springs along the beach, but the movement through the lower gravels is freer and a differential head is produced, r V i i /\4*\,R 1 X V X ^ . X i*^ 3.4-3. o Ms -troo t y 1) X""" \ / - "^ v\ / "^ ■\- v / ,/ V \ s . \ - -^ \ t ] y£ v J \ \ s May 9, \ ay 10, s / i -4./.5\ . -> .-•) 5/.' -■ ;■-■; -\- s - - \ \ " N /\"~- v T N, + r \ May 9, f \ fyiQ. *■ \ X \ ^4 ssi \ > u S ' 40 4.1 >IJ. ~K j \\ er ( lompany. snd of doorsill, The elevation jeneral character as I hose plotted Tlic bench mark established was as foll< west side of building of Huntington Light and above the zero of gage No. 2 is 12.232 feet. The observations here gave curves of the same at Oyster Bay (fig. 8, p. 16). The two curves arc essentially parallel, ami while I lie semidiurnal range of the tide is about S feet, that of the well is abOUl 2 feet. The low water in the well is 3.5 feet, above the highest water in the hay. In fig. :; the high- ami low-tide readings for both the well and the tide have been plotted on a large scale, with the same time values, hut with different vertical values, in order to bring out the amount of time the fluctuations in the well lag behind those of the tide. These curves indicate that for the period of observation the low tide in the well has an average establishment, or lag behind the ocean waters, of two minutes and the high tide eight minutes. On May 11 Mr. Isaiah Bowman made observations with the direct-reading floats on a shallow 6-inch flowing well belonging to the Consolidated lee Company, at Huntington. It is located about a mile from the harbor, at an elevation of 40 feet. The observations were continued for six hours and no fluctuations of any character noted. OBSERVATIONS AT OYSTER BAY, N. Y. During the month following the observations at Huntington Harbor it was found that among the many flowing wells at Oyster Bay four could be observed for a lim- ited time. OYSTER BAY HARBOR. Fig. 4. — Sketch map showing location of wells observed at Oyster Bay, N. Y. These wells are all very near the shore; indeed, the Casino well, which is beneath the floor of a building extending over the water, is always covered at high tide (fig. 4). The others, the Burgess, Lee (or Hill), and Underhill w r ells (fig. 4), are at distances of from 50 to 500 feet from ordinary high-tide mark, though the ground at all is cov- ered when extraordinary wind-aided tides occur. The depths of these wells, as determined by soundings at the time of the observations, were: Casino, 93 feet; Underhill, 114 feet; Lee, 188 feet; Burgess, 155.5 feet. The Casino and Lee are 3-inch wells, and the Burgess and Underhill 2-inch wells. All these wells pass through a surface layer of sand and gravel, then a layer of blue clay 50 to 75 feet thick, and finally penetrate a rather coarse water-bearing sand. In some cases the upper sands and gravels will furnish flowing water, but all the wells observed obtain their supply 14 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. from below the blue-clay layer (fig. 6). This blue clay thins rapidly southward and entirely disappears half a mile south of the wells (fig. 7). It extends under Oyster Bay Harbor and is exposed in the clay pits on the south end of Center Island. <•* LONG ISLAND SOUND ,Hoyt vvell Fig. 5.— Sketch map showing topographic surroundings of wells shown in fig. 4 and location of sections shown in figs. 6 and 7. All these wells were flowing, and in each case, before observations were commenced, lengths of pipe were added until the wells no longer flowed, even at high tide. Float gages similar to those used at Huntington Avere then inserted and the wells covered ^nd amdgravel^. ... _r_ r gr^— Tc, obtain more refined results than were possible with the board gage at Huntington, a 3-inch pipe, perforated at a point several feel above the bottom, was driven in the harbor at the end of a row of piles and at a distance of about 200 feet from the shore (tig. 4). This still box or tide well was fitted with a direct-reading float gage like those used in the artesian wells. This arrangement is not to be rec- ommended during stormy weather, but fortunately during the whole time of obser- vation at this place no trouble was experienced from that cause. 16 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. Observations wei'e commenced on the Casmo, Underbill, and Burgess wells on tbe evening of May 30, by a party in charge of Mr. Isaiah Bowman, and continued, with interruptions on the nights of May 30 and 31 and June 1, to 10 p. m. on June 4. On June 10 and 11 observations were made on the Lee (or Hill) well, covering two high arid two low tides, and for the purpose of comparison the Casino and tide wells were also observed. Observations were generally made every minute for thirty minutes preceding and following the times of high and low water, and from these values the curves shown in fig. 8 were drawn. Times of high and low water were, found by plotting the observations near high- and low-tide marks on a much larger scale, in the manner shown in fig. 3. The values so obtained are indicated on fig. 8, and are given in the following table: Difference in time between high- and low-water stages in four artesian wells at Oyster Bay, N. Y., and the tide in Oyster Bay Harbor. [Time expressed in hours and minutes of 24-hour clock.] HIGH TIDES. 1903. May 30. May 31. June 1. June 2. June 3. June 4. June 10. June 11. Aver- age lag. 14.52 14.43 17.05 16.54 5.19 5.12 18.11 18.02 6.38 6.29 12.22 12. 20 Min- utes. Tide Difference (Jag) ... .09 .11 .07 .09 .09 .02 0.8 15.13 14.43 16.05 15.44 17.15 16.54 5.38 5.12 18.25 18.02 6.56 6.29 Tide .30 .21 .21 .26 .23 .27 24.7 a0.20 23. 48 13.12 12.20 Tide .32 .52 .42 16.02 14.43 17.02 15.44 18.00 16.54 6.20 5.12 19.10 18.02 7.41 6.29 Tide 1.19 1.18 1.06 1.08 1.08 1.12 71.8 LOW TIDES. 20. 28 20.10 9.21 9.04 23.46 12.08 0.47 .38 17. 42 17.33 6.32 6.20 Tide 23.36! 11.55 Difference (lag) . . . .18 .17 . 10 . 13 .09 .09 .12 12.6 10.41 10.00 11.26 10. 53 a 0.10 12.28 11.55 1.04 .38 Tide 23.36 .41 .33 .34 .33 .26 33.4 18.23 17.33 7.26 6.20 Tide Difference (lag) ... .50 1.06 .58 21.25 20.10 10.17 9.04 11.22 10.00 12.06 10.53 a 0.55 23.36 13.11 11.55 1.49 .38 Tide Difference (lag) ... 1.15 1.13 1.22 1.13 1.19 1.16 1.11 75.6 a Morning of following day. Mn 0' o 8>l £ O . Ouj 3* i § 55 M 5< — z 3h LONG ISLAND OBSERVATIONS. 17 The c< the time nlinuat'h v interes 3ERVATIONS WITH SELF-RECORDING GAGES. INSTRUMENTS USED. hi of the observations by means of self-recording gages was due to it. of Mr. F. 11. Newell and Prof. Charles S. Slichter. .Mr. Newell Feet adove low tide O (6 p "« «< - 3- «! 5' o- o? o>? r, P / " OS C Z m ro ' '.53 -" EG !:0l <, & (^ Y a * v c 4. , 14 \ v\ 7 . m 5.1' / i y A 1 / /o y- // r ^ I 1- / // 6 1:46 ,5 55 ,! 10 S3 C z m CO \ si; :38 ^..'0 / / / 1/ ,, 55 /) '-'■ — ' ;2R \ l us t 6:1 ■ , / '■' 7 / J fi // / 1/ ---' ?:iB H :47 // OJ c 2 m 1 , \ \ \ ^ \ ■j . 3E \ 16 56 r:4l 4 / / // ; / visited the. island when the observations at Oyster Bay were in progress, and at once directed that three Friez water-stage registers be purchased. These were supple- irr 155—06— — 2 18 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. mented by a gage constructed at Purdue University from the designs of Mr. Elwood Mead. Shortly after Mr. Newell' s visit, and before the Friez gages had been received, Prof. Charles S. Slichter arrived to take charge of the measurement of the rate of underflow. "* He kindly obtained the loan of five of the gages used by King in his experiments at Madison, Wis. ; a of these, four were week gages and one a one-day gage. The King gages were constructed by H. Green, Brooklyn, from barograph stands; they consist of an ordinary barograph cylinder, driven by a double-spring marine clock, the recording device being a simple lever on a cone bearing with a pen on one end and a place for attaching the float on the other. At the point where the clock motion is transmitted to the drum there was a slight amount of play which King found would introduce into the records an error of one to two hours. A fric- tion brake was, however, subsequently added to overcome this defect. The gages as received on Long Island were adjusted to magnify the fluctuations two or more times; and as this scale was entirely too great for the wells observed, the arm was extended until the ratio was 1:2 and a reduction of one-half thereby obtained. These gages were found to be more sensitive and reliable than any others used. By means of the simple lever with its cone bearing, the friction in this instrument is reduced to a minimum; the pens respond to the slightest movement of the water, and for the faithful reproduction of small fluctuations this simple type of gage is to be highly recommended. In the Mead gage the recording drum is vertical and the pen is carried by a carriage working between two upright guides. The wire supporting the carriage winds about a wheel connected with the wheel around which a wire from the float passes, and is lifted and lowered as the float descends and rises. The float and the wheel to which the pen is attached are so related in diameter that the curve traced is 10/42 of the true scale. There is with this gage, as with most gages where the recording cylinder is driven by the clock, some lost motion at the point of connection. This is particu- larly bad in this instrument. On the Long Beach records great care was used in set- ting the gage and the trouble was avoided, but some of the curves from well No. 8, at Douglaston, are clearly in error two to three hours. In the Friez gage & the recording drum is horizontal and is moved by the float, while the pen is moved by the clockwork. It was found that with the size of float that must be used in wells of small diameter the inertia of the drum in this instru- ment was such that it would not move until considerable head was developed and that small fluctuations were often not recorded. There was also a considerable amount of lost motion in the cogs used in the reducing device; and while an eccen- tric was provided for engaging the cogs closer, this could not be done without so increasing the friction that the instrument was useless. As a whole, this gage is not sufficiently sensitive for this kind of work, and the time element is entirely too small. A water-stage register manufactured by a western house was also used, but the results obtained were not satisfactory because of the poor mechanical construction of the gage. OBSERVATIONS ON WELL OF QUEENS COUNTY WATER COMPANY, 1 MILE WEST OF HEWLETT, N. Y. Through the kindness of the chief engineer of the Queens County Water Company, Mr. Charles R. Bettes, an artesian well 181 feet deep and 3,300 feet south of the company's pumping station (PL II) was covered with a shelter for the protection of the gages and placed at the disposal of the Survey. This well, as is common with the wells of about the same depth sunk near the pumping station, passes through a layer of surface sand and gravel, then through beds of clay and other fine material a Bull. U. S. Weather Bureau No. 5, 1892. b Manufactured by Julian P. Friez, Baltimore, Md. ■ JULY M 12 * THEF A ~r~ \ / s (6 n 'V- / AUT / OF i -^ r BAR AT B ( 26 r RECO - UUtb P r (30C- f\ T^ f\ • ra[~ AT F| BRl (Respect 26 mil* From Fri< i l\ /VLETT, N. > s inverted. \ : U. S. GEOLOGICAL SURVEY WATER-SUPPLY PAPER NO. 155 PL. Ill 1 JUNE 30 ! JULY 1 | JULY 2 I JULY 3 JULY 4 | JULY 5 I JULY 6 JULY 7 | JULY 8 I JULY I 1 ,„, Y ,„ 1903 M 12 M . 12 M 12 ' M 12 M .. 12 . ■ M 12 M 12 : M 12 i 19 I \. ' JULV 10 JULY 11 THERMOGRAPH RECORD ^ 90 AT FLORAL PARK, „ 80 NEW YORK. £ 70 (6 miles north of well.) S 60 -/V ^ •~ /"' / "^~\n r- -JS -i A. y I /^ V ^ -/- / ^_ \ ' r V. V./ / X . / / 45 AUTOGRAPHIC RECORD „§ 46 OF FLUCTUATIONS IN Jg 4? A 181 FT. WELL ~ -° 48 AT FENHURST JuJ (HEWLETT) >< 49 NEW YORK. oj 50 ( Inverted curve) s J- S/L 1 / v/ -\ / I \/ - -" / "\ ,y \ z ^J / ly V i =V —i / \ \ / V i -J V ) V / BAROGRAPH RECORD o> w 3 30.0 AT BRENTWOOD, N. Y. £0 (26 miles north of well.) is 29 - 75 240 RECORD OF PUMPING AT QUEENS CO. WATER CO'S o 20 ° PUMPING PLANT < , 60 AT o § 120 FENHURST, 5 NEW YORK. i 8 ° (3000 feet from well.) 40 J^ -,r I r r r Tl \ ^V 7°- n r ^ } \ in Jl -h M r \ in 1 ,r """l J i \' ^1 ,f p 1 \~ f 1 j .n i v u «L \ ,r L 1 J IH \ if 1 J If 1^ 1 ^f il IT } i, i J L / u \\ "Lr 1 i-n. j lT U II 1 T; t^ m? T-r T. U u 0.00 RAINFALL RECORD AT FLORAL PARK, AND i S 0.50 BRENTWOOD, N. Y. (Respectively 5 miles north and - 26 miles northeast of wells.) H Loo From Friez self-recording gages. a 1.50 > 0.19 - B Rair 008 rood 1 E r° 00 ,1.0 i 3t 8-5 )2 il jP.t. 1. Ex lai itio II r II o uns the at Bre e a itw Fl rat p.;: ind icat ed ly c Otti dli les and the rd ' Br< ntv\ oo« PARTIAL RECORD OF FLUCTUATIONS OF WATER LEVEL IN A 181-FOOT WELL AT HEWLETT, N. Y. From a King gage in charge of Francis L. Whitney and F. D. Rathbun, field assistants. Well curve is inverted. ol'SKKVATIONS WITH SELF-RECORDING GAGES. 19 into a rather coarse gravel, which yields an abunda.nl supply of flowing water." The whole section is of Pleistocene age. There are in the immediate vicinity of the pumping station thirty-two 5-inch wells, ,">:; feel deep, and nineteen 6-inch wells, L50 to 190 feel deep. These are arranged along two lines, one extending northwest and the other southwest from the pumping station. The extreme end of the south line is about 1,000 feet from the pumping station, and the well observed is therefore over 2,000 feet from the nearest pumped well and but slightly to one side of the probable direction of flow. (PI. II.) It was the opinion of Mr. Bettes thai this well was not affected by pumping, and through a misinterpretation of the first records it was believed that this surmise was correct. Considerable discussion was therefore caused when it was found that the well fluctuations simulated the thermograph curve in a remarkable manner, that these fluctuations were inversely related to the tempera- ture (a rise in temperature causing a fall in water), and that the changes manifested themselves with a lag of but one or two hours behind apparently similar temperature fluctuations. (PI. III.) The hourly pumpage was kindly furnished by Chief Engineer Bettes, and this record, when plotted with the well curves, conclusively demonstrated that the rhythmical fluctuations were due more to pumpage than to temperature (PI. III). Fluctuations of a somewhat similar character are produced by temperature changes (see p. 54), and this element is doubtless present in this curve. On PI. Ill the effect of pumpage is clearly shown in the double cusps of the well curves on the night of July 5-6. The temperature curve shows no such variations. Similarly, the records on June 21, 25, and 26 show important differences between the well and temperature curves, which are largely due to pumping. These results are important because of the rapid rate of transmission. The effect of this pumping is felt at a distance of 2,000 feet or more, with a time lag of but one or two hours. This contrasts sharply with the very slow transmission noted in the pressure changes due to tidal loading and to the inflow and outflow along rivers (see pp. 60, 65). It conclusively proves that the supply here is large; that the beds are quite porous, and that the normal water flow is rapid. In the records from the day gage, which was maintained here for the first ten days, the larger time scale brought out very clearly a series of regular minor fluctua- tions which were not clearly defined with the smaller scale of the week records. The most pronounced of this series recurs day after day and has a period of very nearly twenty minutes and a range of 0.06 to 0.08 inch. Besides these vibrations, with a period of twenty minutes, there are several fluctu- ations of smaller amplitude and period. One series has a period of about five or six minutes, but it is so involved that little can be definitely stated regarding it. An instrument with a large time scale, 1 or 2 inches to the hour, and a vertical scale of once or twice the normal would record, at this place, a very complicated series of small recurrent vibrations. OBSERVATIONS AT LONG BEACU, N. V. The deep flowing well of the Long Beach Association at Long Beach, N. Y. (PI. II), offered a most excellent opportunity for the observations of tidal fluctuations. It is situated on a narrow, sandy barrier beach, separated from the main island by a sea marsh 2 to 3 miles wide, cut by narrow tidal channels, and is entirely removed from the influence of any pumping station. This well is 3 inches in diameter and 386 feet deep. The water is obtained in sands of Cretaceous age and rises 2 to 4 feet above the surface of the ground or 10 to 12 feet above sea level. The general geologic relations may be inferred from the diagrammatic cross section given in fig. 1 (p. 9). a Fur detailed record of strata in near-bv wells see Prof. Paper U. S. Geol. Survey No. 44, 1906, p. 225. fig. 6(5. 20 FLUCTUATIONS OF THE WATER LEVEL IN WELLS The section reported by the driller, Mr. W. C. Jaegle, is as follows: Section of well of Long Beach Association, at Long Beach, N. Y. Feet. 1. White sand .' 0-36 2. Dark sand and creek mud 36- 40 3. White gravel, containing saltwater 40- 51 4. White sand 51-55 5. Dark sand 55- 65 6. Whitesand 65- 70 7. White gravel 70- 73 8. Yellowsand 73- 76 9. Blue clay 76-82 10. Yellow gravel 82- 90 11. Creek mud 90-99 12. Dark fine sand, containing lignite .' 99-101 13. White sand . 101-111 14. Dark sand 111-119 15. White sand, with lignite 119-121 16. Blue clay 121-135 17. Fine white sand 135-143 18. Gravel, with salt water 143-145 19. Dark sand 145-156 20. Gravel, with salt water 156-158 21. Clay 158-174 22. White sand, containing at 190 feet a log of lignitized wood 174-192 23. White gravel and salt water ? . . .. 192-196 24. Clay 196-200 25. Fine sand 200-220 26. Solid blue clay 220-270 27. White sand and wood, containing fresh water, sweet and chalybeate 270-276 28. Clay 276-282 29. White sand and wood 282-297 30. Blue clay 297-305 31. White sand, wood, and water 305-308 32. Blue clay 308-317 33. White sand, containing wood and artesian water 317-325 34. Blue clay 325-340 35. White sand and mineral water; has considerable C0 2 , sparkling and effervescent 340-356 36. Blue clay 356-360 37. White sand and pure water 360-378 38. Blue clay 378-380 39. White sand 380-381 40. White clay 381-383 41. Fine sand, with artesian water 383-386 Mr. F. D. Rathbun was placed in Charge of these observations and by a careful readjustment of the Mead gage obtained very excellent curves (PI. IV). Indeed, for this character of work the results from the Mead gage, as set up by Mr. Rath- bun, are better than from the Friez gage. It was impossible to make tide observations at this point, and the values plotted on the curve are taken from those predicted by the Coast and Geodetic Survey a for East Rockaway Inlet, which is 2.8 miles west of the well. The difference in time all. S. Coast and Geodetic Survey Tide Tables for 1903, p. 346. U. S. GEOLOGICAL SURVEY 1903 JULY 7 M 4 8 12 16 20 M. <£ O Q -1. MEAN SEA ' 2 - 5 l HI I I I I M I PARTIAL RECORD OF FLUCTUATIONS OF WATER LEVEL IN A 386-FOOT WELL AT LONG BEACH, N. Y, From a Mead gage in charge of F. D. Rathbun, field assistant, OKSKRY ATIONS WITH SELF-RECORDING GAGES. 21 between the tide at this point and on tlic beach in fronl of the well is probably nol more than tWO Or three minutes, since at New Inlet, •">.:! miles east (if the well, high and low tides occuronly from four to six minutes earlier than at East Rockaway 1 1 1 1« -t . The data regarding the tidal fluctuations in the channel behind Long Beach are very meager. Mr. Paul K. Ames, of the Long beach Association, through whose kindness it was possible to make observations on this well, states that the amplitude of the tide in broad and Long beach channels is about I feet, or the average of the fluctuations in the open ocean at this point, and that in the tides behind Long beach there is a lag of about one hour. Observations a few miles to the west in Jamaica Bay, which is separated from the main ocean by a sand bar similar to that at Long Beach, show that at various points behind the bar the high and low water are from thirty-two to eighty minutes behind the same stages in the ocean, and that the lag at high water is less by from six to thirty minutes than at low water." It may be confidently affirmed that a similar lag occurs in the channels behind Long Beach. The well curve obtained is a very smooth curve, which is apparently the simple resultant of the tides in the ocean and those inside of the bar. The extreme regu- larity of the low-tide values is believed to be due to the modifying influence of the tides in the inner channels, which, because of the shallow character of the outlets into the ocean, would have a much smaller low-tide variation than the ocean. The lag of the high and low waters in the well, compared with the ocean tide, is given in the following table: Table showing difference in lime between high ami low o vater in a 386-foot well at Long Beach, N. Y., and the tide as predicted by the United States Coast and Geodetic Survey for East Rockaway Inlet (PL IV). June July July : July July- July July July 1 July July- July 1903. High water. Well. Time. 2.20 15.20 3 20 16.10 4 05 17.05 7.05 19. 25 7. 35 20 10 8.35 20.50 9.15 21.35 9.55 22.05 10.10 23.05 Tide Time 0.13 12 53 1.10 13 52 2.13 14.49 3.13 15.46 5.58 18.14 6.44 13 55 7.25 19.34 8.03 20.11 8.38 20.46 9.09 21.19 Differ- ence. Hours and min- utes 1 10 1.28 1.07 1.21 0.52 1.29 1 07 1.11 0.61 1.15 1.10 1.16 1.12 1.24 1.17 1.19 1.31 1.46 Low water. Well. Time 18 30 7.00 19 50 8 00 20 50 9 20 22 00 10.30 40 12 20 1.10 12 50 2.00 13. 55 2. 50 14.35 3.20 15.15 3.40 15. 35 Tide Tune 18.09 6.36 19.10 7 32 20 14 8.25 21.16 9.26 23. 59 11. 59 44 12.44 1.24 13.24 2.02 14.04 2.36 14.38 3.11 15. 09 Differ- ence. Hours and min- utes. 0.21 24 0.40 0.28 0.36 0.55 0.44 1.04 41 0.21 0.26 0.06 0.36 0.31 0.48 0.31 0.44 0.25 0.29 0.26 a U. S. Coast and Geodetic Survey Tide Table for 1903, p. 346. 22 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. Table shoiving difference in time between high and low water in a 386-foot well at Long Beach, N. Y. , and tide at East Rockaivay Inlet — Continued. Date. High water. Well. Tide. Differ- ence. Low water. Well. Tide. Differ- ence. July 13 . July 14. July 15. July 16. Time. 11.50 23.00 11.40 23.45 Time. 9.44 21.54 10.21 22.28 Hours and min- utes. 2.06 1.06 1.19 1.17 12.20 0.30 11.03 23.09 1.07 1.21 July 17 Average. 1.14 Time. 4.50 16.55 5.10 17.10 5.55 18.00 6.20 19. 20 7.20 Time. 3.44 15.44 4.19 16. 21 4.55 17.04 5.34 17.55 6.27 Hours and min- utes. 1.06 1.11 0.51 0.49 1.00 0.56 0.46 0.25 0.53 0.40 It will be noted that the lag at high water is greater than at low, which is just the reverse of what occurs in tidal rivers, where the water rises much more rapidly than it falls and the low-water lag is long and the high-water lag short. This is an important result bearing on the relation between the fluctuations of the water in wells and the oceanic tide and clearly refutes the doctrine that the tidal fluctuations here are due to leakage and that the fluctuations are analogous to those of tidal rivers. (See pp. 63-67.) OBSERVATIONS ON THE SCHREIBER WELL NEAR MILLBURN, N. Y. This well is located on the very edge of the sea marsh, about 2 miles south of Bald- win station on the Long Island Railroad (PI. II). It is 8 inches in diameter and the total depth determined by sounding is 288.6 feet. The water will, when piped up, rise about a foot above the surface of the ground. After an unsuccessful attempt to record the fluctuations here with a Friez gage, which gave no results because of the small amplitude of the fluctuations, the King gage used on the Queens County Water Company well near Hewlett was set up and the record obtained from July 17 to August 5. This record shows the most erratic fluctuations obtained on Long Island" (PI. V). In all the other records, while there are always many factors present, certain fluc- tuations can be definitely ascribed to temperature, atmospheric pressure, rainfall, pumping, or transmitted tides, but here either the curves produced by several of these factors have been so superposed that the character of each is thoroughly masked or new factors have been introduced. The most evident characteristic of these curves is the greater rapidity and abruptness in the fall of the water than in its rise. Abrupt drops of this character are known to be produced by changes in barometric pressure and by pumping. It will be noted in this case that these fluc- tuations are not represented in the barograph curve, and a comparison with the record from the 504-foot well at Lynbrook (PI. VI), in which the geologic condi- tions are very similar, shows no correspondence, although the Lynbrook well is clearly greatly affected by barometric changes. The nearest pumping stations are at Rockville Center and Freeport, and these are small village plants. At Rockville Center there were at this time four 8-inch wells, about 50 feet deep, and at Freeport 4 wells, about 35 feet deep. At Rockville Center about 150,000 gallons per day were pumped, and at Freeport about 100,000. These A lib 28 un, f 9-F( eld DOT assis tan US. GEOLOGICAL SURVEY >. 16. PL v 1903. JULY 17 JULY 18 JULY 19 JULY 28 JULY 29 JULY 30 AUG. 3 AUG. 4 AUG. 5 1 8N48M48N48M48N48 M 48N48M48N48M48N48 M 48N48M48N48M4SN48M TIDE CURVE AT NEW INLET ^ 3 AS PREDICTED BY l> 2 U.S. COAST AND GEODETIC SURVEY < j i [5 miles southeast of well.] F- < o "- -1 -2 \ i p / / , ( / \ \ / \ / \ K ^ \ 1 r\ r V ' M EA ./ A l/e A / \ / \ I \ \ 1 V, / / / \ / \ 1 i \ y y ' \ J / ^ V / \ \/ / \ / \ / V ' i v7 V V. / s / .' ! > BAROGRAPH CURVE AT o 3 °- 25 BRENTWOOD, N. Y. E „_ 30.00 - [22 miles northeast ot well] o • y recording gage I a. o z \ Q 6.5 AUTOGRAPHIC RECORD OF w § FLUCTUATIONS OF WATER LEVEL |2 IN A 289 FT. WELL u. o NEAR MILLBURN, N. Y. 8 .„ ,. i S 7 - 5 [Inverted curve] | £ From Friez self-recording gage w 8 fs Ai ' h r r\ 1ST / jjx r- ^ r~v- J A.^ r ; \ / ; \ J \ t f i] \ r m s 1 \ J \j ^ rl r j I THERMOGRAPH RECORD . gj 80 AT FLORAL PARK, N. Y. 5§ 70 [8 miles northwest of well.] Q 6 ° =n\ I S ■-. s V. \ l ^ V 0.00 RAINFALL RECORDS 0.06 AT FLORAL PARK AND BRENTWOOD. N. Y. / E 1.00 (Respectively 5 miles north o. and a 2 6 miles northeast of wells.) £ 1.50 From Fne2 self-recording gages to Explanation: All rains at Brentwood £ 2-00 are indicated with dotted lines and the word "Brentwood." All others 2.50 are at Floral Park 3.00 \ B.„„ 6' ; ' - uS \ 0-12 :,,, V ' B " u°2° \ : \ 1 ' ii 1 \: . PARTIAL RECORD OF FLUCTUATIONS OF WATER LEVEL IN A 289-FOOT WELL NEAR MILLBURN, N. Y. From a King gage in charge of Francis L. Whitney and F. D. Rathbun, field assistants. Well curve is inverted. OBSERVATIONS WITH SELF-RECORDING GAGES. 23 .--i tions are, respectively, L.9 and 2.4 miles from the Schreiber well (PI. [I). The Rockville Center pumping station is, moreover, nearer the deep Lynbrook well than il e Schreiber well, and the Lynbrook well shows no such fluctuations. It is there- fore believed that these fluctuations are not due to pumping. A third hypothesis is that tin- fluctuations arc largely tidal, and that they represent tlif complicated stresses resulting from the culmination of the tides at different times in the neighboring network of creeks and channels. The conditions near this well are regarded as quite favorable for complex tides, hut the resultant of these would he represented by a smooth curve, as is show n by the Long Beach well. The normal tidal curve in such narrow channels would, moreover, show a rapid rise and a gradual fall, while tlu> well curve, is just the reverse. The fluctuations in this well are. so far as known, unique. The geologic condi- tions are believed to correspond in a general way with those in the 368-foot well at Long Beach and the 504-foot well at Lynbrook, both in the same region, but the characteristics of the curves are entirely different and apparently not related to either. OBSERVATIONS AT LYNBROOK, X. Y. A station was established one-half mile west of Lynbrook (PL II). At this point there were two test wells, one 504 feet deep and the other 72 feet, belonging to the Queens County Water Company. Through the kindness of Mr. Franklin B. Lord, president, and Mr. Charles R. Bettes, chief engineer, these wells were covered with a shelter, and a third well, 14 feet deep, driven about 6 feet from the other two. This gave a shallow surface well, a "deep well" (comparable to many of those used at the Brooklyn waterworks pumping stations west of this point) which flowed at the surface for about half the time, and a very deep artesian well, all within a few feet of each other and away from the zone of influence of any jmmping station. About 15 feet from the wells there is a small ground -water fed brook, the bottom of which has an elevation of about 10 feet above sea level; the ground at the wells is 1 1 .3 feet above sea level, and the crest of the low swell, 1,000 feet to the west, about 20 feet (PI. II, p. 16). The surface material is yellow loam, ranging from a few inches to 3 feet thick, then rather coarse sand and gravel. No record was preserved of the strata penetrated in the 72- and 504-foot wells, but a new well sunk during the summer of 1904, about 300 feet west of this group of wells, gave the following section: Section of well of Queens County Water Company, one-half mile west of Lynbrook, N. Y. Tisbury: Feet. 1. Coarse yellow quartz sand; no erratic material 0-29 2. Light-gray sand 29-31 3. Same as No. 1 31-73 Cretaceous? : 4. Light-gray silty clay 73-89 5. Light-yellow medium sand; no erratic material 89-150 Cretaceous: 6. Fine to medium gray lignitic sand 150-158 7. Very fine black micaceous lignitiferous silt 158-200 8-9. Very fine dark-colored lignitiferous sand 200-228 10. Medium light-gray sand, with small amount of lignite 228-340 11. Dark-colored lignitiferous silty clay 340-363 12. Medium dirty-yellow sand, lignitic 363-403 13. Medium to coarse gray sand 403-536 The water in the 504-foot well, during the time of observation, stood from 0.8 foot to 2.2 feet above the surface; in the 72-foot well from 0.6 foot below to 0.5 foot above the surface, and in the 14-foot well from 0.6 foot below to 0.2 foot above the sur- 24 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. face. The water in the 14-foot well rose above the surface three or four times for periods of a few hours. The elevation of the water table under the low swell to the west was probably about 13 feet. Mr. Francis L. Whitney was placed in charge of the observations at this point, and King gages were installed on supports clamped to the well casings. The gages were maintained from June 25 to September 15, and some of the results are given in PL VI. These give a large amount of important material bearing not only on purely scientific problems, but on some of the live economic problems of the region. It will be noticed that as a whole the curves of these wells are parallel. This bears very directly on the old question of the source of the water in the deep wells on Long Island. It has long been a favorite hypothesis that in some mysterious way large quantities of water were introduced by great underground streams from the New England States, and this 504-foot well was one of the wells which were supposed to strike one of these streams. It has already been shown « that the source of the water in the deep wells on Long Island is from the rain that falls on the surface (see p. 10), and the really remarkable agreement of the general shape of these curves furnishes additional confirmation, pointing as it does to close interrelation and a common source. The behavior of these wells during rain storms shows that rain may affect the water level in wells in two ways, (1) without the water reaching the ground- water table, and (2) by actual infiltration and addition to the ground water. In both cases the effect is greatest in the shallow well. In the first case all wells commence to rise as soon as the rain begins, and rise abruptly, sometimes several inches. That this can not be due to actual infiltration is shown by its instantaneous character and by the fact that the water in the shallow 14-foot well, driven entirely in sand, rose above the surface of the ground four times under such circumstances. Such rises, moreover, produce no permanent deflection of the well curve. (See record for Aug. 18-22, PI. VI.) This sudden rise is due to a number of factors. In the first place, when the soil above the water table is filled with air the addition of water to the surface practi- cally seals the outlets for the air and the weight of the rain is transmitted by this confined air to the water table. The effect of such a transmission is to hasten the discharge of the water at the ground-water outlet, and so produce an immediate rise in the streams. In this manner rains which never reach the ground-water table and which do not contribute directly to stream flow may immediately produce a greatly increased stream discharge. It should be noted in this connection that the well always rose before the adjacent brook, although the brook might later reach a higher elevation. In the second case, when the water in the wells is elevated by the actual percola- tion of water, the water rises gradually and reaches its highest point several days or weeks after the rain, rather than in several minutes. In the case of the heavy rains which occurred on August 28, the 14-foot well reached its highest point before noon on the 29th, the 72-foot well at about 6 o'clock on the 29th, and the 504-foot well at noon on the 30th. There are three factors concerned in this last rise: (1) The instantaneous transmission of pressure due to weight of rain on the surface in the vicinity of the well; (2) actual percolation in the vicinity of the well, and (3) pro- gressive deformations resulting from the weight of the rain at more distant points. The rise in the deeper wells is wholly due to the first and third causes. The curve in this case is actually displaced and returns to its former position only gradually, instead of at once, as in the case described above. Barometric changes affect the 504-foot well most, but are occasionally perceptible in the 72-foot well. Temperature changes produce rhythmical daily fluctuations in the 14- and 72-foot wells; in the first the changes are very pronounced, amounting a Prof. Paper U. S. Geol. Survey No. 44, 1906, pp. 67-69. IIDCV1M- I'l'lllMJ \IMTI1 kJU'I I'.UV. « i » < ' ca OK 1 / / 1 1 B I- e"t wood c • 03 1 1 1 - 9,4- TTT.TTfyTTTATtOTCsl Aff TTTTT. WATT7.R, OBSERVATIONS WITH SELF-RECORDING GAGES. 25 t<> as much as I A inches a day. The 504-foot well showed a regular fluctuation with two high and two low waters a day. The fluctuation, however, Is not progressive, and so is not tidal. The curve from the 504-fool well shows agreal number of minor periodic oscilla- tions, but the time scale is not sufficiently great to study them satisfactorily. The most pronounced of the series has a period of about forty minutes. The 72-foo1 well occasionally shows well-marked secondary oscillations, with a period of approxi- mately eighty minutes. For a careful study of these, however, a much larger gage w ith a large time scale is demanded. OBSERVATIONS AT DOUtiLASTON, N. Y. In the winter of 1902 and 1903 a number of shallow wells were sunk along the base of hills, east of Alley Creek, and near "The Alley," an old settlement just south of Douglastoh, X. Y. (PI. Vll), for the Citizens' Water Supply Company. Six of these are flowing wells, and in the other two the water comes very near the surface. Through the kindness of Mr. Cord Meyer and his son, Mr. J. Edward Meyer, president and superintendent, respectively, of the Citizens' Water Supply Company, the flowing wells were piped up beyond the limit of flow and thus prepared for gaging. The relative elevation, depth, and head in these wells are shown in the following table: Elevations in welh of Citizens' Water Supply Company, at Douglaslon, N. Y. Elevation of surface above sea level. Depth of bottom of pipe below sea level. Average height above sea level to which water will rise if piped up. Feet. 17.2 20.2 10. 2 5 5 10.8 10.1 9.8 10.5 Feet. Feet. 17.2 Well No. 1 Well No. 2 20.5 25 28 25 39 35 30 17 19 9 Well No. 3 8.5 Well No. 4 18 Well No. 5 18 Well No. 6 .' 19 Well No. 7 17 Well No. 8 : 15 The strata encountered vary considerably; some of the wells penetrate nothing but sand and gravel, and in others clay beds of greater or less thickness are found. The water is derived from the adjacent hill mass, the height of the ground water in which determines the head in these wells. The tidal marsh to the west is a mass of soft black mud largely covered with a mat of growing vegetable matter, which is sufficiently firm to walk on, but which gives at every step. This surface mat of roots is often sufficiently tenacious to hold up when undermined by the small streams formed by the many springs that occur at the base of the hills, and these streams often flow through underground passages beneath the turf. The underlying mud or black ooze is over 10 feet thick in the upper end of the mud flat, and Mr. D. L. Van Nostrand states that, in driving piles for a dock at the bridge, the depth to "solid ground" was found to be 65 or 70 feet. The arte- sian pressure beneath this mud has caused the ground to rise in several places, with the resultant production of many small rapids (PI. VII). At a number of points near the upper end of the basin, where the mud is thin, the water has broken 26 FLUCTUATIONS OF THE WATEE LEVEL IN WELLS. through and produced low mud cones or mud volcanoes. a The drilling of the arte- sian wells and the fact that they were allowed to flow freely have perhaps in part relieved the pressure here, and during the three months the cones were observed they did not change materially, although on several occasions mud was seen rising from the craters and flowing down the sides. Three hundred feet north of the mud flat and on the east bank of Alley. Creek there is a lesser area of mud which is not covered at low tide. In this there is a marked mud flow, which is likewise probably connected with the artesian waters under discussion. Mr. Francis L. Whitney was placed in charge of the work here, and he prepared wooden shelter boxes covered with tarred paper. These were securely clamped to the top of the well pipes, which were steadied by means of guy lines. & A tide gage was established on the end of the crib of the drawbridge on the main turnpike. (PI. VII. ). The crib furnished a very good still box, and the locality is as near the wells as it was possible to get, for to the south the creek bed is uncovered at low tide. The equipment consisted of 3 Friez gages and 1 Mead gage. The Mead gage was placed on well No. 8 and furnished the only record running through the whole of the time of observation. One of the Friez gages was placed at the draw- bridge during the whole period, but from one cause and another no record was obtained before August 6, and after that time the record was not complete. By shifting the remaining gages records were obtained for a time from all the wells but No. 3. Some of the curves obtained from these observations are shown in PI. VIII. All these wells are clearly tidal, but when the question of the rate of propagation of tidal effect is considered many difficulties are encountered and the extreme com- plexity of the problem at once becomes evident. The curves, while broadly resem- bling each other, show many minor points of difference, which must be attributed to the varying shape of the tidal wave in the mud flat near the wells and the conse- quent complexity and variation of the stresses involved. Thus the records from wells 2 and 8 show that, while the relative amplitudes of the high tides agree perfectly and both show a tendency toward a double cusp at high tide, in well No. 2 the second cusp is characteristically greater, while in well No. 8 the first cusp is often the greater; compare curves from July 27 to 29. The low-tide curves also show marked differences; thus, in well No. 2 there is a continued fall until the tide turns, which it does sharply; in No. 8 there is a long period of stagnation and the curve is rounded when the rise begins. Evidently these curves are not readily comparable with the tide gage at the bridge nor with each other, for each represents the resultant of a different set of forces. The conditions for the production of such differences are very favorable. The semiliquid marsh mud yields readily to all pressure changes, how- ever slight; the liquidity of the mud varies greatly from point to point, and while the artesian water does not commonly escape through the mud covering it may, as shown by the mud cones, do so at any time, and such a point of relief would affect adjacent wells differently. Another factor making exact time comparisons difficult is the small scale of the records and the great amount of lost motion in the Mead gage. The Mead records show unquestionable time errors of one to two hours, and for this reason the end values are more important than the initial ones of each record. Where evident errors occur in the record of the Mead gage for well No. 8 they have been cor- rected as far as possible, and an attempt has been made to indicate on the diagram the various details affecting the time values so far as known. For this purpose the end of each of the original record sheets has been indicated on PI. VIII. « See Prof. Paper U. S. Geol. Survey No. 46, 1906, PI. XXVII, C. b An illustration of the gage box on well No. 4 will be found in Prof. Paper U. S. Geol. Survey No. 44, 1906, PI. XIV. U. S. GEOLOGICAL SURVEY WATER-SUPPLY PAPER NO. 166 PL. VII SKETCH MAP SHOWING LOCATION AND TOPOGRAPHIC SURROUNDINGS OF WELLS OF CITIZENS' WATER SUPPLY COMPANY NEAR DOUGLASTON, N. Y. Black dots indicate location of mud volcanoes. By A. C. Veatch, 1903. LONG ISLAND OBSERVATIONS. 27 OBSERVATIONS OB THE NEW YORK CITY COMMISSION 1 ON ADDITIONAL WATER SUPPLY.^ The Long Island division <>!' the commission on additional water supply under the direction of Mr. W. F. spear, division engineer, during the period from the middle of April to the hist of October, L903, made many observations on the water level in wells mi Long Island. In all aboul 1,200 wells were observed at intervals of from niie to three days by means of steel tapes titted with enp sounders. From these observations Mr. Spear endeavored to obtain the velocity of the downward capillary flow of the water on Long Island. Meteorological stations, equipped with self-recording instruments, were established at Brentwood and Floral Park ( I'l. I, p. 9). It was from these records that the thermograph, barograph, and rainfall curves shown on Pis. Ill, V, and VI were obtained. Mr. Spear likewise obtained from the records of the Brooklyn waterworks data regarding the fluctuations of the water in shallow wells on the south shore and the effect of pumping at Merrick and Agawam. a Spear, Walter E., Long Island .sources: Kept. Commission on Additional Water Supply for the City of New York, Nov. 30, 1903, New York, 1904, appendix 7, pp. 617-806 PART II. GENERAL DISCUSSION OF THE FLUCTUATIONS OF WATER LEVEL IN WELLS. CLASSIFICATION OF CAUSES. The vertical fluctuations of the ground-water table or the changes in the level of the water in wells may be grouped as follows: A. Fluctuations due to natural causes. I. Rainfall and evaporation. 1. Fluctuations not depending on single showers. a. Regular annual fluctuations. b. Irregular secular changes. 2. Fluctuations produced by single showers. a. By transmission of pressure without any actual addition to the ground water. b. By the actual addition of rain to the ground water. II. Barometric changes. III. Thermometric changes. 1. Fluctuation directly related to temperature. 2. Fluctuation inversely related to temperature. a. At the surface of the ground-water table, directly through temperature changes. b. In deeper zones, by pressure changes produced by fluctuations of the preceding class. IV. Fluctuations produced by adjacent bodies of surface water: Rivers, lakes, the ocean. 1. By changes in rate of ground-water discharge. 2. By seepage. 3. By plastic deformation due to varying loads. V. Fluctuations due to geologic changes. B. Fluctuations due to human agencies. 1. Settlement, deforestation, cultivation, drainage. 2. Irrigation. 3. Dams. 4. Underground water-supply developments. 5. Unequal loading. C. Fluctuations due to indeterminate causes. The relation between the fluctuations due to natural causes may be stated in this way: On the broad and irregular curves produced by the secular climatic and geo- logic changes are superposed the regular annual fluctuations, which are perhaps the most characteristic and important of the ground-water fluctuations due to natural causes; and on these, in turn, are superposed the simple rainfall, barometric, ther- mometric, tidal, and flood fluctuations. This complex curve, made up of many regu- lar and irregular elements, is further modified by human agencies. The cumulative effect of these human agencies is irregular and the result is to modify — indeed, often to largely alter — the character of the broad irregular curves produced by secular cli- matic and geologic changes. Yet some of these human modifications have a periodic value which, in the case of cultivation, for example, may greatly change the ampli- tude of the annual fluctuations, or, in the case of pumping or the change of water level behind a milldam, may give rise to rather regular daily fluctuations. 28 ao 1 r 1 ,,-,1 , ~~ \ • — \ r~ A r~ v L ^ ^ \ L> \j \j v — i r — S r \ — \ Vel 1 i> Jo 1 — \ < \ / \ / \ J d u \-J \J W V, v_ J -> hi) u 7 \ i n r 1 1 1 1 1 i i i i 1 1 1 1 1 1 i I 1 ' \ i ' i | 1 1 1 1 1 L rm E IN AC )JO INI NG c RE ,.11 EK ni ■ve Nn 8 1 ost mc tio n is ev de It original record sheets. u. s GEOLOGICAL SURVEY • :r- UPPLY NO. 155 PL. fill 1 1903 ! T» " i T^ 6 8 M ULY 29 | AUG. M 6 12 6~ | AUG. 1 [ *", 1 1 AU 6 3. 25 ■ AU< . 26 AUG A 6 1 2 1 » 6* U ?2 1 l AUG. 29 12 G | AUG. 30 | M 6 12 6 M AUG. 31 I SEPT. 1 | SEPT. 2 | SEPT. 3 | SEPT. 4 M 6 12 6 M 6 1? fi M B 19 r, M i; .o c u „ " ' ' ' ' ' 1 TT1 > I I I 1 1 ' " 1 ' 11 1 ' 1 II 1 n -3.6 -3.3 -3.0 -2.7 -ft- -A- / / f f\ I 7t Tide Curve A A ll A_ f] f \ - -1.5 -1.2 -0.6 0.0 0.3 1,5 2.1 I \ — — (i— |- ' "I _[ / , J u. s. Ge olo 1 S *y Bas i 1 ~T j : 1 T I ^T 1 it J t r - T T 1 1 T 7 T 7 T T r T ~ 3.3 S 3.6 o> 3.9 *, 4.5 ■- 5.1 O 5.4 > 5.7 6.3 l T - J T J 1 / / - 1- J / - u 1 J ! i V \J V T V T [1 1 / ___ _I I 1 Well No ? ^ 1 /-l / r— '"" 1 ^ r \ | 1 ' ,/ l/ J V I -J u /ri /^ /-f e ih?°-M\ '" Wei No H *r \_/~\ y~""\ /~~\ / — ^ r ~\ I f~ \ i~^ r^ ~ " j — i ^~\ '"y rA r * r~\ r~\ r~\ /_% /P* /~ ^ /~i /~v /'"v ^ \ ^~\ ^~\ J^7 V tf \j \f ^ ir " \j w \j / \ / ./ ,/ \y \7 w \ / \ /l\ / \y \y w \y ^ w - v - v - V, ■y/^.y^w^xj^v We 1 No 5 1 iri,*S^ n ~XP\~tXtZL ti UZtLttttXtXl^t^tl jiAfv^i J a -\I4tt * " X tEXt^Al * ^ ^ V >J V Well No 6 U | ^ 1 1 ' " i^ni^i n c D n&irzizTpZ * w^no. J ul ^Ig ^ tf Xfl^AtJLV JJ.XLJJ.JJ.JJ.JJ.JJ.XLJJ.JJ.JXU. lLlZlLXLZxLJXjliJ-lLlJ.±I JJ_JJ_±±JXJJ-JJ-JJ-JJ-J±J^JXJJ-JX.LLJjiLLJXjWjJ_JJ_^ PARTIAL RECORD OF FLUCTUATIONS OF WATER LEVEL IN WELLS NEAR DOUGLASTON, N. Y., AND OF TIDE IN ADJOINING CREEK. From gages in charge of Francis L. Whitney, field assistant. Curves are inverted. Curves for well No. 8 are from Mead gage . all others are from F Tie; : gages. At P°<£">££ * on wel1 curve No ' 8 loSl ' & s B in original records; this has been approximately corrected in this diagram. Short vertical lines indicate ends of original record sneets. FLUCTUATIONS OF WATER IN WELLS. 29 FLUCTUATIONS PRODUCED BY NATURAL CAUSES. RAINFALL AM) EVAPORATION. REGULAR ANNUAL FLUCTUATIONS. GENERAL CHARACTER \M> CAUSE. Woldfich, from a study of nine years' observations on a 19-foot well at Salzburg, con- cl ik led that tlic rise ami fall of tin* ground water stands in no relation whatever to tin: Fig. 9. — Yearly rainfall and water-level curves in shallow wells in middle Europe (alter Soyka). The curves are duplicated for a second year to facilitate comparisons. amount of rain, since with the same quantity of precipitation it at times rises, and again falls, and even with considerably increasing quantities of rain it often falls constantly." .. oVVoldrich, Johann Nepomuk, Mitt. d. Teehn. Klubs zu Salzburg, 1869, Heft 1, Zeitschrift d. Oterreichischen (Jesellschaft tvir Meteorologie, Bd. 4, 1869, pp. 273-279 Penck's Geographische Abhandlungen, Btl. 2, Heft 3. Wien, 1888, p. 23. 30 FLUCTUATIONS OF THE WATER LEVEL IINT WELLS. While so broad a statement is not entirely true for all localities in the North Tern perate Zone, yet it properly emphasizes the fact that the relation between the ground-water fluctuations and the rainfall is not the simple one which might be inferred ; from the statement that the rainfall is the source of the ground water. Observations at many points in the North Temperate Zone have shown that the Geneva, N. Y Bryn Mawr, Pa, Millburn, Long I$land s N. Y Michigan ans/ng } Mich. Geneva, N. Y. ('687- ss) Bryn Mawr.Pa. y (l886-95) Michiqan. Q878/0O-97) A nnArbor/M/c/?. (1885 -/90/J Mil/burn, Long Is/and^.Y. QC97-/903) Lansing, Mich. (1885-/300 Fig. 10.— Yearly rainfall and water-level curves in shallow wells in the United States. The curves are duplicated for a second year to facilitate comparisons. ground water fluctuates in a yearly period with a single maximum and minimum, and that this curve generally does not correspond with the rainfall curve (figs. 9, 10). Indeed, at Frankfurt, Bremen, Berlin, and Briinn, the highest point of the ground water is in the spring months at the time of least rainfall (fig. 9). The yearly FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 31 curves of the ground water are much more regular than the rainfall curve, ami on the whole in general shape they mosl resemble the annual temperature curve (fig. II i. The reason tor this difference is the Bimple one thai the fluctuations of the ground water depend not only <>n the absolute amount of the rainfall, l>ut on the quantity that reaches the zone of complete saturation, or the ground-water table, and ilic time consumed in so doing. The quantity is affected by many factors, among which arc the evaporation from the surface of the ground, the evaporation or transpiration from plants, the quantity retained in the soil above the zone of satu- ration, and the amount that runs directly off the surface without ever penetrating the ground. The time element is affected by the porosity and moisture content of the soil, the character of the covering, and to a greater or less extent by the height of the soil column. The general result is that the water is delivered gradually to the /one of complete saturation, and as the effects of single rains are thus generally — too 90 ^ so *> * % 60 i*> 7 fa // at t A 'ac let}. iht a A, A \l / N 1 A 7 ' s r V / ^_ — V J V 6/ vo r?d H'(7 ter at ^ *ry n Mc f w r / \ s \ / \ V 1 r / \ \ 1 \ / \ \ Pig. 11.— Annual curves based upon mean monthly averages of ground-water level at Bryn Mawr, Pa., a and temperature and rainfall at Philadelphia, Pa.,?> for the period 1886-1895, showing the general resemblance of the ground-water and temperature curves. The exact agreement of the maximum ground- water level and the maximum temperature is unusual. minified and often entirely blotted out, a relatively smooth curve results (PI. IX). This yearly period of the ground water is largely due to the periodic character of the evaporation, including plant transpiration. This depends on the temperature, and the net result for the year is a simple curve of the same shape as the mean tempera- ture curve, although inversely related to it, hence the general resemblance of the yearly well curve to the temperature curve. Were the rain uniform throughout the year, and were there no lag due to transmission or unmelted snow, the maximum ground-water level would occur at the time of the minimum temperature and satu- ration deficit of the atmosphere, or, in the North Temperate Zone, in January. The effect of the irregular distribution of the rainfall is to change the time of the occur- rence of this maximum. A moderate excess of rain in the summer, such as occurs "Observations of W. S. Auchincloss; Waters within the Earth and Laws of Rainflow, 1897. ^Observations of U. S. Weather Bureau; Annual Summary Pennsylvania Climate and Crop Service. 32 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. at Frankfurt-am -Main, causes the maximum to advance to March, while at Bremen, Berlin, and Briinn, where the difference between the summer and winter rainfall is progressively greater, in the order named, the maxima occur respectively in March, April; and May (fig. 9). The extremely great summer precipitation at Munich and Salzburg, together with the low rainfall in January, causes the maximum at those places to advance to July and August. ■ In this connection the observations of (1) Dickinson and Evans, (2) Greaves, and (3) Lawes and Gilbert, near London, are of great interest. All of these observers endeavored to determine the amount of rain actually contributed to the ground Fig. 12. — Results of English percolation experiments. In the Dickinson and Evans experiments the gages were buried in the ground; one was filled with ordinary Hertfordshire soil (a sandy, grav- elly loam) and covered with sod; the other was filled with chalk and covered with a thin layer of soil and sod. In the Lawes and Gilbert observations columns of a rather heavy loam with clay subsoil in their natural state of consolidation were built in with brick and cement; no veg- etation was allowed to grow on the gages, which were surrounded by meadow land. Curves are based on the monthly averages from September 1, 1870, to August 31, 1902. w T ater. Each used vessels with impervious sides and pervious bottoms, sunk level with the surface of the ground. The water percolating through the soil columns was collected and compared with the yield of the adjacent rain gages. In the case of the Dickinson and Evans and the Greaves experiments the boxes w r ere filled with mate- rial supposed to represent the average soil of the region, in both cases a sandy loam. In the Lawes and Gilbert experiments actual blocks of soil were undermined and the results represent the amount of rainfall passing through a heavy loam with a clay subsoil in its natural condition of consolidation, but not covered with vegetation. The average results obtained are given in the following table and are partially shown in a graphic manner in fig. 12. U. S. GEOLOGICAL SURVEY WATER-SUPPLY PAPEF NO . 1 05 PL. IX ! Meiers above sea 32a 310. 300. 290 280 270. S W 260- ; : Water lev P | i„ r . Water level March 1 No. 8 ha River mer BrooK No. 10 ===— E« I'KUM s - " Residual mass curve showing excess rainfall in millimeters July 1886 Dec. 1890. at Hohe Warte station, Vienna. 'lalirl/uchnikr k.t. Caauil-AuUtt ICr s KtttorolOgtl und EnlinagnetiMijus. - ^ ^ 50 ~ o 1 p2 •• '" J V i ' -, r V V i -ioo ; ui Meters above sea Wall No 112. Suifar« elevation 829.99 Bottom of well 276. 4»j \. 290 289 288 286 285 284 283 282 231 280 -/ v> \ Well No. 1. Surface elevation 208.721 Bottom of wall 273.50J Well No. 43. Surface elevation 206.881 n y 276 ° 275 a < 273 272 a 271 o 270 " 269 „ 268 a 267 ■ E Lett ha , n s»r Kat Eels orl "" ' -' ~^ <£ y Bottom of well 268.12 >■£ Well No. 3. Surface elevation <* Bottom ol well 266. 66 j &/ \ .a Bottom of well 264, SN. Woll N...1. Woil.i.-e elt-vdiioii 276 *1\ Bolton, of well 265.24 Well No.fi. buifiue elevation 273. Tt* Well II. Suifa.-e .-!-. at. on 272." 4 *\ "\ .. i- «gs ~"~ ~ 3 ^ Bmt..m of well W e ll No.li. Su.inreeleial.uii Bottom of well ■.■■'." ;i. 26 i. 'Ml ""^t ,^25 WaU No. 8. Surface elevation V^.?&\ ~~^ E 265 264 263 » -•■= ^ j _fl , Fisch* River. Surftce elevation 263.Qsl Bottom of well 261. 75J Fischa " 1 RWj Hammer Brook. Surface elevation 262.321 -*Ld r B ook 261 260 259 258 257 J g § a o o O a 5 s 2 | s s ? E § = C3 3 £ o o a 1 886 1 890 This illustrates regular annual fluctuations, irregular seculars that the well fluctuations are generally independent of single rains. It also shows concentnrJti me of occurrence of the yearly maximum and minimum. ( From Bericht des Ausschusses fur c . i» U. l» 2*0 2J . Water level Aug. - Sept, 1892 J — Water level March 1892 a PROFILE SHOWING LOCATION ANO DEPTH OF WELLS ! | i " * » "-=Ss- -|V- N V- fv It fA ps A \ X J s 1 S \ s. \. *\ 1 'S V f ' VJ A V l K ■/ \ l^ j. V -- \ •■*. f- M 1 .,..; \ \ J / .b«:'.H . s» V / \ / \ ) S 4. s J \ i \ ± N \ / ' v K', \ N 1 s; / \ 1 V_J / / V I \ > A / K \ V \ i \ X' / \ »i ~ if \ \ 1 i \ \ . . - I EVAPORATION. 33 VJS 33 S 03 S o » . S S - ■35 3- •saqoui 09 >-"I'"! 01 ■saq.iin or. •daap s.)i|.nu 09 ,l.,.ip s.lq.HIl Of- •daap saqout 0o 'T-OtU-O^l !: 2 •-> OS «5 OS 00 to : i ? i : i : -. — :i / ic -r i- -i co ci o> ct to -r ci co -r I- o> < K K ci N m ci co io |> Vrsoaso-^ascNoocoocico : i — I - i -' :-' - i r i co CO CO 1^ CO J- oo l~- -T ci m ci m :i « g h 10 O i-C CI to OS O O CO I- w l~ CO as -r as ic t io co io t~ to o o '.' - 1 ia o o -f -T 1 iO H i-l OS ^ i> SJ O lO O to lO UC to to X I~ M i— O fl o n tcOO^ClXi-iifCi-HO^I^ie Jg 00 -^* 00 lO *<1* l£ CO CO CR 00 r-l O «J iH I-! " ' ri « ci b i-- j. c io n n to lo m h ^ t-t oo to ao o oi as i^ co CI r-l tO il to ? CI i-i ^ ^ CI CI CI CS CI CO CI CI •pa«S qSnoiqj 3 II [ | 8 1 I) J J 8 d imu jo a.Siqua.j.ia o t— •< X' CM to CI 0C to t-i O to CO O CO Jgl-li-li-lrMTHrtlHtNi-ttNCNiH SSIUS qjIAV P8J9AO0 ui\!(i[ ai'uiin i.i..| !: qSnoaqj uoijiqoajaj OOOOiCiOiOiOOOOO "-C "'" iC "~ CI CI CI CI »C iC iC »o CNt~t-ClrHr^rHi-lCNt^CICN IlBjm'Ba •avai cij imijiq -o.u.i'l J'ioSi:i'u.i.).l.\[ •ssiu.tf qifAV paj3 -AO0 >1|FI[.I J.l.i.l g qSuojqi m iijiq' i>.i.»- E a E a I ^ Hi? eg cd » >. iC ci •o ii,^ --« ssfiis o , ■ v ■: ■ 3 i.= cj So* °31 . - OJS 1 1 "3 oS > ■;> .. s|| ■ E ^ 1 "T.-o CM p^tli r x > '5^b ^^S 34 FLUCTUATIONS OF THE WATEE LEVEL IN WELLS. Of these the Lawes and Gilbert results are perhaps of the greatest value, because they more nearly represent the normal conditions, and they extend through a sufii- cientlydong period to obliterate temporary variations. While the quantitative values obtained from these experiments differ, the qualitative results, as shown by fig. 12, are essentially the same. All except the hare sand give curves of the same character as those obtained from the actual observation of ground-water fluctuations. The low percentage of water which passes through the soil in summer is emphasized by all, as is also the greater contribution during the winter months. Even in the bare sand, where the water sinks at once and so loses little by evaporation, a downward tend- ency of the curve during the summer months is evident. The effect of the heavy precipitation in the fall is particularly evident in the Lawes and Gilbert results (fig. 12), where it clearly hastened the time of occurrence of the maximum ground- water percolation by about two months. EFFECT OF DEPTH OF SOIL ABOVE ZONE OF COMPLETE SATURATION ON TIME OF OCCUR- RENCE OF YEARLY MAXIMUM AND MINIMUM OF GROUND-WATER LEVEL. It has been suggested that the soil above the ground- water table tends to destroy the effect of single rains by causing the water to be delivered gradually to the zone of complete saturation, whose upper surface is the water table. In the case of the English experiments at Rothamsted (Lawes and Gilbert) and Hemel Hempstead (Dickinson and Evans) the maximum percolation occurs from November to January, yet the wator in the wells in that region, while commencing to rise about December, did not reach its maximum elevation until March, a a delay of about three months. Yet in any attempt to calculate the rate of percolation from these data two difficul- ties are encountered. In the first place the yearly maximum occurred at about the same time in this region in wells of all depths, and, furthermore, the Rothamsted results (fig. 12) show no very important difference in the time in which the water is discharged through 20, 40, and 60 inches of soil so far as it relates to the time of occurrence of the yearly maximum. In the second place the underground water is in motion, a certain amount is discharged at all times, and the amount increases with the head. The case is not, therefore, the simple one where water is caught in a measuring tube, as in the percolation experiments above described, but the water must reach the ground-water table at a rate greater than the rate of the outflow, else no rise will take place. At Wiener Neustadt, Austria, a similar relation has been demonstrated by the observations made between 1883 and 1895, in connection with the Wiener Neustadt deep-well project for the supply of Vienna. 6 These wells are in a valley filling of fluvio-glacial material, somewhat irregular in character, which Suess describes as a series of old deltas. c The wells extend from Fisha River, a spring-fed stream, south- westerly along the Southern Railway. The land and the ground-water table beneath both rise gradually in this direction, but while the water table is at the surface of the ground at Fisha River, 6£ miles south, at St. Agyden, it is from 140 to 170 feet from the surface, the exact depth depending on the time of year. (See section at top of PI. IX. ) The curves obtained from this series of wells, extending roughly at right angles to the slope of the water plane, are entirely concentric, and the maxi- mum and minimum occur at the same time, irrespective of depth of soil above the ground-water table. It would seem to follow from these data that no very satisfac- tory determination of the rate of downward percolation can be made from the rela- tion of the time of greatest precipitation, or percolation, to the time of maximum aClutterbuck, James. Min. Proe. Inst. Civil Eng., vol. 2, 1842, p. 156. bBericht des Ausschusses fur die Wasserversorgung Wiens: Osterreichischer Ingenieur- und Architekten-Verein, 1895. clbid., p. 32; see also Bericht iiber die Erfolge der Wiener Wasserleitungs-Commission, 1864; Karrer, F., Geologie der Franz Josephs-Quellenleitung: Abhandlungen der K.-k. geologischen Reichs,an- stalt, 1877. FLUCTUATIONS DTK TO RAINFALL AND EVAPORATION. 35 ground-water level. The rise of the water is nol determined by the simple delivery of water to the zone of complete saturation, bul by the relation of the water so delivered to the rate of outflow. If the water is lowering, a certain amount is con- sumed in checking thai tendency, and only the excess over the outflow IS available for raising the ground-water level. Moreover, in several carefully observed instances, the depth of the soil above the ground water has been shown to have no effect on the time Of occurrence of the yearly maximum. The short-period observations on Long Island, New York, during the summer of L903, however, gave quite different results from those obtained at Wiener Neustadt. The conditions do not appear to he essentially different ; the glacial sands and gravels of the south plain of Long Island slope gradually to the ocean and in a similar way the valley glacial gravels of Wiener Neustadt slope to Fisha River, and there is apparently no great difference in the irregularity and complexity of the bedding. The Wiener Neustadt or Steinfeld Valley, it is true, is traversed by a large river, the Leitha, whose stages depend on the conditions affecting its headwaters in the moun- tains, but, observations have shown that this stream, because of the silted character of its bed, affects only a few wells in its immediate vicinity, and is not to be regarded as a disturbing factor. (Compare the river stages with well curves on PL IX.) On Long Island the measurements in charge of Mr. W alter E. Spear, department engi- neer of the commission on additional water supply/' showed that, during the summer of 1903, the highest stage of the ground water occurred, as a rule, earlier in the shal- low than in the deeper wells (tig. 13). Where the water level was less than 20 feet from the surface the highest stage of the ground water occurred in April, while wdiere the water level was 60 to 75 feet below the surface it did not occur until August. The increase of the retardation was not always uniform. Thus the highest water in a 35-foot well near Jamaica (No. 551) occurred in May, while in a well of the same depth near Deer Park (No. 388) it did not occur until August, although in a well near Hicksville (No. 237), of about the same depth, the maximum occurred in May. Whether this irregularity is typical or is only a result of the rather peculiar season in which the measurements were made could be determined only by observations covering a period of years, instead of months. It should, however, be stated in this connection that along the south shore, where the Brooklyn water department has observed shallow wells for several years, the curve for 1903 is not greatly different from that of preceding years, indicating, so far as the shallow wells are concerned, that the year is not to be regarded as an abnormal one (fig. 15, p. 39). On the other hand, the results are so at variance with the thirteen years' observations at Wiener Neustadt, which apparently cover similar conditions, that further confirmation of these Long Island results, by additional observations, is needed before any conclu- sions can be drawn. Certainly the Wiener Neustadt data indicate that the depth of the soil above the ground-water table is of no importance in determining the time of occurrence of the maximum ground-water level. On the other Hand, the Long Island observations suggest that a difference in thickness of 60 feet may delay the time of the occurrence of the maximum level four months. The curves showing the result of the Long Island work indicate further that, in the soil in question, single showers frequently produce yery definite effects in shallow wells, and that such effects become less as the depth of unsaturated material above the water table increases. Indeed, in the wells where the water is 30 or 40 feet below the ground, the curves are relatively smooth or the variations bear no evident relation to the rainfall. 6 Spear has attempted to trace the time of rise, due to given showers, from the shallow through the deeper wells, and so determine the rate of a Long Island sources: Rept. New York City Commission on Additional Water Supply, 1904, appen- dix 7, PI. IV, incorrectly numbered PI. VI, p. 792. 6Many of the wells observed by the commission were open, dug wells, which were in use, and the minor fluctuations are partially due to this cause, as well as to barometric and thermometric changes. 36 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 37 downward percolation or downward capillary ll<>\\ . There are certain difficulties in the way of determining the value soughl in this manner. In the firsl place, the fluctuations in the shallow wells can not [>e satisfactorily correlated with those in the deep ones, and the only line which can he followed through the diagram prepared by spear is the time of maximum ground water, which, as indicated above, in this region during the time of observation, in general lags proportionally to the depth. This gives a fairly regular curve and the remaining curves have been inferred on either side of this one. The yearly maximum, however, can scarcely be attributed to a single rain, bul represents rather the culmination of a whole series of events, and hence can nol be used as a basis of such a calculation. In the case of regions like Wiener Neustadt it is clear that the results from calcu- lations of this character would have no meaning, and, indeed, what do the values really represent onLong Island? IRREGULAR SECULAR FLUCTUATIONS. The observations ol Dickinson and Evans and of Lawes and Gilbert with percola- tion gages developed the fact that, as a rule, not only did more water percolate in wet than in dry years, but that the percentage of rain water which passed through the soil columns was usually much greater in wet years than in dry ones. Thus while the average yearly percolation of 1870-1902 at Rothamsted (Lawes and Gil- bert) was 4!» per cent of the yearly rainfall ot 28 inches, in 2222?2S2?5 the year 1878-79, when 40.2 inches of rain fell, 61 percent of the rain water passed through a soil column 60 inches high, and in the year 1S77-78, with a rainfall of 18.2 inches, the percolation was but 36 per cent (see p. 47.) The general tendency — al- though there are exceptional cases, such as recorded at Hemel Hempstead (Dickinson •and Evans) in 1858-59, when, with a rainfall of 28 inches, 2 inches more than the annual average, the percolation was but one-third of 1 per cent instead of the usual 27 per cent — is for the small differ- ences in the annual rainfall to have a rather magnified value in the ground-water fluctua- tions. The yearly variations of the rainfall are generally pro- gressive over rather long periods (fig. 14), and corresponding broad, irregular vari- ations of the ground-water level are produced. On Long Island the shallow wells observed by the Brooklyn waterworks show, besides the annual fluctuations, secular variations corresponding in general with those of the rainfall (fig. 15). Thus the lowest point in both curves is in 1901 and the highest in 1899. Many differences are, however, to be noted between the two curves. The annual curve, though it may be slightly modified, persistently recurs, whatever the rainfall. Note in this connection the regular downward course of the ground water in the latter part of Fig. 14. — Residual-mass curves of rainfall for Long Island, N.Y., Newark, N. J., and Philadelphia, Pa. (After Spear, 1904.) These curves show the cumulative excess or defi- ciency of the total actual rainfall over the total mean rainfall for the periods under consideration, ami as these excess or deficiency values are those which determine the long-period rise and fall of the ground water they indicate the general character of the secular fluctuation of the ground water occurring at these points. 38 FLUCTUATIONS OF THE WATEB LEVEL IN WELLS. 1898 and 1903, when the rainfall curve is rising, and the appearance in 1900 of a typical yearly curve when the rainfall curve falls rather regularly from the spring of 1899 to 1901. Similarly, in the Wiener Neustadt ohservations (PI. IX, p. 62) the secular vari- ations of the ground-water level broadly follow the variations of the rainfall. The observations of Auchincloss a on a well at Bryn Mawr, Pa., which have been plotted by Spear b in connection with the yearly rainfall and temperature curves, like- wise show pronounced annual and secular fluctuations. Here, however, the secular fluctuations of the ground water, while broadly resembling the rainfall variations, show some points of difference. Thus, the minimum of the secular curve of the ground water is in 1885-86, while the minimum rainfall is in 1887, and the general shape of the two curves for 1886-87 and 1888 is by no means parallel. The positions of the maxima, however, agree closely, and there is a general falling off in both curves from 1889 to 1893-94. Though the extreme rains of the latter part of 1889 tem- porarily obliterated the annual curve, it quickly reasserted itself. In general it may be said that irregularities in the yearly curve are due to irregu- larities in the rainfall occurring in the same year. AMOUNT OF ANNUAL AND SECULAR FLUCTUATION. The size of the annual fluctuations depends principally upon (1) the percentage of rainfall reaching the ground water; (2) the amount of free pore space of the strata in the zone affected by the fluctuations, and (3) the relation of the ground-water table to the topography of the region involved. It is relatively self-evident that, where a single well is considered, the range of the yearly fluctuations will vary with the first factor, and that in general the same amount of infiltration will produce a greater fluctuation where the pore space is small than where it is large. It does not, however, follow that in a given region, in beds of the same porosity, the same annual rainfall under the same climatic conditions will produce the same results. Observations have shown that whatever the rainfall or porosity, provided the latter be reasonably constant in the area under considera- tion, the annual fluctuations approach zero at the point of discharge and tend to reg- ularly increase in magnitude from that point to the ground-water divide. c Thus at Wiener Neustadt (PI. IX, p. 62), near the ground-water discharge into Fisha River, where the depth to the ground-water table is about 5 feet, the yearly fluctuation is 3 to 4 feet, while at St. Agyden station, where the water plane is about 150 feet from the surface, the fluctuation is 25 to 30 feet, and the fluctuations in the intervening wells are proportional to their position between these two points. On Long Island the annual fluctuation 2 miles from the shore, at Millburn (figs. 13, 15), is 22 inches, while at the ground-water divide, 8 to 9 miles from the south shore, the fluctuation is about 10 feet. A few observations regarding the amount of the yearly fluctuation at different points have been collected in the table following. Many of these points of observation are located near the points of discharge, and the values as a whole are to be regarded as low. In records for but a few years it is evidently impossible to separate the annual from the secular fluctuations. When, however, the observations cover a considerable period, it is possible to obtain a value for the secular fluctuation. This equals the total range less the average yearly fluctuation. A few such values are given in the table on page 40. aAuchincloss, W. S., Waters within the Earth and Laws of Rainflow, Philadelphia, 1897, p. 9. h Spear, Walter E., Rept. Commission on Additional Water Supply lor the City of New York, 1904, appendix 7, fig. 45, p. 822. c The cross section lrom Watford to the Chiltern Hills midway between Colne and Gade rivers, which accompanies Clutterbuck's discussion of the " Periodic Alternations of the Chalk Water Level under London" (Min.Proc. Inst. Civil Eng., vol.9, 1850, PL VI), is a most excellent diagrammatic illustration of this principle. FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 39 Depfh below surface of ground in feet a ro e •" - 40 FLUCTUATIONS OB' THE WATER LEVEL IN WELLS. i^fi M ,Q O cd C ,U - 5- & a 5- - oft r/i ■SoS CS o3j»i S3 gooi fl +3>H £ R R 2 6 d 6 6 6 $ h ft ft ft ft ft * * > e. s-a °os ^ fe a" •2S3.P •to ft oo O OS O CT> ffl CO Ol OO io oo io to 00 1^ OO 00 lOOOOOCOCOiC-cKOOOO OOCOOOCCXX^OCOOQOOOOOOO O00000000000O000000000O000 I 1 I I I I C2 CM lO CO CD CO CD l-~ CO X^ t~- t-- OOOOOOOOOOCOOOOOOOOOGOOO O CD CO CO CD to F- t- I> t-- 00 OS g be tibaS oS . S3 C"S 2 -Jj oiqd O O id CO ffl CM O 9 •* N N -ct< CO t^ £~ r-H i-H T-t T* 00 ilfss _ o en co as cm oo i-H CM i-H CO i-l ^ CD <4h I CO . ° - COS tv -5 £ .g g a; oi ft ft & > ^ 5 OS o o g.s ti ti o-d • O CJ : ^ : : Q (JO'S 'Sol f» ft 3 a> <1 bis ^ oS 3. OS S3 c « -^0Sg co co -* tH CO CO iC Tt< ^V CN CM- CN CP cu is £ r^ CO O CB O Of; bo <1 .£f .S J | .2 fl bo R 3 S 6 S tpd & 8 W H M En pq ft ft W FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 41 o c : : :_=.-oc 3 « fi s. 5 3 z f 3 P 3 3 3 O W O J O H? (J X. © © o o o o 3 x 5E x i ~ oc x co f. r X ■i -i r- r- i- CD I- 1- / do co oo a go tc / * A si .A < / / SO 'O iC CO OS r-j r-l 00 i-H 00 OS O to H ri (i ti OOOOOOCOOOO -r i^ o co ~ i- :: x co to i-i oo -r co -H .-. i-H lO l> CN .— . OS OS 00 T--I ■ CO 0-1 t-H o .-h as co o o i> o o © CO CO CN CO CO CO 3 -' ;g OS ''- ~ CO ~ o> P. WO 3 c P. . o3 §«J cc w 1-1.14,0 £%< S2 &£ 3 l " -* •^2 as 2*2 a j ■/ 3ft tp » . a.-- 02 ~t- i-! S 1 • ?. °o >JA r-i is *» t- - — :/ ^^llp. H ft oo t^ •- y - aS.2 r-.ci.C5 0)1? bo— OS*"* £3 -^ 2 c< o — r-.' "f oi _7 r o ■ a h ci-O £■52^ ~ g bjo.. ■ be - o t C OIT3 ££ o> o ™ °5 § S flg . ^ ." ,— *G ^^o£..^ > x — • m - Crl a n tg a • » c A cJ'-'S bo : a3 ft CO a) S -i g? 2co r 3 S'go,;^ K O , rt S o "" • S-'^ S co-rio C — - --^ c£ — Maximum ra Rept. New Y Ibid., PI. VI Ann. Repts. Auchincloss, Emery, Fran Soyka, Schw Ibid., p. 34. [bid., pp. 37- Ibid., pp. ii- Ibid., pp. 48- Whitley, He Evans, Sir .b Lucas, Josep Clutterbuck, Extreme ran Min. Proc. Ii e .o « •e w v, ct« ■"• * — a s o a, w ~ pq »i. 3 oT -• o M e rl^^CCjicft o> a> 0)^^^" — 53 W cZ J O O O 42 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. FLUCTUATIONS DUE TO SINGLE SHOWERS. In the foregoing consideration of the relation of the rainfall to secular and annual fluctuations, the most important factor was clearly the amount of water actually contributed to the zone of complete saturation, or the ground water. « In these cases the water table rises or falls because the amount of water received is greater or less than the. outflow. In the consideration of single showers, however, it is found that another factor is of as great or even greater importance. Single showers may affect the water level in a well in two ways: (1) By transmission of pressure without any actual addition to the ground water — indeed, in many cases the elevation of the water in the well is accompanied by an actual depression of the ground-water table; (2) by an actual contribution to the ground water whereby the level of the water table is raised. FLUCTUATIONS PRODUCED BY SHOWERS BY TRANSMITTED PRESSURE WITHOUT ANY INCREASE IN THE GROUND WATER. King observed at Madison, Wis., b that the water often rose in wells very suddenly and sharply during summer rains, when an investigation of the soil showed that it was dry beneath the surface covering wet by the showers. Similar occurrences were recorded at Lynbrook, N. Y., where in a 2-inch well 14 feet deep every rain — during the period of observation, July 17 to September 10 — was recorded and its duration and complexity indicated. Many of these fluctuations produce no perma- nent deflection of the ground-water curve and the evidence that no water from several of these rains reached the ground water is regarded as conclusive (PL VI; also note particularly the behavior of the shallow well August 19-21 and the fluctu- ations produced by rains of July 20, 22, 30, and August 6; the two successive showers of August 6-7 are particularly noteworthy). In the case of wells which are not separated from the main water table by imper- vious layers and in which the water is not under artesian pressure, this is due largely to the hydrostatic transmission of pressure by means of the soil air. When the rain strikes the surface it closes the superficial soil pores or interstices and thus confined and compresses the air in the soil between the surface and the ground-water table. The weight of the rain is thus transmitted to the ground-water table, and the extra head so developed raises the water in wells and increases the discharge at the ground-water outlets. The effect on the stream flow is very analogous to the increase produced by a lowering of the barometric pressure. It is thus possible for a rain to produce instantly a change in the water level in wells and an increase in the ground- water outflow without contributing a drop to the ground water. This has an impor- tant bearing on the calculation of "flood flows" from ground- water-fed streams, for it is evident that in this manner a decided rise in the stream may be produced by a rain from which there is no direct run-off and which does not reach the ground-water table. Two other factors may be involved in the production of the change in level during rains: (1) An actual elastic compression or plastic deformation of the soil, and (2) a change in the capillary conditions. King observed in a shallow well near a railroad track that the passage of a freight train caused a quick rise and fall of the water in the well (p. 75). Apparently the weight of the train compressed the earth and by decreas- ing the pore space caused the water to rise. The weight of the rain might have a similar effect. Under this hypothesis the water would rise on the addition of the rain and gradually fall on its removal by evaporation. aGround water as here used does not include the hygroscopic and capillary water above the water table, or zone of complete saturation. For the purpose of this discussion, water is not considered "ground water" until it reaches the water table. bBull. U. S. Weather Bureau No. 5, 1892, pp. 20, 72-73. FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 43 Regarding the second hypothesis it is well known thai changes in the surface con- ditions greatly affecl the capillary action <>f the soil. At Purdue University il was found thai the addition of a thin layer of soil to the Burface of a lysimeter caused an immediate discharge of water." Tins was attributed t<> a change in the capillary conditions, and it has been suggested that the wetting of the ground surface would produce a similar effect. King, however, observed '' thai a moderate wetting of the surface tended t<> increase the upward percolation, and the effect of wetting the sur- face at Lynbrook would therefore he to diminish rather than increase the rise due to rains which do not contribute to the ground water. At the Colorado experiment station lleadden has observed'' that light rains during dry periods produce a comparatively great increase in the height of the water plane, while in intervals of abundant moisture, when the soil is wet, rains of this character do not produce such an increase; moderate rains are here sometimes accompanied by temporary depressions of the water plane. These observations may be explained on the basis that the soil air is the principal factor and that when the soil is very moist there is so little soil air that no effect is possible with slight showers. The cause of the temporary depression after moderate rains is not evident unless the con- ditions are unfavorable for the transmission of pressure by the soil air, but are such that the increased upward capillary action resulting from the moistening of the sur- face is sufficient to perceptibly decrease the ground water. Where there is an impervious layer between the water-bearing strata and the local ground-water surface, and where there is artesian pressure, the added weight due to the rainfall in all cases acts directly. In case the rain is uniform over a con- siderable area this pressure may be regarded as acting on an elastic body and the same character of results is to be expected in both deep and shallow wells. Thus in the Lynbrook wells on July 22 and August 25 all wells show a sharp vertical rise (PI. VI). On July 22 the rise started in the 14-foot well at 10 p. m., in the 72-foot well at 10.25, and in the 504-foot well at 10.34; on August 25 a somewhat similar lag is noted, the 14-foot well rising at 4.10 p. m., the 72-foot well at 4.20, and the 504- foot well at 4.24. The cause of this lag is not fully apparent. With a direct trans- mission of pressure such as the curve indicates no lag is to be expected. It may be that in this case the soil is to be regarded as having some of the plastic characters shown in other cases. On the other hand, when the rainfall is unequally distributed in time and amount a plastic deformation may result, due to unequal loading, which will give rise to different results in wells of different depths. In the shallow well the zone of influ- ence is relatively limited and the condition in this area may be regarded as fairly uniform. The result is therefore immediate and abrupt, as in the first case. In the deeper wells, however, the increasing zones of influence bring in more factors, which, arriving progressively from different sources, tend to produce a more and more gradual change. Thus, in the Lynbrook wells there were on August 6-7, 11, and 20 abrupt changes in the 14-foot well and a more gradual one in the 72- and 504-foot wells. This plastic deformation in the surficial beds, produced by varying load and the response of the w r ater to it, throws some light on the extreme complexity of the fluctuation recorded in the wells, for it suggests that variation in load, from whatever cause, will produce corresponding fluctuations. The water level in deep wells where an artesian head is developed may thus be very sensitive to local conditions, the effect of local rainfall and of the yearly fluctuations of the local ground-water level being felt to a greater or less degree by transmitted pressure in the deeper zones. aSecond Ann. Rept. Indiana Expt. Station, 18S9-90, pp. 32-33. b Seventh Ann. Kept. Wisconsin Agrie. Expt. Station, 1890, p. 135. cHeadden, William P., A soil study, pt. 4, The ground water: Bull. Colorado Agric. Expt. Station No. 72, 1902. 44 FLUCTUATION'S OF THE WATEE LEVEL IK WELLS. FLUCTUATION OF THE GROUND-WATER LEVEL RESULTING FROM SINGLE SHOWERS, BY ACTUAL PERCOLATION. The fluctuations produced by direct percolation are of a much less abrupt char- acter than those just described; indeed, it is usually the case that the water is deliv- ered so gradually to the water table that no change is noticed. Only in the shallow wells in coarse material can these fluctuations be identified, except in the cases of extraordinary rains, when the result is an irregularity of greater or less importance on the regular annual ground-water curve. On Long Island the shallow wells near the south shore are affected by most of the important rains, although part of the fluctuations recorded are of the character just described. (See figs. 13, 15.) This is due to the coarseness of the surficial mate- rial and to the nearness of the water table to the surface. In the wells in which the ground water is farther from the surface the effect of any rain can not be positively identified. In the Wiener Neustadt records the effect of single rains is entirely obliterated (PL IX), and in long observations of the chalk waters of England the general rule, to which, of course, there are exceptions where large underground caverns are concerned, is that the water is delivered very gradually to the ground- water table. PERCENTAGE OF RAINFALL CONTRIBUTED TO THE GROTTND WATER. METHODS OF ESTIMATION. In connection with this discussion of the fluctuation of the water level it may not be inappropriate to take up the allied question, to which reference has been made at several points, of the percentage of rain contributed to the ground water. Estimates of this character have been made by three methods — (1) by means of the lysimeter, (2) by stream discharge, and (3) by changes in the level of the ground water. ESTIMATION OF PEKCOLATION BY MEANS OF LYSIMETEBS. By the lysimeter method the rain water passing through a column of soil in field conditions is measured directly. The gage used for this purpose consists of a vessel with impervious sides and a pervious bottom, filled with the soil to be tested, and buried so that the surface of the soil in the gage is at the same level as the sur- rounding ground. The discharge through the pervious bottom of the vessel is col- lected by a cone and conducted by a small tube to the measuring gages. In the early forms of the apparatus used by Dalton, 1796, and Dickinson, 1835, surface outlets were provided to discharge the excess rainfall, but these were abandoned when it was found that on the level surface of the gage there was no surface run-off. Many observations have been made along this line, and while the results for long periods clearly have a greater value than those for short periods, some of these short- period values have been included in the table on the following page for the purposes of comparison. Lysimeter results have been subjected to considerable criticism, and very differing views expressed regarding their value. It has been suggested ( 1 ) that the material in the gage is not in the natural condition of consolidation, and that, therefore, the results are too high; (2) that the underdrainage necessary to collect and carry the water from the base of the soil column to the measuring tube introduces an unnatural condition whereby the results are too low; (3) that the surface run-off factor is surpressed and the results are too high. FLUCTUATIONS DTK To RAINFALL AMD EVAPORATION. 45 «J c 3 e 1 C < tc > l> c ) - c I' r d . id C> O td ■^ =9 o 7. 0J.-4 ^J cm CO CM C< CO CM CO ri - X — ^ ^.~ 41 «~- P*» o!) s CO _ > ^ CM O lO C-l to o> -* i.C r. u a S3 J 00 o r-' t-' o> ■o M CO « CJ-M ft 3 ■« _; .= 1/5 to : i is o c CO t^ o> Oi to to OB «C2 u co to 00 01 to lO l^ to 1^ OS P to CO r i ? 1 CM CM CM CM CO « ^ £ £ co p 1 CM OS 00 CI !-H ,_, ,_, ^ ■* CI u$ « 3 *S 3 CO lO o> o ■^ ■* lO 1 lO to to 1^ / 00 CO cc 00 / ^ 00 I- ■ CM *J ^J ^J ^J ^J ^J ^J 'C rH rH O CJ CJ o CJ o ^ o o o Q o *C to 1 1 I to i-H ifi l OO 1 CO CO Oi » C- :: CO" iC to 1^ I / / 00 00 00 OO / 1-1 ■-I i-H rt 1-1 i-H i-l 1-1 1-1 l -1 ^J CJ CJ u o C3 o O o o o o o o rC>0 te 3 -52 o so 3^ a> cj = cj 91 -z z z 03 > B CJ CJ i- 0) CJ CO XJ •a CJ t4 CJ 73 CJ CJ ft "o CJ .2 03 33 01 ^H "o a .2 ^83 o CJ u CJ ft "S 03 CJ a 1 03 u • I X IS Ooo li '7 1 (-. 73 o 'C CJ ft o a a cj ft X 0) o a _o 73 a o a ft& cj a> 73 > co§£ . M 60 a^ ? .5 2 > i — -_ j3 33 «- CJ CJ 32 SSI' CJ t- ft so -a a v£ ,£ 1 sp 1 c - z. c 73 to -r 13 3 ■f a a o SO £ a i CJ = z o 2 5| ii — 7 = 8 <*> « g - 33 IS — - &8 a a O -" f o ™- ^ So T3 CJ . -^ 8irig,g si gco5 u ^a.s-Scj ^.S ft a cj 0613 >"*& 5? ^a "5 -o _; ©Sd »« tu >h a a ~ «•£ 2 * a'-a-e^-a | O-" «l K SO 33 CO u CJ CJ a s a CJ ^3 CJ CI a o «l 0)73 T3 03 <*« o CJ co (5 CJ CJ cj Sj 1 1 •S § cS a •g i ■s g 00 "g g "Sop >.— r-c CJ » "^00 si ll 9 "3 03 CJ a 3 a CJ 73 a S'jO a m CJ "0 CJ 1 CJ 1) si — U 0^3 ■s— 'S ^ ■S 4 tr CC si 1-1 >,=* * 03 >" I- U - O.S so r S« S3 A § a si c c o o o o o o a c 1? -' C ii — cc CJ p - "3 s c3 T ) a a B C t. 1 £ — — j cj o a 1 4 Si s W c c c c o a CJ z ■c X -c -c •d — a a 9 33 sw ~ C c !zi ■>-. • 33 > t»" W CJ > 0) si ,a ■e • a O c a a o c 1 9 6 c d o o ; 'E - S ft p p p p p r 3 9 CJ - "^ a s . ^ w ^' ' 00 ft ^ 46 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. £.P O i-r ih eo w>S s 1 3* -3 0) jh O 0) ,2§>,£ « o3 n"\3 ^2 as 00 ° 2 o3 Pi .3 'H O 2 pi o3 5J ft S C n M_l rr-{ ,r, ^ > w h 1 *- 1 o3 a) 3<^fi 5a s "IS O X :.': •d g as ..d-d o .§^o ^a^s" ££;3-£ £ "> 9 o>_ ^n a2 h +j c.n pi ^c.sa-§ -p| 3>S hocu'C S Pj ?■ +j ■ H MCJf> PI . fii 0) c3,2 O ^ 5 3 c-3 " aSd ill I K Sg =3 go 'C cc 2 oo fe ^ »~p CD "So x tc.2 "C '° ca.2 EiS fi+j PI > O ;>, A° aS A« as . A t3 PI o3.2 "2« o3 o A^ a| ° 3 &i •C 3 4J "^ 03^ 03 > A3 Pl-"S •Sis ^, c3 £"3 ■S"? % a> as Si? aa oj O .3 ^ A ^ OrQ OT3 rr'C 31 dec •5^ a« Co3 'o.g d.S hrR 3X3 ^ OJ ttfel 11 03 a 11 S3 !>>0 a Q pq a 03 S I'll o 2 3 h 03 fltl ■"-„ A H 0) tj'a §WM !z l=R fl-o i3 FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 47 CJ If? o> 00 to o I - 1- t© to 5 00 lO l~ -r s 3 ■* cm o t~ >ra vn ■01 lO 00 ~ en CO I-l M CN to m Ci5 co ,_, rH n IN "1 -1" CN 3 a 3 a 3 ■Oi 00 co CO CO u C-J CO CO ^ H j£ in rt CO y © CC tC 3J B g S SUB'S a) *j sod fl.3 o> s ^•S "c "O >,£|~ .3 .3 3 - 03.5 Srl ■a -o is t- 3. 3 Oi CS C>-3 cS to 2 cS 3 ai 3 — < _~ pj C^-^^c -c tfi oS_.d 3 O ft ft ft 5d6g M o> 01 9) o> T -a . m a ai 03 01 - 015 3 "3 o'C'Z! Flffi 1) £ * ■afj-g is 5 or to SO o H 7? d 0) -' k 1-1 .« I-l J3 ""f^ = agc3

at > to s OJ.ti o3 P O o O C3 ai to CO d^ 5 3'£ 3" > > > ^ S >* 03 03 03 3 °= 3 =i os OS- oJ" OJ OJ QJ 2 1> n 03 03 03 03 co to to co W-d ■a to o> d .S ft is °-> o 01 O >d - A "8 E-t O 5? sss to" 1 ! ft ft rH CO o o> - ' . "S ^ oil- g to _r fl eo S3 '—, .0 < 2w • fid .. oi bS ^ W a! -2 hH c3 . ' "O § S-'ftf S ft ft „- ft » . . g ft "^ M - S - to" J -oSSoo ■^ (H 55 f-H *"^ 03 +J aj ■a -2^" a s S* to ■ -a 1 S*5ti ™ .'c >- d ,0 00 +3 +e fl " >.^— O !-._, (.It 3 u -r 1 .-'6 r r," T. /. Z - „7 3 -^ o> ft •< > - o> - CD to 5 - P - 3J =1 Z > C 2 0) S 1 ^ ft 3 to ■d H O - - 03^ +Z aq O %&& X to — -£ IS 'iT .3 - 3' ft O ai O X - 01 O - X 3 ~ ,c w H 03 jj d c -r pi >> C F- > O ci >-- _) S -1 C £ ti ; ^ >-» 48 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. The objection regarding consolidation is well taken, though it clearly does not apply to the results at Rothamsted, where natural soil columns were used, and where very high results were obtained, or to other gages after the first few years, during which the soil has settled. In the Hemel Hempstead experiments the percolation between 1835 and 1843 was 42.5 per cent of the rainfall, while in the period 1835 to 1853 it was but 35 per cent, a decrease which is perhaps due in part to the gradual compacting of the soil, though it is also affected by the varying rate of percolation, for it must be remembered that the amount of percolation depends more on the time at which the rain falls and the manner in which it is distributed than on the actual amount. Regarding the unnatural condition introduced by the method of drainage, it has been suggested on the one hand that the lower part of the soil column is thus exposed to evaporation, and that not only is there a loss in this manner not accounted for by the measuring gage, but that the dry condition of the basal layer would retard the percolation to a measureable extent and increase the loss by evaporation from the surface of the ground. Ebermeyer has shown, however, that with small lysimeters the lower portion of the soil column is damper than normal," and he proposed to remedy this by constructing larger gages. These defects are clearly ones of con- struction which it is possible to remedy. At Rothamsted essentially' the same results were obtained from soil columns 20, 40, and 60 inches high (fig. 12), showing con- clusively that in this case the natural conditions were not essentially disturbed. Indeed, it is believed that carefully conducted lysimeter observations, extending over long periods, such as are represented by the Dickinson and Evans and the Lawes and Gilbert experiments, give very important values bearing on this question, the Lawes and Gilbert results being particularly important and trustworthy. They indicate that in the climatic condition of middle England, with 28 inches of rainfall, half of the rainfall is contributed to the ground water through a rather heavy soil not covered with vegetation, and half of it evaporated. If the rainfall is greater, the percentage increases, if less, it decreases. Had the ground been covered with leaves, straw, or similar matter the percolation would have been greater; if covered with growing vegetation, less. Lawes and Gilbert estimate, from their observations on plant transpiration, that in this region 2 inches per year would represent the plant transpiration in the area of downs and waste land, where there was very little vege- tation, while with a heavy grass or mangel crop it would amount to 7 inches or more. The average for the whole region was estimated at 3 to 4 inches. This would make the percolation for soil of this character, in the case of downs and waste land, 43 per cent; for the average mid-England district, 39. per cent, and for land covered with heavy grass or mangel crop, 25 per cent or less. It should be noted in this connection that while the most of these observations, including those at Rothamsted, were made in connection with agricultural investi- gations, the Hemel Hempstead and Lee Bridge (Greaves) experiments were made for engineering purposes. The Hemel Hempstead observations were undertaken by a paper manufacturer, dependent on the water power of a spring-fed stream. He argued that stream flow depended on the amount of water which percolated through the soil; that measurements of this quantity would indicate the stream flow to be expected during the following summer. It is stated that he found that the indica- tions of the gage during the winter enabled him to calculate the supply of water trom the stream during the ensuing season, that he had always found the indication per- fectly reliable, 6 and that he was accustomed to regulate the volume of the orders accepted for the summer season by the indication of the gage for the preceding winter, c Olutterbuck adds, though the relation is clearly more of a qualitative than a Reported by JR. H. Scott, Jour. Royal Agnc. Soc , 2d ser., vol. 17, 1881, pp. 66-67. f»Min. Proc. Inst. Civil £ng , vol. 2, 1842, p. 158. clbid., p. 157, FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 4'.) a quantitative one, thai the rise in level of the water in the wells in that region is found to coincide with the readings <>f the Dickinson gages. The lysimeter certainly furnishes a very direct and exact means of determining the amount of water contributed to the ground water at any given point. The principal objection to it is that the block of soil tested is not necessarily representative of the whole area under investigation. Somewhat similar experiments, having for their object the determination of the evaporation from plants, have been conducted by many agriculturists in this country, notably by King, 6 by means of tanks which can be lifted from the ground and weighed. This method is not so applicable to the amount of water contributed to the ground water as is the lysimeter type used above, for the results are less direct, the percolation being inferred from the evaporation generally under more artificial conditions than with the lysimeter. ESTIMATION OF PERCOLATION FROM THE STREAM DISCHARGE. The favorite method of estimating the evaporation of a given drainage basin is to subtract the stream discharge expressed in inches of rainfall over the drainage basin from the average rainfall. Thus it is assumed that — Rainfall — Stream discharge = Evaporation. The stream discharge is composed of ( 1) the rain water which flows into the drain- age channels without penetrating the soil — this is, strictly speaking, the run-off, but as this word is now by common consent used for the whole quantity of water dis- charged by the river, this contribution may be somewhat arbitrarily referred to as the flood flow; and (2) the water which after a greater or less journey through the earth returns to the surface — this may be called the spring or ground- water flow of the river. It has been assumed that the ground-water flow of a river is its low-water flow and that any excess of this quantity can be regarded as flood flow. This is far from being a general fact. As the height of the ground-water table increases the stream dis- charge also increases, and it is possible to have high and low waters dependent entirely on the fluctuation of the ground-water discharge. In streams which cut the ground-water table and are clearly ground-water-fed streams, such as those on Long Island, a rise in the ground-water table changes the position of the head of the stream, and by thus increasing both the head and the area of discharge greatly increases the stream flow. The total range of the ground-water table near the coast is much less than near the ground-water divide, and the discharge during periods of high ground- water level may therefore be disproportionate to the changes in level recorded by the wells near the coast. Because of these great changes in the area of the discharge and the relatively free flow of the surface water, it often happens that the fluctuations of the stream height in the lower part of the stream are greater than the changes in the level in wells in the same region. Heavy rains, with no surface run-off, may likewise produce sudden and consider- able rises by increasing the spring flow by transmitted pressure in the manner described above (p. 42). In the 14-foot well at Lynbrook. (p. 23), besides the sud- den rises recorded for every rain, the water four times during the period of observa- tion rose above the surface. In the first instance the water, much to the amazement of the observer, gushed over the top of the pipe a few minutes after the shower began, and, while after the pipe was raised this did not occur again, the records show that on several occasions the water rose higher than the ground surface. There appear, then, to be great and almost insurmountable difficulties in the way of the satisfactory separation of the stream discharge into spring or ground-water flow and flood flow. « Min. Proc. Inst. Civil Eng., vol. 9, pp. 153, 156. bAnn. Repts. Wisconsin Agric. Expt. Station, 1892, pp. 94-100, 1893, pp. 162-159, 1894, pp. 240-280; 1897, pp. 228-231. irr 155—06 4 50 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. It is the belief of the writer that in the eastern United States the portion of the total stream flow attributed to ground-water contributions is commonly greatly under- estimated. On Long Island Spear, from a comparison of the hydrographs of several of the streams near the south shore with the fluctuation in neighboring wells, has con- cluded that of the total stream discharge but 57 per cent is spring or ground-water flow. a This is an extremely low value, and from a consideration of the various factors involved it is believed that 90 per cent is much nearer the true value. On this basis the flood flow is but 3 or 4 per cent of the yearly rainfall. In the simple equation, Eainf all— Stream flow = Evaporation, no account is taken of the underflow, it being assumed that all the ground water is returned to the stream above the point at which the measurement is made, an assumption which is far from correct. The result of this is to give to the evaporation a value just as much in excess of its true value as there is loss by underflow. Thus on Long Island, where the perco- lation is perhaps 60 per cent of the rainfall, the estimate of Spear b gives the total nor- mal stream discharge as 33 per cent of the rainfall, and the. estimates of the Brooklyn water department are still lower. This, according to the above formula, would give a loss by evaporation of 67 per cent, when it is actually about 40 per cent. It may be assumed, however, except in regions deeply covered with loose superficial mate- rial, such as Long Island, that the loss by underflow is less than the excess by flood flow, and that the total stream flow represents a quantity slightly larger than the percolation. With thh in mind, some idea of the amount of percolation can be obtained from the following values: Rainfall and run-off of drainage basins in the United States. Drainage basin. Watershed of southern Long Island Muskingum River, Ohio Genesee River, N. Y Lake Cochttuate, Mass Mystic Lake, Mass Croton River, N . Y Neshaminy Creek, Pa .'. Sudbury River, Mass Sudbury River, Mass Perkiomen Creek, Pa Connecticut River, Conn . . . Hudson River, N. Y Nashua River, South Branch, Mass. Pequannock River, Conn Years of record. 1888-1895 1890-1898 1863-1900 1878-1895 1868-1899 1884-1899 1875-1900 1875-1902 1884-1899 1872-1885 1888-1901 1897-1902 1891-1899 Average yearly rainfall. Inches of depth. 42.56 39.7 40.3 47.1 44.1 48.07 47.6 46.1 46.38 48.0 43.0 44.2 51.32 44.2 Average yearly stream flow. Percent- age: Stream flow of rainfall. Inches of depth. 14.0 33.0 13.1 33.0 14.2 35.0 20.3 43.0 20.0 45.3 22.33 46.5 23.1 48.5 22.6 49.0 22.75 49.0 23.6 49.0 22.0 51.0 23.3 52.7 27.56 53.7 26.8 60.6 Authority. Spear. Rafter. Do. Do. Do. Freeman. Rafter. Do. Metropolitan water- works. Rafter. Do. Do. Metropolitan water- works. Rafter. In this table the large loss by underflow in the Muskingum and Genesee drainage basins is evident. ESTIMATION OF PERCOLATION FROM CHANGES IN LEVEL OP THE GROUND- WATER TABLE. The broader and more important fluctuations of the ground-water table are clearly due to the infiltration of water, and attempts have been made repeatedly to estimate nRept. New York City Commission on Additional Water Supply, 1904, p. 829. b Ibid., p. 795. FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. 51 the aim mm of infiltration by the rise in the ground water. After ha\ ing determined the available pore space, which is by do means a simple matter, it appears very easy t<» calculate the amount of water which will cause the water Level to rise a few inches or a few feet. The method is an attractive one, with an appearance of exactness and simplicity, and bas often been tried. Theresults have very little meaning, however, for the very important reason that the same rainfall under the same climatic condi- tions will produce very different rises in material of the same porosity; for, as pointed out previously (p. 38), the amplitude of the fluctuation increases with the distance from the ground-water discharge. Thus, with material of the same porosity, an annual rainfall of 25 inches produces at Wiener Neustadl a fluctuation in one well of 1 foot, and in another a fluctuation of 25 feet (PI. IX). It i.s evident that a calcu- lation of infiltration based on the rise of water produced in well No. 1, will give very different values from that of No. 2. and yet it may he confidently asserted that the same amount of percolation is received in both. Similarly, in the chalk region of England, the fluctuation in the same region ranges from a few inches to 50 or 100 feet. The impossibility of accurately calculating the amount of percolation from the rise of the ground-water table is evident. REFERENCES RELATING TO WELL FLUCTUATIONS DUE TO RAINFALL AND EVAPORATION. The bibliography relating to fluctuations of water in wells due to rainfall is natu- rally very extensive, and an attempt has been made to collect a few only of these titles, important either for their general bearing or their special reference to the United States: Acchincloss, W. S. Waters within the Earth and Laws of Rainflow, Philadelphia, 1897. Gives record of fluctuations in well at Bryn Mawr, Pa., 1886-1895, showing annual and secular fluctuations. Barbour, Erwin Hinckley. Water-Sup. and Irr. Paper No. 29, U. S. Geol. Survey, 1899, p. 28; Nebraska Geol. Survey, Kept, of State Geologist, vol. 1, 1903, p. 106. States that wells in Nebraska show an annual fluctuation independent of the rainfall, with the maximum occurring in February. Clctterbuck, James. Observations on the periodic drainage and replenishment of the sub- terraneous reservoir in the chalk basin of London: Mill. Proc. Inst. Civil Eng. [vol. 2], 1842, pp. 155-165; 184::, pp. 156-159. Oh the periodic alternations and progressive permanent depression of the chalk-water level under London: Min. Proc. Inst. Civil Eng., vol. 9, 1850, pp. 151-180, PI. VI. Emery, Frank E. Notes on fluctuations in height of water in an unused well: Eighth Ann. Rept. New York Agric. Expt. Station for 1889, 1890, pp. 374-375, fig. Records monthly observations from December, 1886, to December, ISN'.i, on a 40-foot well at Geneva, N. Y.. which shows a single yearly period independent of the rainfall. Poktier, Samuel. A preliminary report on seepage water and the underflow of rivers: Bull. Utah Expt. Station No 38, 1895. On p. 30, under heading "Effect of subsurface temperature on rate of flow," are given dis- charge, temperature, and rainfall at Denver Water Company's plant at Cherry Creek, from 1888 to 1891. This is an infiltration gallery in fine sand 15 feet below the surface, and the discharge shows a rather regular yearly fluctuation with a minimum in February-March and a maximum, normally, in August-November. This fluctuation is ascribed by Fortier to changes in soil tem- perature. It should be pointed out, however, that, while the annual changes in soil temperature do affect the rate of flow (see p. 59), the yearly maximum is independent of this fluctuation and the agreement here is merely a coincidence. Fkeund, Adolph (secretary). Bericht des Ausschusses fur die Wasserversorgung Wiens, Osterreich- ischen Ingenieur- und Architekten-Verein, 1895. PI. Y, Graphische Darstellung der Wasserstsinde im Steinfelde, 1883-1888. Gerhardt, P. I. Handbuch der Ingenieurwissenschaften, vol. 3, Der Wasserbau, i>t. 1,1892, pp. 46-51. Under heading " Schwankungen des Grundwas ers," gives a summary based largely on the reports of Soyka. Headden, Wilhelm P. A soil study, pt. 4, The ground water: Bull. Colorado Agric. Expt. station No. 72, 1902. Gives data regarding effect of single showers. KING, Franklin II. Fluctuations in level and rate of movement of ground water: Bull. D. S. Weather Bureau No. 5, 1892, pp. 72-74. Discusses instantaneous percolation after rains. 52 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. King, Franklin H. Principles and conditions of the movement of ground water: Nineteenth Ann. Rept. U. S. Geol. Survey, pt. 2, 1899, pp. 100-106. Discusses "Elevation of ground-water surface due to precipitation and percolation," largely from standpoint of porosity. Liznak, Josef. Ueber die periodische Anderung des Grundwasserstandes, ein Beitrag zur Quellen- theorie: Gsea, vol. 17, 1881, p. 330; Meteorol. Zeitschr., Wien, vol. 17, 1882, pp. 368-371. Luegeb, Otto. Die Schwankungen des Grundwassers: Geea, vol. 24, 1888, p. 630. Michigan State Board of Health. Annual reports, 1878-1903. Contain monthly observations of water level at many points in Michigan. Soyka, Isidor. Experimentelles zur Theorie des Grundwasserschwankungen: Vierteljahrsschrift fiir off. Gesundheitspflege, vol. 4, 1885, p. 592. Der Boden (Handbuch des Hygiene und der Gewerbekrankheiten, vol. 2, pt. 3), Leipzig, 1887, pp. 251-351. Contains much of the material incorporated in the following report. Die Schwankungen des Grundwassers mit besonderer Beriicksichtigung der mitteleuropa- ischen Verhaltnisse: Penck's Geographische Abhandlungen, vol. 2, pt. 3, Wien, 1888. Spear, Walter E. Long Island sources: Rept. New York City Commission on Additional Water Supply, appendix 7, 1904, pp. 816-826. Discusses fluctuation in elevation of ground-water surface. Todd, James E. Water-Sup. and Irr. Paper No. 34, U. S. Geol. Survey, 1900, p. 29. The normal yearly maximum is here tentatively referred to the melting of snows or floods. Tribus, L. L. Trans. Am. Soc. Civil Eng., vol. 31, 1894, pp. 170, 391-395. Reports (p. 170) that in driven wells in New Jersey, 50 feet deep, the effect of rain was rarely felt in less than thirty hours; gives curve (pp. 391-395) showing fluctuation of water level in 50-foot well at Plainfleld, N. J., waterworks, 1891-1894. This shows normal annual curves slightly affected by pumping. Woldrich, Johann Nepomuk. Ueber den Einfluss der atmospharischen Niederschlage auf das Grundwasser: Zeitschr. Meteorol., Wien, vol. 4, 1869, pp. 273-279. : — Ueber die Beziehungen der atmospharischen Niederschlage zum Fluss- und Grundwasser- stand: Mitt. d. Techn. Klubs zu Salzburg, pt. 1, 1869. FLUCTUATIONS DUE TO BAROMETRIC CHANGES. CHARACTER AND CAUSE. Changes in air pressure have been observed to affect wells in two ways; in some there is an inflow and outflow of air, and in others a rising and lowering of the water level. a The rise or outflow occurs with a falling barometer, and the depression or inflow with a rising barometer. In the case of flowing wells, when the external air pressure decreases, the air witbin the earth expands, and, as the escape through the soil is greatly retarded by friction and as the well offers a free escape, it finds relief through the well tubing. The power of this blast evidently depends on the area tributary to the well, the loss by friction, and the rate of lowering of the outside pressure. On the other hand, with a rising barometer the external air flows into the pipe to supply the volume lost by the compression of the soil air or earth air. If water is interposed between this included air and the well, the well becomes a rough differential-pressure gage in which the maximum change possible is about 12 inches of water for each mercurial inch of variation in the barometric pressure. Where there- is no soil air suitably confined to produce the result just described, the air and other gases in the water so increase its compressibility that a small, though measurable result may be produced by the direct compression and expansion of the water. In the latter case the depth of water involved is an important factor, and it is probably for this reason that on Long Island, where the earth air occurs only in the porous surficial soil, from which it is relatively free to escape, in the wells at Lynbrook fluctuations due to barometric changes are clearly noticeable only in the 504-foot well, and these are relatively small. An examination of PI. VI shows that there are no indications of barometric influence on the curves from the 72-foot well, but that in the 504-foot well there is a striking resemblance. The similarity is, however, in some places due in part to other causes. Thus the elevations of the water on July 18-19, August 5, and September 16, while closely following the barometric a See in this connection Nineteenth Ann. Rept. U. S. Geol. Survey, pt. 2, 1899, figs. 3, 4, 5; Water- Sup, and Irr. Paper No. 67, U. S. Geol. Survey, 1902, fig. 39, p. 73. FLUCTUATIONS DTK TO BAROMETRIC CHANGES. 58 curve, are partly due to rainfall. This is indicated by the fact that somewhat similar abrupt barometric depressions on July 26, August 20, and September^ 5, when there were no important rains, did no1 produce similar elevations of the water in the well. There is a further point of resemblance and dissimilarity between the curve for the 504-foot well and the barometric curve. The well curve shows a very marked semi- diurnal fluctuation, the two parts of which are generally of about the same value, although they sometimes merge into a pronounced diurnal wave, as <>n August I, 2, and :i. In the barometric curve, although there is a tendency toward a semidiurnal well wave, it is nowhere well marked. The semidiurnal well wave is clearly not tidal. Its resemblance to the barometer curve is, however, sufficiently close to lead to the belief that it is largely barometric, but modified by some other element, perhaps the diurnal temperature wave which shows in the 14- and 72-foot wells (pp. 24-25). REFERENCES RELATING TO WELL FLUCTUATIONS DUE TO BAROMETRIC CHANGES. BLOWING WELLS. Barbour, Erwin Hinck ley. [Blowing wells in Nebraska] : Water-Snp. and Irr. Paper No. 29, V. S. Geol. Survey, 1889, pp. 78-82; Nebraska Geol. Survey, Rept. of State Geologist, vol. 1, 1903, pp. 93-97. Karri. ey, T. On the blowing wells near North Allerton: Proc. Yorkshire Geol. and Polyt. Soc., n.s., vol. 7, pt. 3, 1880, pp. 409-421, PI. VII. Harris, Gilrert Dennison. [Blowing wells in Rapides Parish, La.]: Water-Sup. and Irr. Paper No. 101, U. S. Geol. Survey, 1904, pp. 60-61; Louisiana Geol. Survey, Bull. No. 1, 190, r >, pp. 59-60. Lane, Alfred C. Water-Sup. and Irr. Paper No. 30, U. S. Geol. Survey, 1899, pp. 55-56. Records shallow blowing well in Michigan. Veatch, A. C. Prof. Paper U. S. Geol. Survey No. 44, 1906, p. 74. Records reported occurrence of blowing wells on Long Island, New York. CHANGES IN WATER LEVEL. Atwell, Joseph. Conjectures on the nature of intermitting and reciprocating springs: Trans. Phil. Soc. London, No. 424, vol. 37, 1732, p. 301; Trans. Phil. Soc. London from 1665-1800 (abridged), vol. 7, 1809, pp. 544-550. Describes irregular fluctuations of short interval in springs at Brixam, near Torbay, in Devon- shire called "Lay well." These, he supposes, are produced by the action of natural siphons. Perhaps they are barometric fluctuations. Denizet. Sources sujettees a des variations qui paraissent liees a l'6tat du barometre: Comptes rendus, vol. 7, 1839, p. 799. Reports that springs at Voize are affected by changes in barometric pressure, and that the dis- charge is directly related to the pressure, instead of inversely, as has been proved by more recent work. Gough, John. Observations on the ebbing and flowing well at Giggleswick, in the West Riding of Yorkshire, with a theory of reciprocating fountains: Jour. Nat. Phil. Chem. Arts, ser. 2, vol. 35, 1813, pp. 178-193; Mem. Phil. Soc. Manchester, n. s., vol. 2, 1813, pp. 354-363. The water level in this well fluctuates irregularly at short intervals, and Mr. Gough suggests that the fluctuations are produced by the obstruction resulting from the natural accumulation of air bubbles in the outlet and the relief resulting from their irregular escape. He cities the irreg- gular fluctuations of the Weeding well in Derbyshire and the Lay well near Torbay, as proving that the prevailing idea of the production of these phenomena by natural siphons is erroneous. He concludes, very correctly, that too little is known of the fountain of Jupiter in Dodona and Pliny's well in Coruo to judge of the true cause of the fluctuations mentioned by Pliny. King, Franklin H. [Influence of barometric changes on discharge and water level in wells and springs at Madison and Whitewater, Wis.]: Bull. U. S. Weather Bureau No. 5, 1892, pp. 44-53; Nineteenth Ann. Rept. U. S. Geol. Survey, pt. 2, 1899, pp. 73-77. Latham, Baldwin. On the influence of barometric pressure on the discharge of water from springs: Brit. Assoc. Rept. for 1881, 1882, p. 614; Brit. Assoc. Rept. for 1883, 1884, pp. 495-496. The fluctuation of the Croydon Bourne, due to barometric pressure, on one occasion exceeded half a million gallons per day. Observations in deep wells are also recorded, showing that fluctu- ations are inversely related to the pressure. The fluctuations are attributed to the expansion and contraction of the air and gases in the water. Knightley, T. E. Ebbing and flowing wells [in Derbyshire, England]: Geol. Mag., n. a., decade 4, vol. 5, 1898, pp. 333-334. Suggests that irregularities in flow are due to cavernous limestone, which, by means of natural siphons, gives rise to "intermittent spring phenomena." The possibility of these fluctu- ations being due to barometric changes is not considered. 54 FLUCTUATIONS OF THE WATEE LEVEL IN WELLS. Lueger, Otto. Einfluss der Atmosphiirendruckes auf die Ergiebigkeit von Brunnen und Quellen: Centralblatt d. Bauverwaltung, 1882, p. 8. Milne, John. Seismology. London, 1898, p. 243. Reports two sinkings and two risings of about 5 millimeters in a shallow well near Tokio. The sinkings took place between 2 and 6 p. m. and 2 and 5 a. m. Note: These fluctuations suggest the diurnal barometric wave, but they can be referred to it only if the time given is taken to mean the time at which the water commenced to sink and not the time of the low-water stage. Oliver, Dr. William. Of a well that ebbs and flows . . . : Trans. Phil. Soc. London, No. 204, vol. 17, 1693, p. 908; Trans. Phil. Soc. London from 1665-1800 (abridged), vol. 3, 1809. pp. 585-586. Describes well near Torbay called " Lay well," that ebbs and flows from 16 to 20 times per hour. See Atwell, above, and discussion of minor periodic fluctuations on p. 75. Pliny the Younger (Caius Plinius Caecilius Secundus). Epistolae, lib. 4, epist. ult. In his letter to Licinius Pliny describes the fluctuations in the discharge of a spring near his villa in Como, Italy, which he states ebbs and flows thrice a day. He suggests that the fluctua- tions may be due to " the obstruction of air," tidal action, or some secret and unknown contriv- ance in the nature of a valve. Pliny the elder, in his Historia Naturalis, lib. 2, cap. 106, refers to the same spring, but states that it ebbs and Hows every hour — a statement which is verified by Catanseus, the learned commentator on the Epistles. From the meager and contradictory data given it is unsafe to venture a decided opinion on the cause of these fluctuations, but they may be tentatively regarded as barometric. Roberts, Isaac. On . .- . the variation in atmospheric pressure . . . causing oscillations in the underground water in porous strata: Rept. Brit. Assoc, for 1883, p. 405. States that autographic records from a well at Maghill, near Liverpool, show such fluctuations. Slighter, Charles S. Water-Sup. andlrr. Paper No. 67, U. S. Geol. Survey, 1902, pp. 71-72. Refers to reports of Latham, King, and Lueger. Todd, James E. Bull. Geol. Survey, South Dakota, No. 2, 1898, p. 116; Water-Sup. and Irr. Paper No. 34, U. S. Geol. Survey, 1900, p. 29. Reports that well discharge varies inversely with barometic pressure. FLUCTUATIONS DUE TO TEMPERATURE CHANGES. OBSERVATIONS AT MADISON, WIS.— FLUCTUATIONS VARYING DIRECTLY WITH THE TEM- PERATURE. Iii 1888 King observed in certain shallow wells at Madison, Wis., that the water regularly for a portion of the summer months stood higher in the morning than at night, a Further observations during the period 1888 to 1892 showed that there was . in many wells a diurnal wave, distinctly marked during the summer and dying out in winter, which was clearly not barometric in character, and was not produced by the unequal plant transpiration during the day and night. Suspecting that these changes were intimately related to temperature, King tried the following experiment: A galvanized-iron cylinder, 6 feet deep and 30 inches in diameter, provided with a bottom, and water-tight, was filled with soil, standing its full height above the ground in the open field. In the center of this cylinder and extending to the bottom a column of 5-inch drain tile was placed and the soil filled in about it and packed as thoroughly as practicable. Water was poured into the cavity formed by the tile until it was full, and allowed to percolate into the soil so as to saturate it and leave the water standing nearly a foot deep in the well. When the water in this artificial well had become nearly stationary one of the self-registering instruments was placed upon it.?' In order to avoid any complications due to percolation, the apparatus was provided with a cover which could be put on in times of rain and removed again during fair weather. The first records showed a small diurnal oscil- lation, and as the season advanced these increased in amplitude until finally the water rose in the well during the day of July 8,1.8 inches and fell during the following night 1.84 inches. After these diurnal oscillations had become so pronounced and so constant, a series of thermometers were intro- duced into the side of the cylinder, extending to different distances from the surface, and a record^ kept of the changes in the soil temperature; and the result of these observations was to show that the turning points in the water curve fell exactly upon the turning points of the temperature of the soil in the cylinder. When this fact was ascertained, to show whether the correspondence in the time of the two curves was due to a diurnal cause, other than temperature, which had its turning points so related to those of the temperature as to cause the two to accidentally fall together, cold water was aKing, Franklin H., Observations and experiments on the fluctuations in the level and rate of movement of ground water on the Wisconsin Agricultural" Experiment Station Farm and at White- water, Wis.: Bull. U. S. Weather Bureau No. 5, 1892; also Ann. Repts, Wisconsin Agric. Expt. Station, 1889-1893. b For a figure of this apparatus see Bull. U. S. Weather Bureau No. 5. FLUCTUATIONS DUE TO TEMPERATURE CHANGES. . r )5 broughl from the well and, with a spray pump, applied to the Burface ol the cylinder nil around. ill. water was applied on a hot sunny day, jnsi after dinner, « hen the water was rising In the well, and the resull was an immediate change in the curve, the water beginning to Fall In the well. The water was then turned off, and the result of i his change was to stop the fall of the water in the well, us shown by a change in the direction of the curve. This led to tlic conclusion thai there was a very positive connection between the changes in the soil temperature and changes in the level of the water in the wells, and that the fluctuation varied directly with the temperature; that is, the water in the wells rose with increasing temperature ami fell when the temperature lowered. A specially constructed self-recording soil thermometer showed that at a depth of IS inches the minimum temperature occurred at noon, ami the the maximum a little after midnight. It was therefore argued that at a depth of 3 feet, or the level at which the wells were fluctuating, the maximum and minimum temperature would occur still later, and that the high water which occurred in the wells at \\ II ilation b the 3 ~ ^ shape of the curves. Thechar- acteristic of the air temperature g- — j? curve is a quick rise and slower •< g § fall; well fluctuations directly -r = ~ related to the temperature, as 3 -' 5 those at Madison, therefore s. must show a quick rise and a g !L s slower fall, but in the Lyn- 2.^ g. brook wells there is a quick ~ ^ § fall and slower rise (see PI. VI, <£ | | Aug. 1-3). These evidently 3, * f. belong in quite a different class g, g a from the temperature fluctua- tions observed at Madison, and involve quite a different rela- tion between the soil tempera- ture and the ground water. The soil is a very poor trans- mitter of heat, and there is not only a very rapid diminution of the temperature range with depth, but a very considerable time lag. Swezey's observa- tions « at Lincoln, Nebr., for a period of fourteen years show that while in winter the maxi- mum temperature occurs in the air at about the middle of the afternoon, at a depth of 3 to 6 inches it occurs in the evening, and at 1 foot it is delayed until the following morning; below 1 foot it is scarcely appreciable. In summer the daily range is considerably greater at all depths, the changes are appre- ciable to a depth of at least 2 feet, and are retarded to about the same extent as in winter. At Bronx Park, New York City, the record obtained by MacDougal >> with a Hallock thermograph showed that at 1 foot below the surface the max- imum daily temperature oc- curred between 8 and lip. m., and the minimum between 8 nSwezey, G. D., Soil temperature at Lincoln, Nebr., 1888 to 1902: Six- teenth Ann. Kept. Nebraska Agrie. Expt. Station for 1902, 1903, pp. 95-129; Expt. Station Record, vol. 15, 1901, pp. 4(30-461. t> MacDougal, Daniel Trembly, Soil temperatures and vegetation; Month- ly Weather Review, vol. 31, August, 1903, pp. 375-379. 58 FLUCTUATIONS OF THF WATEE LEVEL IN WELLS. and the maximum daily range was but 2° C. (3.6° F.), which was A o a a a," reached on but two occa- sions. The' total annual variation from June 9, 1902, to May 31, 1903, was 16.2° 0. (29.4° F.). Similar results have been obtained by Cal- lender at Montreal, a As a result of this slow transmission of tempera- ture, the temperature at the ground-water outlet may be at its maximum while a short distance away; where the water table is but a foot or two from the surface, the ground temperature may be at its minimum. Now, the rate of flow of water is greatly affected by temper- ature. Poiseuille found that water at a tempera- ture of 45° C. flowed 2.5 times as fast under other- wise like conditions as water at 5° 0. b This gives rise to the phenomena shown in fig. 17. There is thus produced a distinct and periodic fluc- tuation of the ground water, which is great near the ground-water outlet and decreases rapidly in amplitude as the "distance from the outcrop increases. The fluctuation is produced 03 £ " a water whereby the pressure conditions are constantly changed, and in this re- spect it differs from the Madison fluctuations (p. 54). It is this change in pressure that causes these fluctuations to show in the other wells, even to a depth of 500 feet. The same phe- nomenon of response to loading and relief from load is exhibited in some of the rainfall fluctuations described above (p. 42) and in the sympathetic tidal fluctuations (p. 65) . f-Callender, Hugh L., Proc. and Trans. Royal Soc. Canada for 1895, 2d ser., vol. 1, sec. 3, p. 79, fig. 4. b Quoted by King, Nineteenth Ann. Kept. U. S. Geol. Survey, pt. 2, 1899, p. 82; see also Carpenter, L. C, Seepage and return waters from irrigation: Bull. Colorado Agric. Expt. Station No. 33, 1896, pp. 42-44. FLUCTUATIONS DUE TO TEMPERATURE CHANGES. 59 These data suggesl thai the annua] changes of the soil temperature may proauce a Bomewhat analogous effect, the warm summer temperature assisting in the depletion or lowering of the water near the ground-water outlets, and the cold winter tempera- ture, by rendering the Mater more viscous, retarding the outflow. In one respect the result would be in the same direction as the annual ground-water fluctuations, ami this is perhaps to be considered one <>f the minor factors. It would also tend to make the time of occurrence of the maximum and minimum of the yearly fluctua- tion earlier near the ground-water outlets than on the divides — or just the condi- tion observed on Long Island (p. 35). OBSERVATIONS AT SHERLOCK, KANS. Diurnal fluctuations due to temperature changes were observed by Mr. Henry C. Wolff, at Sherlock, Kans., in 1U04, while working under the direction of Prof. C. S. Sliehter." Mr. Wolff reports that the wells are low in the evening and high in the morning, and that there is no important time lag between wells where the water line is ti inches below the surface and those where it is :; feet below. In a few wells where the water level was about .'! feet from the surface, which were observed long enough to show the shape of the curve, the characteristics of the Madison curve are shown — that is, there is a long fall and sudden rise. DIURNAL FLUCTUATIONS OF CACHE LA POUDRE RIVER, COLORADO. Carpenter has observed very regular diurnal changes in the height of Cache la Poudre River near Fort Collins, Colo.'' Here the high water occurs at from 4 to 6 a. m., the low water at 8 p. m., and the extreme range of the daily change in river level noted was about 1 foot. The curves show the same characteristics as those observed by Wolff at Sherlock, Kans. ; there is a long fall and a sudden rise, and they are therefore directly related to the temperature. Carpenter concludes that these fluctuations are due to differences in daily melting in the snow fields, and that the occurrence of the high water in the morning is due to the distance from the snow fields. While this is a very possible explanation, and in the writer's opinion it is probably the correct one, it is desirable to have the matter checked by other obser- vations. If the fluctuations are purely due to daily waves moving down the river, due to melting snow, gages at other points should show the maximum and minimum at different times; if the fluctuations are due to variation in rate of ground-water discharge and are analogous to those described above, the time will lie the same at different points. REFERENCES RELATING TO FLUCTUATIONS PRODUCED BY TEMPERATURE CHANGES. Besides the references given above to the discussions of King, Sliehter, and Wolff, it is desirable to add here the early reference of Pliny to fluctuating wells belonging to this class: Puny the Elder (Caius Plinius Secundus). Historia Naturalis, lib. 2, cap. 106 [Pliny's Natural History, Bostock and Riley's translation, Bohn's Libraries, vol. 1, 1887, pp. 133-134]. " In the island of Tenedos there is a spring which, after the summer solstice, is full of water from the third hour of the night to the sixth." "The fountain of Jupiter in Dodona . . . always becomes dry at noon, from which circumstance it is called ' The Loiterer.' It then increases and becomes full at midnight, after which it again visibly decreases." Hardouin notes that there-is a similar kind of fountain in Provence called "Collis Martiensis." These fluctua- tions clearly belong to the class produced by temperature changes. FLUCTUATIONS PRODUCED BY RIVERS. Rivers may produce fluctuations of the water level in wells in three ways: (1) By changing the height of the ground- water discharge; (2) by seepage or actual con- tributions to the ground water, and (3) by transmitted pressure or plastic deformation. "The underflow of the Arkansas Valley in western Kansas: Water-Sup. and Irr. Paper No. 153, U. S. Geol. Survey. ''Carpenter, L. G., Bull. Colorado Agric. Kxpt. Station No. 55, 1901, figs. 2-6. GO FLUCTUATIONS OF THE WATER LEVEL IN WELLS. FLUCTUATIONS PRODUCED BY CHANGES IN RATE OF GROUND-WATER DISCHARGE. In regions where the sides of the channels are pervious and the ground water con- tributes to the stream flow, the water table, after a period of long drought, slopes regularly down to the water surface. If the stream rises through causes not asso- ciated with local ground-water conditions, the ground-water table is found likewise to rise to a greater or less extent. This is accomplished in part by an outflow from the river and in part by the accumulation of the ground-water flow, which can not so readily escape under the new conditions as under the old. If this new level were permanently maintained a complete readjustment would take place, and a line roughly parallel to the initial position of the ground-water table would be developed; and if there were no new outlets developed by this elevation of the ground water, wells at all distances would be similarly affected. Actually, however, the river is con- stantly changing; a series of waves of unequal height and duration, representing the high and low waters, are constantly traveling down every stream, and no stage lasts long enough for the establishment of a perfectly graded ground-water table, even were there no other factors involved. The result of this unceasing change is that the fluctuations are greatest near the river and become imperceptible at very short distances. This is due not only to the rapidity of the fluctuations in the river, but to the slow rate of outflow and accumulation. Observations made by Hess « along Aller River, near Celle, Germany, in 1866, give the values expressed in the table below: Table showing lag of high- and loiv-water stages in wells along Aller River, behind high- and low-water stages in the river. Distance from river. High water, Feb.- Mar., 1866. Low water, Mar. 7, 1866. High water, Apr. 1, 1866. Test well No.— High wa- ter in well be- hind high water in river. Rate per day. Low water in well be- hind low water in river. Rate per day. High water in well be- hind high water in river. Rate per day. 1 Meters. 47 140 351 468 584 Days. G 5 17 19 21 Meters. 10 28 21 24 28 Days. 2 3 4 7 Meters. 23.6 47.0 88.0 67.0 Days. 4 5 10 Meters. 12.5 2 28.0 3 35.0 4 5 Observations made by Slichter in very porous gravels in western Kansas, at Gar- den, Sherlock, and Deerfield, while showing a more rapid transmission than those just indicated, give very marked time lags. The conditions here are different in the respect that the ground water does not materially add to the stream flow, and the rise of the water in the wells is due wholly to seepage. The slope of the water plane is not toward the river, but downstream, at a rate very nearly the same as that of the stream. At Sherlock the water plane on July 27, 1904, sloped gently to the river from test well No. 5, but the water in test wells Nos. 2, 3, and 6 was lower than the river. Between 11 and 4 o'clock the river rose 1.6 feet, and then fell gradually. The beginning of the rise was felt in well No. 3, 400 feet north of the river, in less than two hours; in wells No. 2, 900 feet north of the river, and No. 5, 550 feet south of the river, in between three and four hours; in well No. 4, 1,000 feet south of the a Hess, Beobachtungen iiber dasGrundwassersdernorddeutschenEbene. Zeitschr. des Architekten und Ingenieurvereins in Hannover, vol. 16, 1870, quoted by Sovka, Der Boden, Leipzig, 1887, pp. 262- 267, figs. 23-25; and Gerhardt, Der Wasserbau, Leipzig, 1892, pp. 49-50, PI. I, figs. 7, 8. The diagrams given by Soyka and Gerhardt do not check with the values given in the text, and as the figures are clearly carelessly drawn the text values are reproduced here. , FLUCTUATIONS PRODUCED HY KIYKRS. (51 river, in four hours. Well X<>. •>, 2,500 feet from the river, fell during the whole period of observation. The difference in time in the occurrence of the maximum is expressed in the following table: Difference in time between high water in Arkansas River and wells on its banks, near Sher- lock, Kans., July, 1904. No. of well. Distance from river. Lag. Feet. Hours. « -100 3-6 a 900 12 «500 30 6550 108+(?) '■1,000 L08 '■2,500 w "North. & South. cNo rise in five days. It will be noted that the most rapid transmission was toward the north, where the water plane sloped away from the river, while the slight rise of the water plane toward wells Nos. 4 and 5 produced a very marked retardation. These observations seem to indicate that the rate of transmission is greater when the water plane slopes from the river than when it slopes toward it and help to explain the great retardation observed at Celle. In no instance in this Kansas work were the effects of floods observed in wells at distances of more than one-fourth of a mile from the river. In case there is open connection between the river and the well, such as might be afforded by limestone caves, changes in level may be felt at considerable distances with but very little time lag. Very rapid fluctuations would, however, be obliter- ated here, for the well would act very much as the still box used in tidal work, which consists of a large well connected with the ocean by means of a relatively small passage opening at some distance below the water surface. The wave action is entirely obliterated, because the water does not have time to flow in and out of the well in the period between fluctuations. The gradual changes of the tide are, however, exactly recorded. But direct cavernous connections between wells and waterways are rare, and generally river changes act through the interstices of the soil in the way observed at Celle and Sherlock or by transmitted pressure as described below. FLUCTUATIONS PRODUCED BY IRREGULAR INFILTRATION FROM RIVERS WITH NORMALLY IMPERVIOUS BEDS. Besides rivers which have pervious sides and into which the ground water is always free to flow or out of which the water flows whenever the river level exceeds that of the ground water, there are many rivers which, under normal conditions, so plaster their beds with fine silt that water is unable to flow either in or out. Nearly all rivers carrying large amounts of fine silt a are normally in this condition, and it is thus that the water in many delta regions is able to flow at heights above the surrounding land, and rivers in other regions flow at heights much above the normal ground-water level. Thus the Leitha at Katzelsdorf, near Wiener Neustadt, flows at a height of 10 to 70 feet above the level of the ground water at that place, the height depending on <«For an example of the silting power of a elear stream see Freeman, John R., Percolation through Embankments and the Natural Closing of Leaks, Boston Soc. Civil Eng., June 20, 1888. 62 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. the stage of the ground water, although there is at all times considerable seepage. (PL IX.) The Rio Grande in a similar way flows from central New Mexico to the sea, always above the ground-water level. When the rivers silt up their beds, where the water plane slopes down to the river and there is a tendency to discharge into the river, the silting develops an artesian head which causes the water to rise above the level of the river in wells sunk within the channel. Such occurrences have been reported by Salbach in the Elbe. a This Elbe occurrence may, however, be due to a sheet of clay not connected with the river-silt deposits of the present regime. During floods such rivers frequently scour out their beds and establish a connection between the surface and the underground waters. When the river is above the ground-water table, this leakage will raise the water level; when the reverse is the case, it may, by permitting a discharge of the artesian water, cause the water level in a near-by well to lower. Slichter has found at Mesilla Park, N. Mex., that the greater part of the underflow in the valley is derived in this way from the flood waters of the Rio Grande. & Pre- ceding the flood of October 5, 1904, the ground-water level was several feet below the river channel, although the river contained considerable water. The effect of the flood was well marked at well No. 7, three-fourths of a mile from the river, but did not affect well No. 6, 1.3 miles from the'river, in seventeen days. The rate of trans- mission is evidently quite similar in this case to those given above. Quite different from these slow rates of change in the ground- water level due to river changes are the changes in the London wells ascribed by Clutterbuck and Buckland c to floods in Colne River at Watford, Hertfordshire. According to these observers a rise in the Colne produces a rise in the London wells, 15 miles distant, in a few hours. These floods are due to heavy rains and it seems much more probable that the observed rise is due to the weight of the rain on the local London area than to the transmitted pressure from a flood 15 miles distant. Fluctuations of this character due to rainfall have been observed at Lynbrook, N. Y. (p. 42), and no such rates of lateral transmission have been observed either from the river flood or tides, with the possible exception of the tidal wells at Lille, France (p. 64.) Certainly there is no evidence of such great underground caverns between Lille and the sea as this rate of transmission would require, though its very occurrence, if conclusively proven, would indicate some such connection. FLUCTUATIONS DUE TO A PLASTIC DEFORMATION PRODUCED BY VARYING VOLUMES OF "WATER CARRIED BY RIVERS. The alternations of load due to the irregular waves, whose crests are the high- and low-water stages, which are constantly passing down every river, produce fluctua- tions anagalous to those produced by tides, though lacking their periodic character. They resemble also the sympathetic fluctuations produced in the 72- and 504-foot wells at Lynbrook due to variations in load produced by rainfall and thermometric changes (pp. 43, 58). The zone in which these fluctuations will be distinctly recogniz- able will be limited to a mile or two in the immediate vicinity of the river. REFERENCES RELATING TO FLUCTUATIONS OF THE WATER IN WELLS PRODUCED BY RIVERS. Buckland, Dr. Min. Proe. Inst. Civil Eng. [vol. 2], 1842, p. 159. Reports that wells in London rise in a few hours alter floods at Watford, 15 miles distant. Clutterbuck, James. Observations on the periodical drainage and replenishment of the subterra- neous reservoir in the chalk basin of London: Min. Proc. Inst. Civil Eng. [vol. 2], 1842, pp. 156, 158; 1843, p. 162. Same as under Buckland. "Salbach (Baurathat Dresden, Saxony), Experiences had during the last twenty-five years with waterworks having an underground source of supply; Trans. Am. Soc. Civil Eng., vol. 30, 1893, p 314. '< slichter, Charles S., Observations on the ground waters of Rio Grande Valley; Water-Sup. and Irr. Paper No. 141, U. S. Geol. Survev, 1905, p 27. i' Min. Proc. Inst. Civil Eng., 1842, pp. 158, 159, 1843, p. 162. FLUCTUATIONS DUE To LAKE \NI> OCEAN LEVEL8. 63 Fuller, Mybon L. Notesonthe hydrology of Cuba: Water-Sup. and In. Paper No. 110, U.S.Geol Burvey, 1906, Records, on p. 189, fluctuations of spring level at Vento, Cuba, due to changes In level "f Almendares River, evidently acting through :i free connection of large size In the limestom . Harris, Gil bert D. Dnderground waters of south Louisiana: Water-Sup, and Err. Paper No. 101, U.S. Geol. Survey, 1904, p. 14; Bull. Louisiana Geol. Survey No. l. L905, p. I. Discusses effect of Mississippi River on level of wells along Us banks. <;i i;n irdt, P. Per Wasserbau, vol. 1, pt. 1, 3d ed., 1892, pp. 48-49. Gives Bess's observations on Aller River, Slichter, Charles S. Observations cm the ground waters of Rio Grand Valley: Water-Sup. and irr. Paper No. ill. CJ.S.Geol. Survey, 1905, pp. 18,25 28,30. Gives observations on effect of outflow from river on ground-water level at El Paso, Mesilla Park, and Berino. The Underflow Of the Arkansas Valley in western Kansas: Water-Sup. and Irr. Paper No. 153, (J. S. Geol. Survey. (lives revitlts of observations on the influence of floods in Arkansas River on the water level in wells at Garden, Sherlock, and Peetiicld, k'ans., in the summer of 1901. Soyka, [sidor. Pie Sch wank unveil des Grundwassers: Penck's Geographische Abhandlungen, vol. 2, pt. 3, 1888. Chapter 3, "DieBeziehungen des Grundwassers zu den oberirdischen Wasserlaufen," contains an excellent discussion of middle European conditions. Tiio.m assay, Raymond. Geologie pratique de la Louisiane, 1860. Contains an entirely fanciful discussion of the seepage of the water from Mississippi River. Todd, .i ames E. Geology and water resources of a portion of southeastern South Dakota: Water-Sup. and Irr. Paper No. 34, U.S. Geol. Survey, 1900, p. 29: Bull. South Dakota Geol. Survey No. 2, 1898, p. 116. Records that many deep wells have a greater discharge when Missouri River is high, and sug- gests that the increased hydrostatic pressure checks the leakage. Veatch, A. C. Geology and underground water resources of northern Louisiana and southern Arkansas: Prof. Paper U. S. Geol. Survey No. 46. Records fluctuations of water level in artesian wells at Fulton, Ark., agreeing with stages of Red River. These are ascribed to pressure of river water acting at the outcrop of the water- hearing sands in river bed several miles from the town. This probably is a case of transmitted pressure, the blue clay over the water-bearing layer acting as a diaphragm and producing fluctu- ations in wells near the river. FLUCTUATIONS 1'RODUtED BY CHAXGES IN LAKE LEVEL. Variations in lake levels < if whatever cause produce fluctuations of the level in wells along their slum's (1 ) by cheeking the rate of outflow when the ground water is draining freely into the lake and (2) by transmitted pressure and deformation as in ocean tides described below. The pressure of deep-seated waters mightalso be slowly affected by the weight of the sediment deposited in the lake beds. This would tend to equalize itself by back flow, and is perhaps a factor of no great importance, except when considered for very long periods. King has, however, obtained a flow artifi- cially by ordinary sedimentation in a tank." FLUCTUATIONS PRODUCED BY VARIATIONS IN THE OCEAN LEVEL- TIDAL WELLS. As partially indicated in the references on page 67, wells and springs which fluctuate with the tide have been observed on nearly all coasts and under many different geo- logic conditions. These fluctuations are produced in three ways: (1) By transmis- sion of pressure through open cavities or passageways affording a free communication between the wells and the ocean; (2) by a checking of the rate of discharge of the normal ground-water flow through porous beds freely connecting with the ocean; and (3) by adeformation of the strata due to the alternate loading and unloading of the tides. In this last case, instead of leakage being an important factor, as it is in the lirst two, the fluctuations are greater the more nearly complete the separation of the oceanic and ground waters. ''Nineteenth Ann. Kept. V. S. Geol. Survey, pt. 2, 1899, pp. 79-80. 64 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. FLUCTUATIONS PRODUCED BY CHANGES IN RATE OF OUTFLOW. The first two classes differ more in the rate at which the change takes place and the character of the zone of influence than in the manner in which it is produced. If the ocean level is raised the first effect is to check' the velocity of outflow; but before any change occurs in the level of near-by wells it is necessary that water accumulate at the point of observation by actually flowing in. • This change, then, is clearly dependent on the same factors which influence the rate of flow, and in underground caverns where the velocity is a question of miles per day this accumulation will be rapid, there will be a relatively short lag, and the distance from the shore to which the rise can be propagated before the water begins to fall will be comparatively great. Its influence will, however, be restricted to wells along limited lines, following the course of the underground passageway. On the other hand, when the water is flow- ing through the interstices of porous strata, where the motion is one of feet per day instead of miles, the accumulation will be slow, the lag proportionately greater, and the zone of influence, while not extending so far from the ocean, will perhaps occupy a larger area, because of its uniform distribution along the coast. When the ocean level falls the reverse will occur, Where there is considerable velocity, as in a cavernous opening, the velocity of outflow retards the effect of the rising tide and hastens that of the falling tide and there is then, as in tidal rivers, a greater lag at low than at high water. When the outflow is slow, as from porous beds, the velocity is not sufficient to exert a very great retarding influence. The fluctuations in porous materials along the seashore are clearly the same, in character and cause as those occurring under similar conditions along river courses (see p. 60), and the same great time lag is to be expected. The difference is only in the very regular periodic nature of the oceanic fluctuations. Nearly all the shallow tidal wells noticed along the seacoasts belong to this class. Such are clearly the tidal wells « reported near Bombay and along the Malabar coast; at Barren Island, in the Andaman Sea; at Perim Island, in the Red Sea; at South Foreland light-house, Kent; the shallow wells at Seagirt, N. J.; and perhaps the wells at Newton Nottage, Glamorganshire, Wales, and Chepstow, Monmouth, England. At Perim Island, in shallow pits at a distance of 60 to 300 feet from the shore, the lag is such that the high water in the wells appears to agree with the low water in the ocean. A similar lag is reported in a well 14 feet deep and 400 feet from the river at Chepstow. At Newton Nottage, where a shallow well 500 feet from the ocean was observed by Man- dan, the lag is but three hours. Experience has shown that wells which are sunk entirely in porous beds near the seacoast should have their bottoms about midway between high and low tide; if they are deeper there is generally an infiltration of salt water. Such wells are commonly dry at low tide, but frequently furnish good supplies of fresh water at high tide, at which time it is necessary to obtain all the water used for domestic and other pur- Wells dependent on underground cavernous openings, such as required by the first case, are quite rare. The fluctuations of the Iceland springs, reported by Hallan de Roueroy, are perhaps connected with such open fissures, though any opinion from the data presented is but a hazard. Perhaps the most remarkable occurrence of tidal fluctuation is that reported at Lille in an artesian well at the citadel. These fluctuations, which amount to 0.415 meter, are referred by Bailly b to the tides of the English Channel, 30 miles away, with a time lag of but eight hours. The only basis on which these fluctuations can now be explained is a supposition of a relatively open connection between the well and the ocean, which, it must be admitted, is a very unsatisfactory hypothesis. a See references, pp. 67-69. bComptes rendus, vol. 14, 1842, pp. 310-314. TIDAL WELLS. 65 TIDAL FLUCTUATIONS IN WELLS PRODUCED BY PLASTIC DEFORMATION. Besides the shallow wells, depending on ordinary porous surficial beds, there are, along ^nd near the seacoast, many deep artesian wells which show tidal fluctuations. In many of these wells there arc clearly no underground caverns involved, the water-bearing beds being ordinary porous strata in which the water Hows through the small interstices at a rate to be expressed in feel rather than miles per day, and in which accumulation or depletion by simple flowage will he corresponding!} slow . There is, moreover, every reason to believe, in some eases where there are thick clay beds above the water-bearing strata, which are known to he continuous lor many miles, that there are no near suboceanic outlets of importance. In case there is some distant outlet it is evident from the slow rate of change shown in the examples given above, where there was a sudden increase or decrease in the volume of river water ( pp. 60, 61 ), that the fluctuation produced by a simple checking or hastening of the rati' of outflow could he propagated but a short distance, and that a long period of time would he necessary for even that. There is, however, in the case of waters under artesian head a new factor intro- duced which is of very great importance. The pressure of the artesian water exerted against the retaining cover, which may he assumed at present to he clay, tends to elevate it, thus placing the clay under an upward stress. The addition of any weight on the surface tends to disturb the equilibrium. If there is no outlet and the weight is applied uniformly, the additional weight can not change the position of any portion of the mass, except to the very slight degree of the elastic compressibility of the water and the soil. If, however, there is any escape for the confined water, such as would be afforded by a well tube, the mass will yield and the water be forced up in the tube. "Were the clay layer perfectly elastic, or in the condition of a stretched elastic membrane above a perfectly mobile body, there would be no time lag, and the water in the well would exactly follow the fluctuations of the ocean w 7 aters; but the clay is not to be regarded as an elastic diaphragm, and the water-bearing sand is not a perfectly mobile body; moreover, for such a deformation to be felt in a well water must be transferred from one point to another, and this involves a time element. The deformation is essentially a plastic one; the clay yields to the super- posed weight and the water is lifted in the well, but if there were no pressure from below the clay could not return to its original position. In the case of tides along the coast only the portion of the clay layer under the ocean is loaded, and that load- ing is a progressive one from a distant point toward the shore. The effect is a defor- mation in which the clay layer is depressed under the ocean and elevated under the land. When the weight is removed the artesian pressure tends to reestablish the old conditions of equilibrum, and the clay layer is lifted under the ocean and sinks under the land. If the artesian pressure is high, compared with the tide when the ocean water com- mences to fall after high title, this pressure lifts the clay quickly and thus tends to shorten the high-tide lag in a near-by well; as the tide falls the high pressure enables the clay to follow the tide closely; at low tide the artesian pressure is clearly in the ascendancy and the clay still rising in the ocean area. As the tide begins to rise it must overcome this artesian pressure before any deformation occurs, and the rising curve in the well therefore lags more behind the tide than the falling curve. Under such conditions the high-tide lag is less than the low-tide lag. Conversely, when the artesian pressure is low compared with the tide, at high tide the feeble artesian pres- sure but slowly lifts the clay weight and the lag is long; at low tide, when the clay diaphragm is high, the greater tidal weight quickly overcomes the feeble resistance of the artesian water and the lag is short; it may then happen that the low-tide lag will be less than the high-tide lag.« It is evident that between the two extremes thus aThis plastic deformation should not be confused with the elastic deformation of the earth which Darwin has considered in his calculations of the effect of tides on seacoasts. He assumes that the ire 155—06 5 66 FLUCTUATIONS OF THE WATER LEVEL IN" WELLS. indicated there are all possible variations, and that the thickness and plasticity of the beds above the water-bearing layers are important modifying factors. The fluctua- tion in a well in such cases does not furnish an exact measure- of the amount of deformation; it furnishes only a fair indication of the variation in pressure at the particular point at which the well is sunk. The maximum effect is felt at the seacoast- near low-tide mark and gradually decreases inward, disappearing in a few miles. It is less if there is leakage from sub- ocean springs, for in such cases the escape of the water decreases that available for the elevation of the water in the tube. * On Long Island the tidal fluctuations observed in the wells at Huntington, Oyster Bay, Long Beach, and Douglaston are clearly of this character, all depending more on the deformation of the overlying layer through tidal load than on changes of dis- charge in leakage. At Huntington (p. 10), Oyster Bay (p. 13), and Douglaston (p. 25) the lag is greater at low than at high tide, as would be expected from the great head and shallow depths, while at Long Beach, where the head is low, the water-bearing sand fine, and the thickness of overlying strata great, the reverse is true (p. 19). . The Oyster Bay observations give the following values: Summary of observations on tidal wells at Oyster Bay, N. Y. Well. Casino ... Burgess . . Lee Underhill Depth. Feet. 93 155 188 114 Distance from ordi- nary high tide. Feet. In water. 50 ■ 100 500 Low-water Minutes.- 12.6 33.4 58.0 75.6 High-water Minutes. 8.0 24.7 42.0 71.8 It will be noticed that the lag here increased with the distance from the shore, and that the low-water lag increased more rapidly with depth than the high-water lag. The cause of the very small difference between the high- and low-water lags in the Underhill well, the one farthest from the shore, is not clear, but it is apparently related to the lessening of the tidal influence with the increasing distance. The Long Beach well is affected both by the tide in the ocean and in the channels behind it, the curve being, as would be expected, a simple resultant of the two stresses. Because Of the shallow bars at the openings the irregularity in the height of the inner low tide is less than in the ocean, and the effect of this difference is shown in the greater regularity of the low-tide heights in the well than in the ocean (PI. IV, p. 20). Here the high-tide lag is one hour and nineteen minutes and the low-tide lag forty minutes, when compared with the ocean, while compared with the tide behind the bar the high-tide lag is practically nothing and the low-tide lag is nearly two hours. To this same class belong nearly all the deep artesian wells along the seacoast which fluctuate with the tide. The phenomenon observed is the result of actual deformation, and the occurrence of tidal fluctuations in deep wells does not, as has been commonly supposed, prove a connection between the water-bearing strata and the ocean. Examples of this type are afforded by the deep wells along the New Jer- earth has an elasticity equal to twice that of the stiffest glass, and the elastic compression produced by loading a sphere of such material of the same size as the earth with a tide ot 5 feet is calculated on the supposition that the ocean is in the shape of a narrow canal. According to this the tides on the Atlantic coast may cause the land to rise and iall as much as 5 inches. See Darwin, G. H., On variations in the vertical due to elasticity of the earth's surface: Brit. Assoc. Rept., 1882, p. 388; Phil. Mag., 5th ser., vol. 14, 1882, pp. 409-427. The Tides, Boston and New York, 1898, pp. 139-143, quoted by Milne, J., Nature, vol. 38, 1883, p. 367, Seismology, London, 1898, pp. 236-237. TIDAL WELLS — BIBLIOORAPBY. 67 Bey coast, the wells at Pensacola, Fla., the deep wells al Greenwich Hospital. Lon- don, ami the wells along the Lincolnshire and Yorkshire coasts. Shelford " has presented a very clear diagram of the conditions on the Lincolnshire coast. This shows overflow springs occurring at the top of a porous layer and the base of an impervious one, and tin- relation between ordinary overflowing and tidal wells, all depending on the same strata. Here, as is almost universally the case, the tidal wells occur only on the shore, and the wells 2 and •"> miles inland are nol affected. In explaining the phenomenon Shelford supposed thai the water found an outlet in Silver Pit, a deep hole in the ocean bed about L8 miles from the coast, which he has represented in his drawing as the ground-water outlet, and that changes in level in the discharge produced the tidal fluctuations. Such a simple change in the rate of discharge could affect the wells IS miles distant only if there were a large open cavernous connection. That there is no such cavern is shown by the fact that, the effect diminishes very rapidly in passing inland, entirely dying out in i' miles. There is no reason why the effect should he propagated IS miles to the coast and then suddenly cease, when tidal wells of the same character penetrating a thick clay bed and obtaining water in the upper pomus layer of the chalk occur along the whole Lincolnshire and Yorkshire coast. In many cases springs near the coast, deriving their supply from the water beneath the clay, are likewise tidal. The cause of this phenomenon in the Bridlington Quay wells, Yorkshire, was correctly given by Inglis in 1817. He recognized in the clay layer a moving diaphragm affected by the tidal pressure from above and the artesian pressure from below. REFERENCES RELATING TO TIDAL FLUCTUATIONS IN WELLS AND SPRINGS. Anonymous. Louden Athenaeum, August, 1860; Jour. Franklin Inst., vol. 72, 1S61, pp. 309-310. States that tides in wells near the sea are universal, and records their occurrence about Bom- hay and along the Malabar coast wherever the material dug through is porous. Wells dug in trap rock are not tidal. Baili.y. Rapport sur les variations observers dans la depense du puits artesien de l'hopital militaire de Lille et dans les hauteurs de la colonne d'eau quand on a interrompu l'ecoulement: Comptes rendus, vol. 14, 1842, pp. 310-314. Gives observations showing that fluctuations, having a range of 0.415 meter, are clearly tidal and occur eight hours behind the tide on the adjacent coast between Dunkerque and Calais. Reports that tidal wells also occur at Noyelle-sur-Mer, Departement de la Somme, and at Ful- ham, London, England. Braithwaite, Fredrick. Min. Proc. Inst. Civil Eng., vol. 9, 1850, p. 168; vol. 14, 1855, pp. 507-522. "At Greenwich Hospital, London, the land springs ebb and flow 2 feet 6 inches, the sand springs, 3 feet, and the chalk springs, 4 feet 6 inches every tide." The total depth of the chalk well referred to is 149 feet. Christie, James. Jour. Franklin Institute, vol. 101, 1901, p. 193. Reports fresh-water well near shore which fluctuates with the tide. Clutterbuck, James. Min. Proc. Inst. Civil Eng., vol. 9, 1850, p. 170. Explains tidal fluctuation in wells on basis of leakage between high- and low-tide marks. Min. Proc. Inst. Civil Eng., vol. 14, 1855, pp. 510-511. Wells at Ramsgate, England, are sunk to half-tide level. These begin to fall at half tide, are dry at low tide, and begin to rise at half tide on the flood. Mm. Proc. Inst. Civil Eng., vol. 19, 1860, p. 33. Reports that wells at Portsmouth, England, are tidal, and concludes that this proves a free connection -with the sea. Desaguliers, Rev. J. T. An attempt to account for the rising and falling of the water of so:ne ponds near the sea, etc. Trans. Phil. Soc. London, No. 384, vol. 33, 1724, p. 132; Trans. Phil. Soe. London from 16G5-1800 (abridged), vol. 7, 1809, pp. 39-41. Reports well at Greenhithe in Kent, between London and Gravesend, which appears to fluctu- ate inversely with the tide. This he explains by imagining a natural siphon. Douglas, James Nichols. Min. Proc. Inst. Civil Eng., vol. 47, 1879, p. 88. Chalk well at South Foreland light-house, Kent, England, 283 feet from face of cliff, 280 feet deep, with bottom level with half tide, has a peculiarity common to many wells of this region in that it is dry at low tide and filled with pure spring water at high tide. aShelford. W., Min. Proc. Inst. Civil Eng., vol. 90, 1887, p. 69. 68 FLUCTUATIONS OF THE WATER LEVEL FN WELLS. Frazer, Persifor. Notes on fresh-water wells of the Atlantic beach: Jour. Franklin Inst., 1890, vol. 130, p. 231. Reports well at Sea Girt, N. J., 20 feet deep, which rises and falls with tide in ocean 150 feet distant Hallan de Roucroy. Comptes rendus, vol. 12, 1841, pp. 1000-1001. States that well at Lille, France, shows tidal fluctuations. Hutton, Capt. F. W. Trans, and Proc. New Zealand Inst., 1895, vol. 28, 1896, p. 655. States that artesian wells at New Brighton are affected by the tide. Inglis, Gavin. On the cause of ebbing and flowing springs [at Bridlington, Yorkshire]: Phil. Mag., vol. 50, 1817, pp. 81-83. ■"When the recess of the ocean lessens the pressure upon the upper surface, the hydraulic pres- sure on the under stratum must raise the whole mass in proportion as the force is superior to the resistance. The return of the tide brings with it the weight and altitude of its mass of water and acts on the flexibility of the clay as a pressure would on a hydraulic blowpipe." King, Fbanklin H. Fluctuations in the level and rate of movement of ground water: Bull. U. S. Weather Bureau No. 5, 1892, pp. 52-53. Suggests that tidal fluctuations may be produced in wells by coastal deformation. JMandan, H. G. Note on ebbing and flowing well at Newton Nottage [Glamorganshire, Wales] : Abs. .Proc. Geol. Soc. London, 1898, pp. 85-86; Nature, vol. 58, 1898, pp. 45-46. Shallow well 500 yards from shore ebbs and flows with the tide; lag about three hours. Mallet, F. R. Ebbing and flowing wells: Nature, vol. 58, 1898, p. 104. Shallow wells in volcanic ash on Barren Island, Andaman Sea, show tidal fluctuations clearly due to retardation of leakage. McCallie, S. W. A preliminary report on the artesian-well systems of Georgia: Bull. Geol. Survey Georgia No. 7, 1898, p. 112. Reports three artesian wells at Tybee Island, near Savannah, Ga., 240 feet deep, one of which is affected by the tide. Moore, H. C. A well intermitting inversely with the ebb and flow of the tide: Trans. Woolhope Nat- uralists Field Club, 1892, pp. 23-24; Jour. Manchester Geol. Soc, vol. 10, 1894, pp. 223-224. Well at Chepstow, Monmouth, fluctuates inversely with the tide. Shallow pits on Perim Island, Straits of Bab el Mandeb, Red Sea, 20 to 100 yards from shore, are full of fresh water at low tide, empty at high tide; explained on basis of time required for filtration. Pearson, Rev. W. Observations on some remarkable wells near the seacoast at Brighthelmstone and other places contiguous. Jour. Nat. Phil. Chem. Arts, vol. 3, 1802, pp. 65-69. States that shallow wells at Brighton fluctuate with the tide, but with a lag of two hours. He ascribes the fluctuation to retardation of leakage. Pliny the Elder. (Caius Plinius Secundus.) Historia Naturalis, lib. 2, cap. 106: (Pliny's Natural History, Bostock and Riley's translation, vol. 1, 1887, pp. 134-135). " There is a small island in the sea opposite the river Timavus, containing warm springs which increase and decrease at the same time with the tide of the sea." Riviere. Comptes rendus, vol. 9, 1839, p. 553. Spring at Givre, canton Montiers-les-Maux, fluctuates with tide. Robert, E. Comptes rendus, vol. 14, 1842, pp. 417-418. Reports that springs near Buder, Olafsen, and Paulsen, Iceland, ebb and flow with the tide. Roberts, Isaac. On the attractive influence of the sun and moon causing tides ... in the under- ground water in porous strata: Rept. Brit. Assoc, 1883, p. 405; see also Proc. Liverpool Geol. Soc, vol. 4, pt. 3, 1881, pp. 233-236. Reports that in a well sunk in Triassic sandstone in which the water rose 60feet above sea level, autographic records showed solar and lunar tides. (See p. 69.) Shelford, W. Min. Proc. Inst. Civil Eng., vol. 90, 1887, p. 68. Describes wells 200 feet deep, on the North Sea, near Louth, Lincolnshire, which fluctuate 3 feet with spring tides. Sinclair, W. F. Ebbing and flowing wells: Nature, vol. 58, 1898, p. 52. Describes well at Alibag, near Bombay, in sand dunes about 25 yards from high-tide mark, . which fluctuated with the tide after heavy rains when the ground water level was high. Tide in well occurred later than that in ocean. Storer, Dr. John. On an ebbing and flowing stream discovered by boring in the harbor of Brid- lington [Yorkshire]: Phil. Trans., vol. 105, pt. 1, 1815, pp. 54-59; Phil. Mag., vol. 45, 1815, pp. 432- 436. S., W. On ebbing and flowing springs: Phil. Mag., vol. 50, 1817, p. 267. States that wells. near Hull under conditions similar to those at Bridlington are not tidal. Trautwine, J. C, jr. Jour. Franklin Inst., vol. 51, 1901, pp. 193-194. Explains tidal wells on basis of free discharge in ocean, as from an open tube; changes in pres- sure at discharge change water level in wells as if they were piezometers along a conduit. Tribus, L. L. Trans. Am. Soc. Civil Eng., vol, 30, 1893, p. 695. Mentions tidal fluctuations in wells at Pensacola, Fla., li miles from the shore front, 4 to 6 inches in diameter, and from 60 to 280 feet deep. Water rises 16 to 17 feet above sea level and GROUND-WATER TIDES GEOLOGIC CAUSES. 69 fluctuates 6 to 10 inches daily with the tide. He supposes, therefore, that thej tap subterranean rivers which have free connection with the ocean. Note: The tides ai Pensacola are rather irregular, with :i small semidiurnal and large diurnal value, mid it is quite possible thai a portion of the fluctuation observed is due to barometric and thermometric changes. Vermeole, C. »'. Water supply Eor wells: Ann. Rept. New Jersey Geol. Survey for 1898, 1899, p. 163. States thai many wells along thecoasl of New Jersey show tides ci irresponding in period, but not in time of occurrence, with the tides oi the ocean, and with a smaller range. Wood, James G. Jour. Manchester Geog. Soc, vol. L0, 1894, pp. 237-239; Abs. Proc. Geol. Soc. London, 1898, p. 86. Reports well near thai described by H. C. Moore (see above), and suggests that well is fed by water coming along fault, which passes under the river; that at high tide this fault is closed, cutting off supply, and at low tide opens again, allowing an influx; and that therefore well fluctu- ates inversely with tide. (Note: A simple leakage would, on accountof slow propagation of change, explain the phenomena quite as well, and more naturally.) Wool. man, LEWIS. Artesian wells in New Jersey: Ann. Rept. New Jersey Geol. Survey for 1898, 1899, pp. 76, 78, 79. Records that the height of water in many artesian wells along the New Jersey coast fluctuates with the tide. At Venter fluctuations were noted in a well 813 feet deep, which had a range of 7j to 1 1.; inches, and a lag of approximately forty-five minutes. Similarly at Avalon, in a well 925 feet deep, the fluctuation observed had a range of from 1(H to 15) inches. Young, Rev. G. and J. Bird. A Geological Survey of the Yorkshire Coast. 4°. Whitby, 1822; 2ded., 1828. Ebbing and Mowing springs, Bridlington, pp. 22-24; intermittent springs, pp. 27-28. TIDES IX THE GROUND WATER PRODUCED BY DIRECT SOLAR AND LUNAR ATTRACTION. The ground water has not an extended level surface like the ocean, where the tides range from nothing to 50 feet, or even the Great Lakes, where the tidal fluctua- tion is but a few inches. The ground-water table is comparatively level only over areas which are but a fraction of the size of the Great Lakes, and direct ground- water tides would be of extremely small size. It seems quite unlikely, therefore, that the fluctuations in the Maghull (Liverpool) well are due to direct solar and lunar attraction, as Roberts" suggests, but, as King& has already pointed out, are rather to be ascribed to the action of the ocean tides on the near-by coast. FLUCTUATIONS DUE TO GEOLOGIC CAUSES. In regions of abundant rainfall the ground-water table is but a subdued reflection, of the surface topography, and any changes in the topography will therefore change the position of the ground-water table. If a stream valley is filled by sedimentation, the ground water is raised over the whole tributary area up to the ground-water divide; if the stream valley is eroded, the water level is in like manner lowered. Similarly, if a lake is produced by a landslip or destroyed by the erosion of its outlet," or the ocean level is changed by orographic movements, the ground-water table like- wise is changed. To these broad generalizations certain exceptions are to be men- tioned. If a river is intrenched in an impervious layer overlain by porous strata, it is evident that the position of the impervious bed in the bank, when the water level in the river is below it, is the factor which determines the position of the ground- water table. A stream may thus lower its bed without affecting the adjacent ground water. Examples of this kind are found in the Isar at Munich and the Salzach at Salzburg, both of which have deepened their beds in recent times, due to regulating works, without lowering the adjacent ground water, because the deepening was entirely in impervious material. ( ' Solution and deposition by percolating waters may cause a gradual depression or elevation of the water level; solution, by increasing the porosity and consequent rate of flow, will enable quite a quantity of water to escape along certain lines and so lower the water level; deposition, in an opposite way, will raise it. a Roberts, Isaac, Rept Brit. Assoc. 1883, 1884, p 405. & King, Franklin H., Bull. IT. S. Weather Bureau No. 5, 1892, p. 54. "Soyka, Penck's Geographische Abhandlungen, vol. 2, pt. 3, Wien, 1888, pp. 60, 63. 70 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. In regions where the rainfall is low and the ground-water table is below the level of the rivers changes in topography naturally have little effect, except where the erosion is sufficient to cut the ground-water table. Generally in such regions the rivers contribute to the ground water by seepage", and the amount so contributed becomes small when the conditions are favorable for the deposition of silt with which the rivers plaster up their beds. During flood periods the rivers scour out the silt and again allow the percolation of water. Earthquakes may produce fluctuations due to several causes: Small fluctuations .may result directly from the earth's tremors; a deformation without faulting may produce changes in pressure; and faulting may make new ground-water outlets which will cause the water in neighboring wells to rise or fall according to their relation to the faulting. Geyser phenomena may produce both periodic and irregular fluctuations of the water level, and Slichter has suggested that the peculiar periodic fluctuations at Uri- sino Station, New South Wales (p. 76) , may be due to such a cause. In this connection it may be well to refer to the hypothetical siphon, or Tantalus- cup arrangement, which the old philosophers gave as an explanation of intermittent springs," a theory which has survived in Houston's Physical Geography, a work still used in the high schools in some parts of this country. & From a geologic stand- point the existence of such a siphon arrangement as this theory postulates may be regarded as almost impossible because of the difficulty of finding an air-tight passage. The fluctuations are now known to be due in many cases to causes not understood at the time this hypothesis was advanced, and in the light of our present knowledge an intermittent spring depending on a natural siphon for its action would be regarded .as a most exceptional phenomenon. It would be necessary to do more than prove that a spring or well ebbs and flows to establish the existence of such a siphon. FLUCTUATIONS PRODUCED BY HUMAN AGENCIES. EFFECT OF SETTLEMENT, DEFORESTATION, AND CULTIVATION ON THE LEVEL OF WATER IN WELLS. It is well known that many hillside springs throughout the entire eastern United States which furnished water when the country was first settled are now dry, that large areas of former marsh land are now in cultivation, and that streams on which boats plied in the early days are no longer navigable. The rainfall records do not indicate that there have been any radical climatic changes, and the changes are clearly the' result of human occupation, c . Part of this is due to the fact that large areas have been artificially drained by tiles, ditches, or absorption pits; beaver dams and other stream obstructions, such as the Great Red River Raft, have been removed, with the consequent drainage of greater or less areas. < l Some of the hillside springs have merely been buried as the soil washed in from the surrounding lands, while others have been affected by the drainage of the lower lands. Different kinds of vegetation use different amounts of water f - and affect the surface a See Regnault, Pere, Phil. Conversations, vol. 2, conversation 6: Dechales, Tract. 17 de Pontibus Naturalibus, etc.; Desaguliers, Rev. J. T., Trans. Phil. Soc. London, No. 384, vol. 33, 1724 (abridged edition Trans. 1665-1800, vol. 7, pp. 39-41) ; Atwell, Joseph, Trans. Phil. Soc. London, No. 424, vol. 37, 1732 (abridged edition Trans. 1665-1800, vol. 7, pp. 544-555). bFor a more recent suggestion of the same theory, see Knightly, T. E., Geol. Mag., n. s., decade 4, vol. 5, 1898, pp. 333-334. cSee, in this connection, King, Franklin H., Bull. U. S. Weather Bureau No. 5, 1892, p. 42; Lane, Alfred C, Water-Sup. and Irr. Paper No. 30, U. S. Geol. Survey, 1899, pp. 54-55. dProf. Paper TJ. S. Geol. Survey No. 46, 1906. eThe literature on the amount of water transpired from plants and evaporated from the earth under different conditions is very extensive, but the results are neither readily comparable nor readily applicable to natural conditions, because of the differing and in many cases unnatural condi- tions under which these experiments have been tried. For a review of the literature, see Harrington, M. W., Review of forest-meteorology observations, and Fernow, B. E., Relation of forest to water supply; Bull. Bureau of Forestry, U. S. Dept. Agric. No. 7, 1873; Lueger, Otto, Die Wasserversorgung der Stadte: Der stadtische Tiefbau, Bd. II, pp. 176-177, 196-205, bibliography, pp. 143-161; Wollny, Ewald, Expt. Station Record, vol. 4, 1893, pp. 531-533; King, F. H., The Soil, New York, 1895, Drainage and Irrigation, New York, 1899. FLUCTUATIONS PRODUCED HV HI'MAN AGENCIES. 71 evaporation in differenl ways, and a change in the plant covering or crop over large areas may clearly result in a broad elevation or lowering of the water level. Simi- larly, certain methods of cultivation conserve more moisture than would find its way into the ground under certain natural conditions, w liile others allow large quantities to flow off tin- surface. Fertilizers and manures affect the rate of percolation in dif- ferent ways; some greatly hasten and others retard the percolation of the soil water. The relation of cultivation to the position of the ground water is therefore very com- plex, ami it is clearly possible to have different results on adjacent fields. In regard to the effect of forested areas on percolation it should he pointed out, on the one hand, that i 1 I a portion of the rain water, varying from 8.5 to oil per cent" of the yearly rainfall, is caught in the crowns of the trees and is evaporated without reach- ing the ground; i '2 ) the absorption capacity of the forest litter and moss is great, and water can be contributed to the ground water only after this is saturated; while the evaporation from this surface is slow, it is to be considered evaporation from a satu- rated surface, and the net result may be greater than from a region where the water sinks rapidly into the ground; (3) the old litter or humus is, according to the experi- ments by Riegler, Ebermeyer, and Wollny, practically impervious, and, while the fresh litter may absorb large quantities of water, the impervious humus or rotted lit- ter prevents the water from reaching the ground water; (4) the roots of the trees in some cases draw' from ground water that is entirely out of the reach of ordinary field plants. Moreover, the direct observations of Ototzky b and Henry and Tolsky c yield the positive result that in Russia and France the level of the ground water is decidedly lower under forests than under cleared land. The results of Ototzky' s observations are summarized in the Experiment Station Record in the following words: This is a translation from the Russian giving the results of a hydrological survey in the steppes region. The conclusion is reached that, physico-geographical conditions being the same, the level of ground water is lower in forests than in adjacent steppes or in general in neighboring open spaces. The level falls as forests are approached, the fall sometimes being very sudden, and it is more marked in case of old forests than new. On the other hand, it should be pointed out that the stream flow from forested regions is more constant than from unforested ones,^ and as this is to be considered as due to ground-water phenomena it indicates a greater percolation. On the plains, groves and hedgerows by acting as wind breaks tend to elevate the water level by decreasing the surface evaporation. It is well known to agriculturists that it is possible to cultivate the soil so that the evaporation will be greater than under natural conditions or so that the moisture will be conserved. It is thus possible to either increase or decrease the ground water by cultivation. In the semiarid region of the Middle West, where the rainfall is light, the secret of the so-called " dry or arid farming" is to so prepare the soil as to insure the percolation of all the rain water or of a very large percentage of it, and to prevent its escape by evaporation. To accomplish this various methods of subsoil- ing, subsurface rolling, and surface mulching, either by pulverizing the soil or by the addition of straw or manure, have been employed, in some cases with marked suc- cess. I am informed by Prof. Charles S. Slichter that in western Kansas, where the Campbell system is employed, the ground water has in places been raised several feet by the increased percolation. a See Bull. Division of Forestry, U. S. Dept. Agric. No. 7, 1893, pp. 100-101, 130-131, and references therein given; also Lueger, Wasserversorgung der Stadte, p. 197. bOtotzky, P., Ann. Sci. Agron. 1897, vol. 2, No. 3, pp. 455-477, pis. 2; review, Expt. Station Record, vol. 9, 1898, p. 1041. i-Henrv, E., and Tolsky, A., Ann. Soc. Agron. 1902, vol. 3, No. 3, pp. 396-422; review, Expt. Station Record, vol. 15, 1903. p. 125. dSee Bull. Division of Forestry, U. S. Dept. Agric, No. 7, 1892, pp. 158-170; Hanson, Marsden, Com- parison of low-water discharge from a timbered with that from a comparatively timberless area: Water-Sup. and Irr. Paper No. 46, TT. S. Geol. Survey. 1901, pp. 40-47. 72 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. EFFECT OF IRRIGATION. Irrigation, almost without exception, raises the ground-water level, and in regions where there is no natural ground- water outlet so placed that it furnishes a sufficient natural escape for the underflow, elaborate systems of tiling and pumping are neces- sary to keep the water level from reaching the surface in the low places and con- verting them into marshes or alkali flats. Carpenter reports that in the Cache la Poudre Valley, Colorado, the water level has been raised 20 to 40 feet.« The effects of irrigation in the King River Valley, California, are shown in Water-Supply and Irrigation Paper No. 58, PI. XXVI. On Long Island only limited areas have so far been irrigated, but these bid fair to rapidly increase. On account of the very porous character of the soil and the fact that all the water used must be obtained from the ground water of the region involved, there is no danger of serious raising of the ground- water level; indeed, the net result here of extreme irrigation, which would have to be done by pumping, would be a lowering of the ground-water level to the extent of the added loss by evaporation and plant transpiration. When the water for irrigation is supplied wholly from springs, as it is at one or two points near Flushing, or where supplied from the city waterworks, as at Elmhurst, 6 the result is a local raising of the ground- water level EFFECT OF DAMS. In regions where the ground water is tributary to the stream channels the effect of the ponding of streams, except where the material of the bed of the reservoir is entirely impervious, is to raise the ground-water level. As the pond or reservoir is relatively permanent, the ground water generally has time to adjust itself to the new conditions, and an elevation is produced which is persistent as long as the reser- voir lasts. Thus on Long Island, where dams were built in all the little streams at an early day, the effect has been to abnormally raise the ground-water level over considerable areas. In mill ponds of this character the use of the water during the day and the accumu- lation during the night give rise to a periodic fluctuation of the water in the wells along their banks which tends to accentuate the temperature effect. In regions where the ground- water table is below the stream, ponding will increase the leakage, though this may naturally check itself in time by the deposition of silt and colloidal material. EFFECT OF UNDERGROUND WATER-SUPPLY DEVELOPMENTS. Underground water is developed for water supply in one of four general ways: (1) By subsurface dams, (2) by infiltration galleries, (3) by pumping from single wells or groups of wells, and (4) by flowing wells. EFFECT OF SUBSURFACE DAMS. In regions where there are valleys with impervious sides filled with porous mate- rial a dam across the valley will pond this underflow and force it to the surface. This has been employed in many regions of the West where dry stream beds with considerable underflow abound. The effect of such a structure on the ground-water level is shown in Water-Supply and Irrigation Paper No. 67, Pis. V, VII. EFFECT OF INFILTRATION GALLERIES. Infiltration galleries may either raise or lower the ground-water level. When con- structed along the line of contact of a pervious and impervious bed they may act in ^Carpenter, L. G., Seepage or return waters from irrigation: Bull. Colorado Expt. Station, No. 33, 1896, p. 4. bBull. Office Expt. Stations, U. S. Dept. Agric., No. 148, 1904. FLUCTUATIONS Dl'K T< > I'l'MI'INO. 73 a way analogous to a subsurface dam. Where deep in pervious layers they offer a new outlet ai a lower level than the natural one, and so depress the water plane. This effect in the same as thai in a pumped well, except that here the cone of depres- sion is greatly Lengthened in one direction. EFFECT OF PUMPING. Toe tirst effect of pumping is to develop a more or less symmetrical cone of depres- sion, of which the well is the center. The steepness and slope of the cone depend mi such factors as the porosity, rate of flow, rate of pumping, and uniformity of soil. The effect of such a depression in the porous material on Long Island is to lower the water in adjacent wells and drain the near-by ponds and marsh areas. a The effect of this lowering of the water table has a marked effect on the stream flow on Long Island, as is shown by the following table, compiled by L. B. Ward: Effect of ground-water pumping in diminishing stream flow, from 187S to 1899, in the old watershed of the Brooklyn waterworks, comparing five-year periods.^ Period. Aver- age annual rain- fall. Average annual rainfall collect- ed, referred to watershed as a whole. Driven-well supply. Area of water- shed. Ex- pressed as rain- fall. Daily per square mile. 1873-1877 1878-1882 1883-1887 1889-1893 189.5-1899 Inches. 43. 33 41.58 43.30 45. 05 43.14 Per cent. 25. 07 29. 60 31. 60 38.43 36.32 Inches. ' 10.86 12.31 13. 68 17. 31 15. 67 Square miles. 52.30 55.14 64. 42 65. 54 66. 44 Inches. ( & ) 2.95 5. 85 7.76 Gallons. (») (") 140,392 278, 383 369, 581 Other pumped sources of supply. Daily total per square mile derived from all sources in the wa- tershed. Water collected as stream flow, referred to 50 square miles of watershed. Period. Ex- pressed as rain- fall. Daily per square mile. Daily per square mile. Expressed as rain- fall. Amount. Propor- tion of total. 1873-1877 Inches. 0.18 .99 2.30 4.17 2.74 Gallons. 8, 659 47, 063 109, 041 198, 605 130, 224 Gallons. 517, 206 585, 978 651, 506 824, 195 745, 983 Gallons. 532, 034 594, 310 518, 071 455, 153 327, 122 Inches. 11.17 12.48 10.88 9.56 6.89 Per cent. 25. 79 1878-1882 30. 02 1883-1887 25.13 1889-1893 21.22 189.5-1899 15. 96 a Merchants' Association of New York, The Water Supply of the City of New York, 1900, p. 186. bBeganinl883. While all pumped wells cause a cone of depression, in regions where the ground water moves rapidly and where the demand does not exceed the supply the recovery is very rapid, as is shown by the figures prepared by W. E. Spear from the records of the Brooklyn water department. ° Several points are noteworthy about these diagrams. The water-bearing strata, in the deep and shallow wells, in each case are separated by rather fine material which may usually be called a clay. There is a distinct flow below the clay layer with a velocity, as shown by Slichter, of 6 feet per day, while that immediately above a See discussion in Prof. Paper U. S. Geo!. Survey No. 44, 1906* pp. 78-79. !>Rept. New York City Commission on Additional Water Supply, 1904. appendix Pis. XI, XII. 74 FLUCTUATIONS OF THE WATER LEVEL FN WELLS. the clay layer is but 18 inches. Near the surface the velocity is 3 to 15 feet per day. It is therefore interesting to notice at Agawam, which is essentially a deep-well station, the sympathetic effect of pumping in the lower layer on the water in the upper. There is also an important difference in the depression produced in well No. 11 and well No. 16. This indicates local irregularities in porosity. At Merrick the wells connected to the suction are all shallow hut one; the effect of pumping is therefore more marked in the shallow than the deep wells. ■ It is here quite normally greatest in the center well. The recovery after pumping is very rapid in both cases, indicating that the supply is a free one and that the plants have not overdrawn it. The records for a 181-foot well at the Queens County Water Company pumping station, near Hewlett, N. Y., show regular fluctuations due to pumping (PL III, p. 18). This well is 3,000 feet from the pumping station and 2,000 feet from the nearest pumped well, and the records showed fluctuations of such a regular rhythmical character that they were at first thought to represent fluctuations due entirely to temperature changes. Further consideration in connection with the record of pump- ing from the station shows that the fluctuations are almost wholly due to pumping, although there is perhaps a slight temperature effect involved. The response of the water level in the deeper wells to changes of pressure at the surface, due to rainfall, tidal, barometric, and thermometric fluctuations, suggests that the removal by pumping of the surface ground water over an artesian stratum will, by relief from load, produce sympathetic fluctuations in deep wells where there is absolutely no connection between the water-bearing strata. EFFECT OF ARTESIAN-WELL DEVELOPMENTS. The universal experience in artesian basins has been that after a time the head decreases. This is due to many causes. When the whole basin is affected, it indicates that the outflow or pumping from the wells exceeds the inflow from the porous strata, and a gradual decrease is to be expected until these factors are balanced. A well or group of wells may be influenced by interference from a single well favorably situated. Thus the drilling of a well in which the outflow is many feet below that of near-by wells quickly affects the head of the higher wells. Very often where but a few wells here and there are affected the decrease in head is to be regarded as wholly due to leakage, either on the outside of the casing or by the failure of the casing through corrosion. All artesian wells are sooner or later pumped, and the effect of the pumping is often to lower the head over wide areas. The diminution of the head in the chalk wells in London during the early part of the nineteenth century is well shown by Clutterbuck.« EFFECT OF LARGE CITIES ON THE WATER LEVEL. Aside from the general lowering of the water level in cities due to pumpage, another factor tending in the same direction is a decrease of the inflow of rain waters. The mass of buildings, paved streets, and drainage systems absolutely prevent the infiltration of rain water over wide areas. This loss is, however, in part replaced by leakage from the water and sewer systems. The loading resulting from the placing of large and heavy buildings on small areas will have the same effect as loading due to any natural causes, except that the former is so gradual that in most cases the water has time to escape laterally. It is conceiv- able, however, that the loading may exceed the rate of outflow and a temporary measurable increase in the artesian head be produced, but this is of such slight value as to be of theoretical rather than practical importance. The effect is practically nothing when the building rests on bed rock, as at New York, and reaches its maxi- aMin. Proe\ Inst. Civil Eng., vol. 9, 1850, pi. vi, p. 180. FLUCTUATIONS DUE TO INDETERMINATE CAUSES. 75 mum al points like New Orleans, Galveston, and other coasl towns underlain by unconsolidated Tertiary :me produced by large fires, which in a short time would remove a large weight from a limited area. These pressure effects would be noticeable only in wells in which the water is already under artesian head and when the overlying beds have considerable plasticity. They would clearly be greatest in unconsolidated materials and would decrease with the thickness df the strata above the water-bearing layer. EFFECT PRODUCED BY LOADED FREIGHT TRAINS. The sensitiveness of the water in wells to any change in load at the surface is strikingly illustrated by the oscillation produced by slowly moving freight trains at .Madison. Wis. This is described by King as follows: a While the self registering instrument was upon well No. 48, it was observed that there were frequent records of sharp short-period curves shown upon the sheet, which at first were supposed to be the result of accidental jars which the instrument sustained; but the frequency of their occurrence and the fact that they always indicated elevations of the water led to a closer scrutiny and their final association with the movement of trains past the well. On the eight-day instrument these fluctua- tions are shown as single dashes. Imt with the one-day form the curve was open. The well in which these disturbances occur is situated about 140 feet from the railroad track and has a depth of 40 feet. It is tubed up with 6-inch iron pipe to the sandstone, 37 feet below the surface, and the water has a mean depth of about 20 feet in it. The strongest rises in the level of the water are produced by the heavily loaded trains, which move rather slowly. A single engine has never been observed to leave a record, and the rapidly moving passenger tains produce only a slight movement, or none at all, which is recorded by the instrument. The figure shows the curve to be produced by a rapid but gradual rise of the water, which is followed by only a slightly less rapid fall to the normal level, there being nothing oscillatory in character indicated by any of the tracings nor observable to the eye when watching the pen while in motion. The downward movement of the pen usually begins when the engine has passed the well by four or five lengths, and when the pen is watched, it may be seen to start and to descend quite gradually, occupying some seconds in the descent. This is very similar to the various pressure effects noted above, due to tidal, baro- metric, and rainfall loading, and to transmitted fluctuations due to variations in local load produced by temperature changes, except that the time of lateral transmission is rather shorter, and it is not clear that the water is under artesian head. FLUCTUATIONS DUE TO INDETERMINATE CAUSES. SMALL FLUCTUATIONS. The extreme susceptibility of the water level in wells to pressure changes would lead one to expect many minute fluctuations; and, indeed, all the well curves show a great number of such fluctuations superposed on the larger fluctuations produced by the dominant element at that point. Many are clearly compound waves of very complex character and represent the resultant of many forces. They emphasize the continued state of unrest of the earth's surface. These fluctuations can be properly studied only with instruments having both a large vertical and time scale, and their elucidation would necessitate corresponding meteorologic instruments of great delicacy. On the day gages at Hewlett (p. 18) there is a distinct series of minor fluctuations with a well-defined period of about twenty minutes. These greatly resemble the minor oscillations in the tidal curves at many points. 6 a Bull. U. S. Weather Bureau No. 5, 1892, pp. 67-68. bSee Airy, Sir G. B., On the seiches or n on tidal undulations of short period at Malta: Phil. Trans. Royal Soc, 1878, pp. 1'23-138 Dawson, W. Bell, Notes on secondary undulations recorded by self- registering tide gages, Trans. Royal Soc. Canada, sec. 3, 1895, pp. 25-26; Illustrations of remarkable secondary tidal undulations in Nova Scotia, Trans. Royal Soc. Canada, sec. 3, 1899, pp. 23-26. Duff, A. W., Secondary undulations shown by recording tide gages; Trans. Nat. Hist. Soc. New Bruns- wick, 1S97; Am. Jour. Sci., 4th ser., vol. 3, 1897, pp. 406-412; Am. Jour. Sci., 4th ser., vol. 12, 1901, pp. 123-139. Denison, F. Napier, The Great Lakes as a sensitive barometer: Canadian Eng., Oct. -Nov., 1*97; Secondary undulations oi tide gages, Proc. Can. Inst., n. s., vol. 1, 1897, pp. 28-31; The Great Lakes as a sensitive barometer: Proc. Can. Inst., n. s., vol. 1, 1897, pp. 55-63; The origin of ocean tidal secondary undulations: Proc. Can. Inst., n. s., vol. 1, 1897, pp. 134-135. 76 FLUCTUATIONS OF THE WATER LEVEL IN WELLS. The secondary oscillations in the tide curve at Swansea, England, have a time interval of fifteen to twenty minutes; at Malta, twenty-one minutes; and at Sydney, twenty-six minutes; while Denison has observed on Lake Huron oscillations with periods of fourteen, eighteen, twenty-two, and forty-five minutes. As* no such secondary tidal oscillations have been observed near Long Island, and as the Hewlett well is at such a distance from the coast that it is not affected by tides 4 feet high, these oscillations are clearly not of transmitted ocean origin. Denison' s observations led him to the conclusion that many of the secondary oscillations are due to baro- metric fluctuations, and the occurrence of these fluctuations in wells must be regarded as strong confirmatory evidence of his conclusion. Besides these fluctuations with a period of twenty minutes, there are several other minor vibrations with smaller amplitudes and periods; one series seems to have a period of five or six minutes, but is not very sharply defined. In the wells atLynbrook (p. 23) minor fluctuations with periods of forty and eighty minutes have been clearly recognized in a mass of still smaller fluctuations. FLUCTUATIONS AT MILLBURN, N. Y. Extremely irregular fluctuations with a range of as much as 1 inch were obtained from a well at Millburn, N. Y. (PI. V, p. 22). These are quite different from any of the other curves obtained and no cause can be assigned for these irregularities. Not the least strange part of the curve is that its general character changes sharply on July 29. (See discussion, pp. 22-23.). FLUCTUATIONS AT URISINO STATION, jSTEW SOUTH WALES. The fluctuations reported by Professor David « at Urisino Station, between Wanaar- ing and Milparinka, in the northwest corner of New South Wales, 200 miles from the ocean, are unique. Two subartesian wells, one 1,680 and the other 2,000 feet deep, in which the water rises to within 15 or 20 feet of the surface, show regular rhyth- mical pulsations with a range of 4 to 5 feet every two hours. That is, there are here six almost equal "tides" of large size every twenty -four hours. Prof. Charles S. Slichter has suggested the very probable explanation that the fluctuations are due to a sort of periodic geyser phenomena. This is quite competent to produce the fluc- tuations observed and the high temperature of the water in this basin lends consider- able color to the suggestion. a David, T. W. E., Notes on artesian water in New South Wales and Queensland: Jour, and Proc. Royal Soc. New South Wales for 1893, vol. 27, pp. 429-430. I N DEX. A. Page. Agawam, N. Y., pumping at, effect of 71 Agram, Hungary, annual fluctuations in well at 41 Air. fluctuations produced by pressure trans- mitted through 7-8,24, 12-43 Airy, G. B., on minor tidal fluctuations at Malta 75 Alfriston, England, annual fluctuations in well at 41 Alrnenderes River, Cuba, fluctuations in springs produced by 63 Aller River. Germany, well fluctuations produced by 60 Ann Arbor. Mich., annual and secular fluc- tuations in well at 40 Annual fluctuations, character and cause of 29-34 ■ latcs of maximum and minimum, fac- tors affecting 34-37 diagrams showing 30,31,32 (PI. IX), 39 Arkansas, well fluctuations produced by Red River in 63 Artesian well developments, effect of, on water level 74 Atwell, Joseph, on fluctuating springs in Devonshire 53 Auchincloss, W. S., on annual and secular fluctuations at Bryn Mawr 38, 51 Avalon, X. J., tidal fluctuations in well at. 69 B. Baden, Austria, annual fluctuations in well at 41 Bailly, , on tidal well at Lille, France.. 64,67 Barbour, E. H., on blowing wells in Ne- braska 53 on annual well fluctuations in Ne- braska 51 Barometric changes, effects of. 7-8, 24-25, 52-54, 76 effects of, bibliography of 53-54 diagram showing 24 (PI. VI) Barren Island (Andaman Sea), tidal wells on 64,68 Berlin, Germany, annual fluctuations in well at • 40 annual rainfall and water-level curves at, figure showing 29 Bettes, C. R., aid of 18-19 Bibliography of fluctuations, due to baro- metric changes 53-54 due to ocean tides 67-69 due to rainfall 51-52 due to rivers 62-63 due to temperature 59 Page. Blowing wells, occurrence and bibliogra- phy of 53 I '.on i Kay, India, tidal wells near 64, 67-68 Bowman, Isaiah, well observations by 13,16 Braithwaite, Frederick, on tidal-well fluctu- ations at London 67 Bremen, Germany, annual fluctuations in well at 40 annual rainfall and water-level curves at. figure showing 29 Brentwood. X. Y., barograph record at, fig- ure showing 18 (PI. Ill), 22 (PI. V),24 (PI. VI) observations at 27 rainfall record at, figure showing. 18 (PL III), 22 (PI. V),24(P1. VI) Bronx Park, N. Y., soil temperatures at, observations on 57, 58 Brooklyn waterworks, effect of pumping at, on stream flow 73 Briinn, Austria, annual fluctuations in well at 40 annual rainfall and water-level curves at, figure showing 29 Bryn Mawr, Pa., annual and secular fluc- tuations in well at 40,51 annual rainfall and water-level curves at. figure showing 30 ground-water curves at 38 figures showing 30, 31 Buckland, Doctor., on London well fluctua- tions 62 C. ("ache la Poudre River, Colo., diurnal fluc- tuations of 59 rise of ground water along 72 Caleves, Switzerland, percolation experi- ments at 46 Callender, H. L., on soil temperatures 58 Capillarity, effect of, in producing well fluctuations 8, 42-43, 57 effect of surface changes on 43 Carpenter, L. G., on fluctuations in Cache la Poudre River, Colo 59 on rise in ground water along Cache la Poudre River, Colo 72 Caterham, England, annual fluctuations in well at 41 Caverns, tidal fluctuations transmitted t h n >ugh 62-64 Celle, Germany, well fluctuations produced by Aller River at 60-61 Charnock, Charles, lysimeter experiments at Ferrybridge by 46 77 78 INDEX. Page. Chelgrove, England, annual fluctuations in well at 41 Cheshire, England, annual fluctuations in well at 41 Christie, James, on tidal fluctuations in wells '. 67 Cities, effect of, on water table 74-75 Citizens Water Supply Co., wells of, fluctua- tions in, plate showing . . 26 (PI. VII) wellsof,locationof,mapshowing. 26 (PI. VII) observations on 25-26 Clutterbuck, James, on lysimeters 48-49 on tidal wells in England 67 on well fluctuations at London 51, 62, 74 Cochituate, Lake, Mass., basin of, rainfall and run-off in 50 Colne River, England, well fluctuations at London due to 62 Colorado, annual fluctuations in infiltration gallery in 51 annual rainfall and water-level curves in, relations of 43 ground-water fluctuations in 43, 59 Connecticut River, Conn., basin of, rainfall and run-off in 50 Consolidated Ice Co., Huntington, N. Y., observations on well of 13 Cretaceous clays, lack of continuity of, on Long Island 10 flgureshowing 9 Cretaceous sands, figure showing 9 water in 10, 19 Croton River, N. Y., basin of, rainfall and run-off in 50 Cuba, spring fluctuations produced by Al- menderes River in 63 Cultivation, changes in ground-water level due to 70-71 Czernowitz, Austria, annual fluctuations in well at 41 D. Dal ton, John, lysimeter experiments of 44-45 Dams, changes in ground-water level due to 72 Darwin, G. H., on elastic deformation of the earth 55-66 David, T. W. E., on well fluctuations in New South Wales 76 Dawson, W. Bell, on secondary tidal fluctua- tions 75 Debreczin, Hungary, annual fluctuations in well at .- 41 Deformation, elastic, of earth, G. H. Dar- win on 65-66 Deformation, plastic, of earth, well fluctua- tions due to 8, 28, 42-43, 62-63, 65-68, 74-75 Denizet, , on spring fluctuations at Voize, France, due to baromet- ric changes 53 Denison, F. Napier, on secondary fluctua- tions of Lake Huron 75-76 Denver Water Co., infiltration gallery of, fluctuations in 51 Desaguliers, J. T., on tidal fluctuations in England 67 Dickinson, John, lysimeter of 44-45 Dickinson, John, and Evans, John, percola- tion experiments of . 32-34, 37, 45-46, 48 percolation experiments of, diagram showing 32 Discharge, point of, distance from, relations of fluctuations and 38, 49, 51 rate of, fluctuations due to 57, 61, 63-64 Douglas, J. N., on tidal well in Kent, Eng- land 67 Douglaston, N. Y., marsh near, description of 25-26 marsh near, map showing 26 (PL VII) mud volcanoes near 25-^6 wells at, description of 25 fluctuations in 26 figure showing 28 (PI. VIII) lag in 25, 66 location of, map showing 26 (PI. VII) observations on 10, 25-26, 66 errors in 18, 26 Drainage, changes in water table due to. . . 70 Dubois, H. J., well record by 12 Duff, A. W., on secondary tidal fluctuations in New Brunswick 75 E. Earthquakes, changes in water table due to... 70 East Rockaway Inlet, tide curves at, figure showing 20 (PI. IV) Eastdean, England, annual fluctuations in well at 41 Ebermeyer, Dr. E., on forest litter 71 on lysimeters 48 Elbe River, Germany, artesian wells in bed Of :.... 62 Emery, F. E., on annual fluctuations in well at Geneva, N. Y 51 England, annual fluctuations in wells in. . . 41, 51 barometric fluctuations in springs and wells in 53-54 blowing well in 53 decrease of head in artesian wells in. . . 74 fluctuations due to rivers in 62 percolation experiments in 32-34, 44-47 tidal fluctuations in wells in 64, 67-69 Erosion, changes in water table due to. 69 Europe, annual and secular fluctuations in wells in 34-35,38,40-41,51,61-62 annual rainfall and water-level curves in, figures showing. 29,32 (PI. IX), 62 Evans and Dickinson. See Dickinson and Evans. Evaporation, loss by 50 Evaporation and rainfall, fluctuations due to 29-42 F. Farrley, T., on blowing well in England. . . 53 Fenhurst, N. Y. See Hewlett, N. Y. Fires, effect of, on water level 75 INDKX. 79 Page. Flood Bows, contribution by ground water to 8, 12,49 definition of 19 effect of showers on 12 Si i also Stream flow. Floral Park, N. Y., observations at 27 rainfall records at, figures showing. . . 18 i PI. III),-.'.' il'l. Vi,-ji (PI. VI) thermograph records at, figures show- ing.. 18(P1. [II),22(P1.V),24(P1.VI) Fluctuations in ground-water level, amount of 7,38 11. 51 causes of 7,28-76 classification of 28 - also Annual fluctuations; Artesian wells; Bibliography; Capillarity; cities; Dams; Deformation; For- ests: Geysers; Geologic changes; Irrigation: Pumping: Rainfall; Secular fluctuations: Showers; Soil; Streams; Tides. Forest, effect of, on ground-water level 71 Fortier, Samuel, on underflow 51 France, barometric fluctuations in springs in : 53 forests in, effect of, on ground-water level 71 percolation experiments in 45 tidal fluctuations in wells in 02, 64, 67-69 Frankfurt, Germany, annual fluctuations in well at 40 annual rainfall and water-level curves in wells at, figures showing 29 Freund, Adolf, report of, on Vienna water- works 51 Frazer, Persiflor, on tidal well at Seagirt, N.J 68 Friez gage, description of 18 Fuller, M. L., on spring fluctuations in Cuba, due to river changes 63 Fulton, Ark., fluctuations due to stream flow in well in 63 G. Gages, direct-reading, description of 10-11 observations with 10-17 self-recording, descr'ntion of 17-18 observations with 18-26 Galleries, infiltration, changes in water table due to 72-73 Gasparin, Doctor, lysimeter experiments of 45 Genesee River, N. Y., basin of, rainfall and run-off in 50 Geneva, N. Y., annual fluctuations in well at 40,51 annual rainfall and water-level curves in well at, figure showing 30 Geneva, Switzerland, percolation experi- ments at 45 Geologic causes, changes in ground-water level due to 69-70 Geological Survey, U. S., observations by. . . 10-27 Georgia, tidal fluctuations in wells in 68 Gerhardt, P., on annual fluctuations 51 on fluctuations due to stream flow 63 Pag Germany, annual and secular fluctuations in wells in 10 percolation experiments in 16 17 well fluctuations due to streams in 60-62 Geysers; well fluctuations due to 70,76 Gilbert and I. awes. Sei La wesand Gilbert Gorlitz, Germany, percolation experimi nl at 40 Gough, John, on barometric fluctuations in Yorkshire wells 53 Graz, Austria, annual fluctuations in well at 10 Greaves, Charles, percolation experiments of 32-33, 16, 18 Green, II., gages of , description of 18 Ground water, definition of 12 topography and, relations of 38,69-70 See also Bibliography; Capillarity; Cities; Dams; Deformation; Forests; Flood flow; Geysers; Geologic changes: Human agencies; Irri- gation; Pumping; Rainfall; Showers; Streams; Stream flow; Temperatures; Tides. Ground-water curves, relation of rainfall curves and, figures showing 18 (PI. Ill), 22 i PLY), 24 (PI. YI), 29, 30, 31, 32 (PI. IX), 36, 39 Ground-water divide, distance from, effect of, on fluctuations 38, 49 figure showing 9 H. Hallan de Roucroy, on spring fluctuations in Iceland 04 tidal fluctuations at Lille, France. . . 68 Harris, G. D., on blowing wells in Louisi- ana 53 on relations of Mississippi River to well fluctuations 03 Headdon, W. P., on effect of showers on ground water 43, 51 Hemel Hempstead, England, annual per- colation and rainfall curves at, figure showing 32 percolation experiments at 32-34, 45-46, 48 Henry, E., and Tolsky, A., on relations of forests and ground water 71 Hess, , on fluctuations due to stream flow 60 Hertfordshire, England, annual fluctuations in well in 41 Hewlett, N. Y., wells at, description of 18-19 wells at, location of, maps showing 9 (PI. I), 16 (PL II) observations on 18-19, 75-76 pumping of, effect of 74 effect of, figure showing.. 18 (PL III) tidal fluctuation in, figure showing. 18 (PL III) Hicksville, N. Y., annual fluctuations in wells at 40 Hudson River, N. Y., basin of, rainfall and run-off in 50 Human agencies, ground-water fluctua- tions due to 70-75 80 INDEX. Page. Humus/ absorptive capacity of 71 Huntington, N. Y., wells at, description of. 10-13 wells at, location of 11 location of, figure showing 11 observations on 10-13 tidal fluctuation in, figure showing. 12 lag in 10,06 Huntington Light and Power Co., well of, location of, figure showing 11 well of, observations on 10-13 record of 12 Huron, Lake, secondary tidal oscillations on , 76 Hutton, P. W., on tidal fluctuations at New Brighton, England 68 I. Iceland, springs in, tidal fluctuations of... 64,68 Infiltration from rivers, fluctuations due to. 60-61 Infiltration galleries, changes in water table due to 51,72-73 Inglis, Gavin, on tidal fluctuations in springs in Yorkshire 68 Innsbruck, Austria, annual fluctuations in well at 40 Irrigation, changes in water table due to . . 72 J. Jaegle, W. C, well record by 20 Japan, barometric fluctuations in well in.. 54 Jgvington, England, annual fluctuations in well at 41 Josephstadt, Austria, annual fluctuations in well at 40 K. King, P. H., citations of . . 42-13,51-56,63,68-69,75 evaporation experiments by 49 Klagenfurt, Austria, annual fluctuations in well at 40 Knightly, T. P., on barometric fluctuations in Derbyshire wells 53 Krakau, Austria, annual fluctuations in well at 40 L. Lag in well fluctuations 13, 17, 19, 21-22, 24, 34-36, 60-62, 64-66 Lake level, changes in, effect of, on water table 63,69 Lane, A. C, on blowing well in Michigan. . 53 Lansing, Mich., annual fluctuations in well at 40 annual rainfall and water-level curves at, figure showing ' 30 Latham, Baldwin, on barometric fluctua- tions in springs in England 53 Lawes, John, and Gilbert, J. H., percolation experiments of 32-34, 37, 47-48 percolation experiments of, figure show- ing 32 Leakage from rivers, effect of, on water level 61-63 Leitha River, Austria, effect of, on ground water 35, 61-62 Lille, France, tidal wells at 62,64,67-69 Lincoln, Nebr., soil temperatures at, observa- • tions on 57 Liverpool, England, barometric and tidal fluctuations in well near 54, 68-69 Liznar, Joseph, on periodic fluctuations in wells 52 London, England, annual and secular fluc- tuations in wells at 41, 51 barometric fluctuations in wells at 53 diminution of artesian head at 74 fluctuations due to rivers at 62 percolation experiments near 46 tidal fluctuations in wells at 67 Long Beach, N. Y., well at, description of 19-20 well at, location of, maps showing. . 9 (PI. I), 16 (PI. II) observations on 10, 19-22 errors in 18 record of '. 20 tidal fluctuations in 18, 21-22, 66 figure showing 20 (PI. IV) lag in 19, 66 Long Island, blowing wells on 53 ground water on, source of 10 hydrologic conditions on 9-10- figure showing 9 similarity of, to those at Wiener Neustadt 35 irrigation on 72 map of southern part of 16 (PL II) map of western end of 9 (PI. I) observations on 10-27 ponds on, effect of 72 pumping on, effect of 73 rainfall curveson.figuresshowing. 18(P1. Ill), 22 (PI. V), 24 (PI. VI), 30,36, 37,39 section of 9 secular fluctuations on 37-38 south side of, rainfall and run-off on. . . 50 stream flow on 10, 50 topography of 9 underflow on 73-74 wells on, fluctuations in 10-27, 35, 37-38, 40, 42-44, 49, 52-53, 57-59, 62, 66, 74-76. fluctuations in, figures showing 12, 16,18 (PL III), 20 (PL IV), 22 (PL V), 24 (PL VI), 26 (PL VII), 28 (PL VIII), 30, 36, 39. observations on 9-27 Louisiana, blowing wells in 53 fluctuations of wells in, due toMississippi River 63 Lueger, Otto, on annual fluctuations in wells 52 on barometric fluctuations in wells and springs 54 Lynbrook, N. Y., wells at, description of . . . 23 wells at, fluctuations in 22, 42-43, 49, 52, 57-59, 62, 76 location of, map showing 9 (PI. I), 16 (PL II) observations on 23-25, 57-59 record of 23 INDKX. 81 I'age. Lyalmeters, descriptions of •"'-'. 1 1 it objections to 44,48 19 observations with 32,34, n 19 results of, figure showing 32 M. McCallie, S. W., on tidal well fluctuations in Georgia (>S McDougal, I). T., on soil temperatures 57-58 Madison, Wis., temperature and well fluc- tuations at 54-57 well at, fluctuations in, due to showers. 42 fluctuations in, figure showing 56 record of 5G Maghull, England, tidal and barometric fluctuations in well at 54,68-69 Malabar coast, tidal wells on 64,67 Mallet. F. R., on tidal fluctuations in wells in Wales 68 Malta, .secondary tidal oscillations at 76 Manchester, England, percolation experi- ments at 45 Mandan, H. G., on tidal wells 64,68 Map of Douglaston and vicinity 26 of Oyster Bay an d vicinity 13, 14 of property of Huntington Light and Power Company and vicinity .. 11 of southern Long Island 16 (PI. II) of western Long Island 9 (PL I) Maurice, , lysimeter experiments of.... 45 Mead, Elwood, gage of, description of 18 Merrick, X. Y., pumping at, effect of 74 Mesilla Park, N. Mex., underflow at 13, 62 Meyer, Cord, aid of 25 Meyer, J. E., aid of 25 Michigan, annual and secular fluctations in wells in 40 blowing well in 53 rainfall and water-level curves in, fig- ures showing 30 water level in, observations on 52 Mill ponds, effect of, on ground-water level 72 Millburn, N. Y., well at, fluctuations in 22- 23, 38, 40, 76 well at, fluctuations in, figures show- ing 22 (PL V) 30,39 location of, map showing 9 (PL I), 16 | PL II) observations on 10, 22-23 rainfall and water-level curves in, figures showing 30, 39 Milne, John, on barometric fluctuations in well in Japan 54 Moore, H. C, on tidal fluctuations in wells. 68 Mud volcanoes, location of, near Douglas- ton, map showing 26 (PL VII) occurrence of, near Douglaston 25-26 Munich, Germany, annual fluctuations in well at 40 annual rainfall and water-level curves in well at, figure showing 29 percolation experiments at 47 Muskingum River, Ohio, basin of, rainfall and run-off in 50 Muskingum River < >hio < Continued. basin of, underflow in 10 Mystic Lake, Mass.. basin of. rainfall and run -oil' in 50 \. Nashua River, Mass., basin of, rainfall ami run-off in 50 Nebraska, annual II net nations in wells in.. ■•! blowing wells in 53 soil t em ] 'em tu res a t, observations on . . 57 Neshaminy Creek, Pa., basin of, rainfall and run-off in 50 New Inlet, X. Y., tide curve at 22 (PI. V) \e» Jersey, annual fluctuations in wells in. 52 artesian wells in, fluctuations in 69 tidal fluctuations in 66-69 New York, annual fluctuations in well at Geneva 51 soil temperatures at Bronx Park, obser- vations on 57-58 See also Long Island. New York City commission on additional water supply, observations of, on Long Island 27 observations of, results of, figure show- ing 18 (PL III) , 22 (PL IV), 26 (PL VII), 36 Newark, N, J., residual-mass curves of ram- fall at 37 Newell, F. H., aid of 17-18 North Allerton, England, blowing well at . 53 O. Oliver, William, on minor periodic fluctua- tions of well in England 54 (•range, France percolation experiments at. 45 ( itotsky, P., on forests and ground water .. 71 Oyster Bay, N. Y., sections at, figures show- ing 14,15 wells at, location of, figures showing. . . 13, 14 observations on 15-17, 66 tidal fluctuations in 17 figure showing 17 P. Paleozoic rocks, occurrence of 10 occurrence of, figure showing 9 Pearson, W., on tidal-well fluctuations in England 68 Pennsylvania, annual fluctuations in well in 51 Pensacola, Fla., fluctuations in wells at 67-69 Pequanock River, Conn., basin of, rainfall and run-off in 50 Percolation, amount of, estimation of. 32-37, 44-51 rainfall and, relations of, figure show- ing 32 stream flow and, relations of 49-50 Perim Island (Red Sea), tidal wells on 64, 68 Perkiomen Creek, Pa., basin of, rainfall and run-off in 50 Philadelphia, Pa., rainfall curve at 31,37 IRR 155—06- 82 INDEX. Plastic deformation. See Deformation, plas- tic. Pleistocene gravels, of Long Island, occur- rence of 10 Pliny the Elder, on barometric fluctuations in wells in Italy 54 on temperature fluctuations in wells in Italy , 59 on tidal fluctuations in wells in Italy . . 68 Pliny the Younger, on barometric fluctua- tions in wells in Italy 54 Point of discharge. See Discharge, Poisenille, — , on temperature and viscosity. 58 Prag, Hungary, annual fluctuations in well at 40 Pressure, transmitted, fluctuations pro- duced by 7-8, 21, 28, 42-43, 62-63, 65-68, 74-75 Pumping, effect of, on ground water 52, 73 effect of, on ground water, plate show- ing 18 (PI. Ill) on stream flow 73 Purdue University, lysimeter experiments at 43 Queens .County Water Co., wells of . . . 18-19, 23-25 wells of, fluctuations in, figures show- ing 18 (PI. Ill), 24 (PI. VI) R. Railway trains, effect of, on water table ... 42, 75 Rainfall, contribution to ground water by. 44-51 effect of, on ground water 7, 24, 29-52 excess of, effect of 37-38 effect of, figure showing 34, 36, 37 figures showing 18 (PI. Ill), 22 (PI. V),24(P1. VI), 29, 30, 31, 32, 36, 37, 39 fluctuations due to 7, 24, 29-52 bibliography of 51-52 percolation of, amount of . 38 effect of 44 observations on 32-34 residual-mass curves of, figures show- ing 32,36,37,39 Statistics of 40-41, 45-47, 50 stream flow and, relations of. 7-8, 10, 24, 49-51 See also Showers. Rainfall curves, relation of ground-water curves and, figures showing.. 18 (PI. Ill), 22 (PI. V), 24 (PI. VI), 29, 30, 31,32,36,39 Rathbun, F. D., work of 18 (PI. Ill), 20 (PI. IV), 22 (PI. V) Red River, Ark. , fluctuations in wells along. 63 Residual mass rainfall curves, figures show- ing .... . 32, 36, 37, 39 Riegler, , on forest litter 71 Rio Grande, relation between water table and bed of 62 Risler, , lysimeter experiments by 46 Rivers. See Streams. Riviere, , on tidal fluctuations in spring at Givre 68 Robert, E., on tidal fluctuations in Iceland springs 68 Roberts, Isaac, on barometric fluctuations in well at Maghull 54 on tidal fluctuations in well at Maghull. 68-69 Rothamsted, England, percolation experi- ments at 32-34, 48 percolation and rainfall curves at, fig- ure showing 32 Run-off, statistics of 50 See also Stream flow; Flood flow. Salbach, B., on artesian wells in Elbe River. 62 Salt water, expulsion of, from formations of Long Island 10 infiltration of, prevention of 64 Salzburg, Austria, annual fluctuations at . . 40 annual rainfall and water-level curves at, figure showing 29 observations at 29 Saturation, zone of, depth of soil above, effect" of, on ground-water fluc- tuations 34-37 Schrieber, Adolf, well of, observations on.. 22-23 Seagirt, N. J., tidal wells at 64, 68 Secular fluctuations, amount of 40 diagrams showing 32, 39 occurrence of 37-38 range of 40 Sedimentation, changes in water table due to 63, 69 Seepage, effects of _ 61-63 Shelf ord, W., on tidal fluctuations in Lin- colnshire wells 67-68 Sherlock, Kans., fluctuation due to temper- ature change at 59 ground- water movement at 60-61 Showers, effect of, on wells 35-37, 42-44, 51 effect of, on wells, diagrams showing... 24 (PI. VI), 36, 39 Sidney, secondary tidal oscillations at 76 Silt in rivers, effect of, on ground water 61, 70 Sinclair, W. F., on tidal fluctuations in Bom- bay wells 68 Siphons, natural, hypothetical, question of. 53,70 Slichter, C. S., aid of 17-18 on fluctuations in ground water due to stream flow 62-63 on ground-water movements in Kansas. 60 on rise of ground water in Kansas 71 on underflow on Long Island 73 on well fluctuations 54 on well fluctuations in New SouthWales. 76 Soil, air in, pressure transmitted to ground water by 7-8, 24, 42-43 depth of, effect of, on ground-water fluc- tuations 34-37 temperature of, relations of well fluctu- ations and 54-59 Solution, changes in water table due to 69 Soyka, Isidor, figures by 29 on annual fluctuations 52 on fluctuations due to rivers 63 Spear, W. E., figures by 36,37 observations by 27, 35, 50 on flood flow 10 INDEX. 83 Page. spear, w. E.— Continued. on fluctuations on Long Island 52 due to pumping 78 Springs, intermitting, occurrence <>f 58, 70 Still box. use of ,; l Storer, John, on tidal fluctuations in wells in Yorkshire 68 stream flow, effect of pumping on 73 estimation of percolation from 49-50 ground water ami, relations of 7-8,49-50 rain fall ami, rela I [l >nsof 7-8, 24, 49-50 Si '" also Flood flow. streams, fluctuations in wells due to 59-63 t! net nations in wells due to, bibliography of - 62-63 infiltration from 61 -62 silt in, effect of 61-62 fluctuations in springs due to 63 plastic deformation due to 62 Streams, silted, relation of ground-water level and 61-62 Subsurface dams, effect of, on ground-water level 72 Sudbury River, Mass., basin of, rainfall and run-off in 50 Sussex, England, annual fluctuations in wells in 41 Swansea, secondary tidal oscillations at ... 76 Swezey, G. D., on soil temperatures in Ne- braska 57 Switzerland, percolation experiments in.. 45-46 Szegedin, Hungary, annual fluctuations in well at 41 T. Temperature changes, depth and, relation of 57-58 effect of, on well.fluetuations. 8,24-25, 51,54-59 bibliography of 59 figure showing 58 nontransmission of, figure showing. 56 Tharoud, Germany, percolation experi- ments at 46 Thomassey, Raymond, on seepage from Mississippi River 63 Tides, effect of, on still box 61 effect of, on well fluctuations.. 8, 10-26,63-69 bibliography of 67-69 observations on 10-26 Todd, J. E., on annual fluctuations 52 on barometric fluctuations 54 on fluctuations in South Dakota wells due to Missouri River 63 Tolsky and Henry. See Henry and Tolsky. Topography, relations of ground-watertable and 38 Trautwine, J. C, jr., on tidal fluctuations in wells 68 Tribus, L. L., on fluctuations in New Jersey wells due to rainfall 52 on tidal fluctuations in Florida 68-69 Trieste, Austria, annual fluctuations in wells at 40 Tybee Island, Georgia, tidal fluctuations in well at 68 Pagi Underflow, loss by >'• United States, rainfall and ground-water curves in, figures showing..:... 30,81 UrUSinO Station, New South Wales, well 1] net nations at 76 V. Valley Stream, X. Y., observations on well fluctuation at 10 See also Lynbrook and Hewlett. Van Nostrand, D. L., on depth of mud near Douglasti m 25 Veatch, A.i'., mi Arkansas and Louisiana fluctuations due to streams 63 on Long Island blowing wells 53 Vento, Cuba, spring fluctuations at, due to Almenderes River 63 Ventor, M. J., on tidal fluctuations in well at 69 Vermeule, C. C, on tidal fluctuations in wells in New Jersey 69 Yon Mollendorf, G., lysimeter experiments by 46 \Y. Wales, tidal fluctuations in 68 Ward. L. B.. on pumping on Long Island.. 73 Wells. See Annual fluctuations: Artesian wells; Bibliography; Capillar- ity; Cities; Dams: Deforma- tion; Forests; Geysers; Geologic changes; Irrigation; Pumping; Rainfall; Secular fluctuations; Showers; Soil; Streams; Tides. Wells, blowing, occurrence and bibliogra- phy of 53 Wells, sea-coast, peculiarities of 64,67-68 Whitney, F. L.. work of 18 (PI. Ill), 22 (PI. V),24 (PI. YD, 26,28 (PI. VIII) Wiener Neustadt, Austria, water level at, compared with that on Long Island 35 water level at, fluctuations of 38,41,51 fluctuations of, figure showing 32, (PL IX) observations on 34-35, 38 percolation and, relations of 41 Wisconsin, barometric fluctuations in wells in 53-57 Woldrich, J. N., on effect of rainfall on ground water 29, 52 Wolff, H. C, observations by 59 Wollny, E., lysimeter experiments by 47 on forest litter 71 Wood, J. G., on tidal fluctuations 69 Woolman, Lewis, on tidal fluctuations in New Jersey wells 69 Y. Y'orkshire, England, percolation experi- ments in 46 Y'oung, G. and J. Bird, on tidal fluctuations in Y'orkshire wells 69 CLASSIFICATION OF THE PUBLICATIONS OF THE UNITED STATES GEOLOGICAL SURVEY. [Water-Supply Paper No. 155.] The serial publications of the United States Geological Survey consist of ( 1 ) Annual Reports, (2) Monographs, (3) Professional Papers, (4) Bulletins, (5) Mineral Resources, (6) Water-Supply 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 publications may 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 obtained, free of charge (except classes 2, 7, and 8), on application. 2. A certain number are delivered to Senators and Representatives in Congress for distribution. 3. Other copies are deposited with the Superintendent of Documents, Washington, D. G, from whom they may be had at prices slightly 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 ol 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; O, Underground waters; P, Hydrographic progress reports. This paper is the fifty-second in Series O, the complete list of which follows (PP=Pro- fessional Paper; B=Bulletin; WS= Water-Supply Paper) : SERIES O, UNDERGROUND WATERS. WS 4. A reconnaissance in southeastern Washington, by I. C Russell. 1S97. 96 pp., 7 pis (Out of stock.) WS 6. Underground waters of southwestern Kansas, by Erasmus Haworth. 1897. 65 pp., 12 pis. (Out of stock.) WS 7. Seepage waters of northern Utah, by Samuel Fortier. 1897. 50 pp., 3 pis (Out of stock.) 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. 61 pp. (Out of stock.) W's ho. Waterresourcesof 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 31. Geology and water resources of a portion of southeastern South Dakota, by J. E. Todd. 1900. 31 pp., 19 pis. WS 53. Geology and water resources of Nez Perces County, Idaho, Pt. I, by I. C. Russell. 1901. 86 pp., 10 pis. (Out 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.) I II SEEIES LIST. WS 55. Geology and water resources of a portion of Yakima County, Wash., by G. 0. 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 Waldemar Lindgren. 1903. 62 pp., 4 pis. WS 78. Preliminary report on artesian basin 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. 6§ 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 C. M. 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. Puller. 1904. 98 pp., 11 pis. WS 102. Contributions to the hydrology of eastern United States, 1903, by M. L. Puller. 1904. 522 pp. WS 104. Underground waters of Gila Valley, Arizona, by W. T. Lee. 1904. 71 pp., 5 pis. ■ WS 110. Contributions to the hydrology of eastern United States, 1904; M. L. Puller, geologist in charge. 1904. 211 pp., 5 pis. PP 32. Geology and underground water resources of the central Great Plains, by N. H. Darton. 1904. 433 pp., 72 pis. (Out of stock.) WS 111. Preliminary report on underground waters of Washington, by Henry Landes. 1904. 85 pp., lpl. 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. Puller, geologist in charge. 1904. 285 pp., 18 pis. WS 118. Geology and water resources of east-central Washington, by P. 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. Puller. 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 the Salt River Valley, by W. T. Lee. 1905. 194 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 and others. 1905. 394 pp., 34 pis. WS 137. Development of underground waters in the eastern coastal plain region of southern California, by W. C. Mendenhall. 1905. 140 pp., 7 pis. WS 138. Development of underground waters in the central coastal plain region of southern California, by W. C. Mendenhall. 1905. 162 pp., 5 pis. WS 139. Development of underground waters in the western coastal plain region of southern California, by W. C. Mendenhall. 1905. 105 pp., 7 pis. WS 140. Field measurements of the rate of movement of underground waters, by C. S. Slichter. 1905 122 pp., 15 pis. WS 141. Observations on the ground waters of Rio Grande Valley, by C. S. Slichter. 1905. 83 pp., 5 pis. WS 142. Hydrology of San Bernardino Valley, California, by W. C. Mendenhall. 1905. 124 pp., 13 pis. WS 145. Contributions to the hydrology of eastern United States; M. L. Fuller, geologist in charge. "1905. 220 pp., 6 pis. WS 148. Geology and water resources of Oklahoma, by C. N. Gould. 1905. 178 pp., 22 pis. WS 149. Preliminary list of deep borings in the United States, second edition, with additions, by N. H. Darton. 1905. 175 pp. PP 46. Geology and underground water resources of northern Louisiana and southern Arkansas, by A. C. Veatch. 1906. WS 153. The underflow in Arkansas Valley in western Kansas, by C. S. Slichter. 1906. 90 pp. SERIES LIST. Ill \vs 154. The geology and water resources of the eastern portion of the Panhandle of Texas, by C. N. • Gould. 1906. 64 pp., 15 pis. \vs 155. Fluctuations of the water level In wells, with special reference to Long Island, New York, by A. C. Watch. 1906. 83 pp. The following papers also relate to this subject: Underground waters of Arkansas Valley in eastern Colorado, by G. K. Gilbert, in Seventeenth Annual, I't. II; I'reliminan report on artesian waters oi a portion of the Dakotas, by X. II. Darton, in Seventeenth Annual. Pt. II; Water resources of Illinois, by Frank l.everctt. in Seventeenth Annual. I't. II; Water resources of Indiana and Ohio, by Frank Leverett, to Eighteenth Annual, Pt. IV; New developments in well boring and irrigation in eastern south Dakota, by N. II. Darton, in Eighteenth Annual, I't. IV; Rock waters of Ohio, by Edward Orton, in Nineteenth Annual, I't. IV. Artesian-well prospects in the Atlantic coastal plain region, by X. II. Darton, Bulletin No. 138. Correspondence should be addressed to The Director, United States Geological Survey, Washington, D. C. June, 1906. O LB Mr '08 I 1 1 1 1 1 LIBRARY OF CONGRESS