WATER. 



the Rrezmper fecond in feet will be 6.5 times the fquare root 

 of the height. , • • 



If the height be meafured in inches, then the velocity in 

 feet per fecond will be 1.88 times the fquare root of the 

 height, nearly. It will be eafy to put the funnel into the 

 moil rapid part of the ftream, by moving it about to different 

 places, until the difference of altitude in the two tubes be- 

 comes the greateft. In fome cafes, it will happen that the 

 immerfion of the inftrument will produce a httle eddy in the 

 water, and thus difturb the accuracy of the obfervation ; 

 but keeping the inflrument immerfed only a few feconds 

 will correft this. The wind alfo would affeft the accuracy 

 of the experiments ; it is therefore advifeable to make them 

 when there is little or no wind. 



By means of this inftrument, the velocity of water at va- 

 rious depths in a canal or river may be found with tolerable 

 accuracy, and a mean of the whole drawn. Where great 

 accuracy is not required, the bent tube with the funnel at 

 bottom will alone be fufficient, becaufe the furface of the 

 water will be indicated with tolerable precifion, by that 

 part of the prifmatic frame for the tube which has been 

 moiflened by the immerfion. 



M. Pitot likewife propofed that a fimilar inflrument fhould 

 be ufed inflead of a log, to determine the rate at which a 

 Ihip fails. For this purpofe, in the middle of a veffel, or as 

 near as can be to the centre of its ofcillations, place two 

 tubes of metal of three or four lines in diameter, one of 

 them being ftraight, and the other bent at bottom and 

 enlarged into a conical funnel. The lower ends of both 

 are to dip into the water in which the veffel fails, and 

 there will be no evil to apprehend from orifices fo minute. 

 Into thefe metallic tubes, two others are clofely fitted at 

 a convenient height for the obfervations. The water will 

 rife, in the firft of thefe tubes, up to its level on the outfide 

 of the fhip ; and in the fecond, up to a certain height, which 

 will indicate as above the velocity of the veffel. For the 

 funnel being turned towards the prow of the fhip, it will, in 

 confequence of the motion, be affefted in hke manner, as 

 if it were plunged into the ftream of a running water. The 

 aftual velocity of the veffel is found by the fame rules as 

 that of ihe current. This method has been repropofed in 

 this country, without any acknowledgments to M. Pitot. 

 We do not, however, recommend its adoption on board a 

 (hip ; for, notwithftanding its theoretical ingenuity, it is hable 

 to many fources of error in praftice, and would not, it is 

 probable, furnifh more accurate meafures of a ihip's way, 

 than thofe deduced from the common log. 



In the praftical ufe of M. Pitot's inftrument, a great 

 difficulty is experienced from the ofcillations of the water 

 in the tubes, which it is not eafy to prevent, and a mean 

 height of the ofcillating water muft be taken. 



M. Du Buat made trials of the inftrument, and found it 

 could not be trufted for any other purpofes than to give the 

 ratios of different velocities. He found the inftrument was 

 better without the ftraight tube, and he employed only one 

 tube with its lower end turned horizontally, in the direc- 

 tion of the ifream, it was made of tinned plate inflead of 

 glafs, and fuf&ciently large to admit a float to fhew the 

 height of the water in the tube. Inftead of making the end 

 of the tube an open trumpet -mouth, he ufed to clofe it by 

 s flat plate, with a fmall perforation in the centre to admit 

 the water through it, or in fome cafes feveral fmall perfora- 

 tions. In this way, the water will rife in the tube, juft the 

 fame as if it was open ; but the ofcillations of the column 

 will be avoided, or greatly diminifhed. 



The hydraulic quadrant has been recommended by feve- 

 ral authors, for meafuring the velocity of water. 

 I? 



It confifts of a fmall quadrant with a divided arch, and 

 having two threads moving round its centre. One of thefe 

 is fhort, and carries a plummet which always hangs in air, 

 and ferves to place the quadrant in its true pofition. The 

 other thread is longer, and carries a weight whofe fpecific 

 gravity is greater than that of water, and which plunges 

 more or lefs deep in the current as the thread is lengthened. 

 The inftrument is held over the water, fo that the plummet 

 of the long thread hangs in the water, and the force of the 

 current will remove it from the perpendicular, whilft the an- 

 gular diftance from the other thread, which is a vertical 

 hne can be afcertained by the divifions on the arch of the 

 quadrant ; the quantity of this deviation from the perpendi- 

 cular is the meafure of the force, and confequently of the 

 velocity of the current. Boffut has fhewn, that the force 

 of the current is as the tangent of the angle which one 

 thread makes with the other, and gives direftions for ufing 

 this inftrument to try a current at different depths. 



Dr. Brewfter, in his edition of Fergufon's leftures, re- 

 commends a fmall and light wheel, like an underfhot water- 

 wheel, with float-boards on its circumference. It is provided 

 with an apparatus to afcertain and record the number of 

 turns it makes, and is held in the ftream, fo that the water 

 may aft upon the float -boards to turn it round ; and from the 

 number of turns it makes in any given time, the velocity 

 of the ftream may be computed. He direfts the wheel 

 to be made of the lighteft materials, and about ten or 

 twelve inches in diameter : it is furniftied with four- 

 teen or fixteen float -boards. The centre of the wheel is 

 perforated with a hole, and tapped to receive a delicate 

 fcrew or wire, which forms the axis upon which it re- 

 volves, with as little friftion as pofTible. At each end 

 of the fcrew or axis, is a handle to hold it by, and to fup- 

 port the wheel ; and to one of thefe handles an index is 

 fixed, pointing to divifions on the circumference of the 

 wheel, which confift of 100 parts. This index fhew» 

 the aliquot parts of a revolution, whilft the number of 

 threads which the wheel advances on the fcrew fhews the 

 number of whole turns it makes. 



To prepare this inftrument for ufe, the wheel muft be 

 turned round upon the fcrew until it arrives quite at one end 

 of it, and till the index points to zero of the divifions on the 

 rim of the wheel ; then hold the axis or fcrew horizontally 

 by the two handles, fo that the floats dip in the water and 

 turn the wheel round upon the fcrew. 



By means of a ftop-watch, or a pendulum, find how 

 many revolutions of the wheel are performed in a given 

 time, a minute, for inftance. Multiply the mean cir- 

 cumference of the wheel, /. e. the circumference deduced 

 from the mean radius, meafured from the centre of im- 

 pulfion upon the float-boards to the centre of the wheel, 

 by the number of revolutions, and the produft will be the 

 number of feet which the water moves through in the given 

 time. On account of the friftion of the fcrew, the refift- 

 ance of the air, and the weight of the wheel, its circum- 

 ference, will move with a velocity a little lefs than that of 

 the flream ; but the diminution arifing from thefe caufes, 

 may be eftimated with fufficient precifion for all the pur- 

 pofes of the praftical mechanic. 



This, we think, is one of the bell ftream-meafurers, becaufe 

 it will give a correft meafure of the motion at the furface 

 of the water ; but it will not give the velocities at the dif- 

 ferent depths beneath the furface, nor do we know any 

 machine which vnU. effeftually anfwer that purpofe. 



By means of this inftrument, we can obtain the velocity 

 of the furface with greater accuracy than perhaps by any 

 other means ; but to afcertain the quantity of water which 



ftiaU 



