An artificial overflawtitg TFell. 5 77 



and it was not thought fafe to fink any deeper. A double thicknefs of fteining was madS 

 about fix feet from the bottom upwards, and a borer of 5-!^ inches diameter was made ufc 

 of. A copper pipe of the fame diameter with the borer was driven down the bore hole to 

 the depth of 24, at which depth the borer pierced through the rock into the water ; and by 

 the manner of its going through it muft probably have broken into a ftratum containing 

 water and fand. At the time the borer burft through, the top of the copper pipe was about 

 three feet above the bottom of the well : a mixture of fand and water inftantly rufhed in 

 through the aperture of the pipe. This happened about two o'clock in the afternoon, and by 

 twenty minutes paft three o'clock the water of the well ftood within 17 feet of the furface. 

 The water rofe the firft 124 feet in eleven minutes, and the remaining 119 feet in one hour 

 and nine minutes. The next day fcveral buckets of water were drawn out, fo as to lower 

 the water four or five feet ; and in a (hort time the water again rofe within 1 7 feet of the fur- 

 face. A found line was then let down into the well in order to try its depth. To our 

 great furprife, the well was not found by 96 feet fo deep as it had been meafured before the 

 water was in it ; and the lead brought up a fufficient quantity of fand to explain the reafon 

 of this difference, by fhewing that the water had brought along with it 96 feet of fand into 

 the well. Whether the copper pipe remained full of fand or not, is not eafy to be deter- 

 mined ; but I fhould rather be inclined to think it did not. 



After the well had continued in the fame flate fevcral days, the water was drawn outfo as 

 to lower it eight or ten feet; and it' did not rife a ain by about a foot fo high as it had rifen 

 before. At fome days interval, water was again, drawn out, fo as to lower the water as be- 

 fore J which at each time of drawing rofe lefs and lefs, until after fome confiderable time it- 

 would rife no more ; and the water being then all drawn out^ the fand remained perfectly 

 dry and hard. I now began to think tbe water loft ; and confequendy that all the labour 

 and expence of finking this well, which by this time were pretty confiderable, had been in< 

 vain. There remained no alternative but to endeavour to recover it by getting out the" 

 fend, or all that had been done would be ufelefs ^ and although it became a more difficuit 

 tafk than finking a new well might have been, yet I determined to undertake it, becaufe t* 

 knew another well might alfo be liable to be filled vf\t\\ fand in the fame manner that this 

 was. The operation of digging was again neceflarily reforted to, and the fand was drawn 

 up in buckets until about 60 feet of it were drawn out ; confequently there remained only 

 36 feet of fand in the well : that being too light to keep the water down, in an inftant it 

 forced again into the well with the fame violence it had done before j. and the man ,.ho wa^ 

 at the bottom getting out the fand was drawn up almoft fufFocated, having been covered all 

 over by a mixture of fand and water. In a fhort time the water rofe again within ly feet of 

 the furface, and then ceafed to rife as before. When the water had ceafed riiing the found- 

 ing line was again let down, and the well was found to contain full as much fand as it did the 

 firft time of the water's coming into it. 



Any fuit'ier attempt towards recovering the water appeared- now in vain j and mr.ft peo- 

 ple vv<«ild, 1 believe, have abandoned the undertaking. I again coniidered, that tne labour 

 and the expence would be all loft by fo doing ; and I determined without delay to tet about 

 dr.iwing the fand out through the water by means of an iron box"' made for that purpofe,> 

 without giving it time to harden as before. The labour attending on this operation was 

 ¥ery great, as it was neceflary continually to draw outthe fand, and thereby to prevent the' 



iand 



