eee 27 F726 L AN INQUIRY ye FC) VA MORE PERFECT FORM OF WATER-WHEEL BY J. P. FRIZELL | i. BOSTON 60 CONGRESS STREET 1897 tga. pe? oo Se arc ‘Boker, ; s ‘ Tro - An Inquiry ase toma More Perfect Form of Water-Wheel A WATER-WHEEL is a wheel for applying the power of water to some useful purpose. Water-power implies two things: a flow of water and a fall. We will express the former generally by the symbol g in cubic feet per second, the latter by h in feet. Work is mechanical effect. It implies resistance overcome, and is conveniently represented by the raising of a weight. y t Yi Z Draft-tube SJfJu eae (ZIZILLLL LLL LLL LLL LAY A 43 SS OS ZZZZZZLE Fig 17 as in the ordinary case. Any quantity between the capacity of both wheels and the capacity of the small wheel can be used with little loss. Below the latter quantity the defect appears in full force. The form of bucket adopted in this wheel, although of good efficiency at full gate, is one of the worst forms extant at part gate. However small the quantity of water admitted to the wheel, it is certain to issue in a uniform stream from the orifices of discharge with 26 A More Perfect Form of Water - Wheel. greatly diminished velocity. This wheel is also liable to losses from which ordinary types are exempt. A turbine, in order to give its best effect, must run with a velocity which bears a certain ratio to that due the head. The connections must be such as to admit of this velocity when the shafting and machinery run at their normal speed. Two wheels of different diameter on the same shaft, cannot both run at this velocity. Moreover the wheel out of action is kept in rotation as a useless drag upon the other. Fig. 18 5. In the forms of wheel shown in Figures 18 and 19, the guide serves the double purpose of guide and gate. It controls the width of the guide passages by turning upon a hinge or joint under the action of the regulator. In 18, the outer end of the guide is fixed; the inner end is susceptible of an outward movement, limited by the adjoining guide. In 19, the inner end is jointed, the outer end rotates. The passages are closed when each guide touches the adjoining guide. The same remark may be made of this arrangement as of the Swain wheel. It obviates the loss incident to a change of direction and velocity of the eiftering water. But it in no way affects the second source of loss, viz. that arising from the expansion of the ~l A More Perfect Form of Water - Wheel. 2 stream in the wheel passages. To avoid this latter loss would require the wheel passages to be contracted in the same proportion as the guide passages, when the discharge is reduced. From this brief survey of the present state of the art of wheel- building we are, I think, justified in asserting that no existing form of wheel is free from grave, inherent and unavoidable defects, defects which are material at full discharge, and become more and more marked as the discharge diminishes. No existing form is consistent in design with the highest degree of efficiency or with the well estab- lished principles of hydraulics. They have resulted from tentative methods and from partial and incomplete knowledge, not from a thorough and comprehensive study of the whole subject. It appears to me, also, that the only hope of developing a perfect water-wheel lies in a radical departure from existing forms, every one of which is intrinsically defective. The whole subject of improvement turns upon this question: Are the guides necessary and indispensable to the efficient action of the wheel? Iam convinced that they are not. It appears to me that in a wheel surrounded by a free space sufficient to allow the water to 28 A More Perfect Form of Water - Wheel. move without constraint, it would naturally take the direction and velocity most conducive to the efficient action of the wheel. ‘This, in a center-vent wheel, means a velocity, the tangential component of which is equal to the velocity of the exterior circumference of the wheel. What reason have we for supposing that the water will take that velocity’?* We have the very simple and satisfactory reason that the water cannot otherwise enter the wheel. ‘This answer, how- ever, carries another question with it, viz.: Though there is no doubt that the water would take the required velocity, would not that move; meut be attended with serious loss of energy? I think not, for these reasons : — K Fig. 2O 1. When an orifice is opened for the escape of water under press- ure, the water will approach the orifice. If the orifice recedes, the water will follow and overtake it. 2. A movement of rotation, under the conditions supposed, is in accordance with the natural tendency of the water. Water in a cir- cular vessel, discharging through a central orifice, spontaneously takes a movement of rotation.+ * This would be true for an orifice opening normally, as 7, Fig. 20, or backwards, as atk. It would not be true for an orifice opening toward the direction of rotation, as @. In that case, tle water would be “ scooped” into the wheel without taking the full ve- locity of the wheel. + When a heavy particle (i.e. a particle having weight) moves freely under the action of a force directed toward a fixed point, the line joining it with the fixed point describes equal areas in equal times. This general proposition appears from Fig. 21. Suppose a A More Perfect Form of Water - Wheel. 29 3. It is a universal principle of nature that every movement is performed with the minimum expenditure of energy. Water, in the case supposed, will enter the orifices of the wheel with the least loss of energy possible under the existing conditions. Now we know that if the water reaches the orifice with the rotatory velocity of the wheel, the loss of energy will be slight; and since we know that ex- isting conditions admit of this velocity, and that the actual loss will be as small as is consistent with existing conditions, we are entitled to assume that the actual loss will be slight. 4. If we assume that the water enters the wheel with a less ve- locity than that of the floats, and suddenly acquires the motion of the latter, this sudden accession of velocity does not necessarily im- ply any loss of energy. Loss of energy may occur in several ways. D Cc B A Fig. ra | heavy particle to move with uniform velocity.in the line A D,— A B, BC, CD, being the equal distances moved in successive elements of time. Let O be any point whatever. The triangles A OL, BOC, COD, are the areas described in equal times with reference to 0. These triangles are all equal to each other, having equal bases and the common height A H = the perpendicular distance of O from A D. Now introduce the supposition that, at @, the particle is acted on by a force directed toward QO, which would cause it to move a distance ( / in the element of time. At D draw DG, parallel to CO, and on it lay off DE=CF. Draw CE, then C EO is the area described in the element of time under consideration. Join E O, and draw OG, perpendicular to DG. The triangle C DZ = O D £, both having the base D /, and the common altitude 0G. The triangle CH O=CDO—(CDE—DKE)+O0DE—DKE=CDO. There- fore, the areas described in successive elements of time are equal; and in order to make these areas equal, the velocity must increase as the particle approaches the fixed point. This is the law discovered by Kepler, and it governs the motions of the planets. A particle of water in the wheel-case is under conditions very similar to those of a planet in free space. It enters the case with a certain velocity. As soon as it enters, it is acted on by a force directed toward a fixed point, i.e. the center of the wheel. It does not move straight toward the wheel, but circles around it with an increasing velocity. 30 A More Perfect Form of Water- Wheel. One of these is the communication to the water of motions other than those necessary for its entrance into the wheel, i.e. useless movements. Commotions in water commonly arise from irregulari- ties in the channels, or in the circumstances of movement. No such irregularity exists here. Loss of energy occurs when water passes orifices with contraction. No appreciable loss can occur here from that cause. ‘The contraction, if any, is only that due to the radial component of the velocity, which is not over one-fifth of the tangen- tial component. Each orifice of entrance has the contraction sup- pressed on one side, and it is rounded on two others. On the fourth side, contraction can only exist in virtue of the excess of the wheel’s velocity over the tangential component of the water’s velocity. Nei- ther contraction nor commotion therefore can exist as a source of serious loss. The most natural conception of the phenomenon is this: The water which has passed the tips of the floats is in motion with the full ve- locity of the wheel, the next outside film a little slower, the next slower still, etc., the acceleration being communicated by fluid fric- tion. The loss of energy consists in the fact that, of two consecu- tive films of water, the one nearest the wheel moves a little faster than the one more remote, so that the energy expended is not fully represented by the velocity acquired. ‘The energy represented by the velocity imparted to the film of water is not lost. The loss, however, from fluid friction in this wheel is no greater than in any wheel of equal surface (speaking now of the general surface of the wheel, not that containing the orifices) and equal velocity. In any wheel, the film in contact with the wheel moves with the velocity of the wheel, the adjoining film a little slower, the next slower still, etc., precisely as in this. I have, therefore, become convinced that the guides, although in existing forms of wheel, necessary for constructive reasons, are in no sense essential to the efficient action of the wheel. That they are attempts to force upon the water a direction and velocity which it would take spontaneously if relieved of constraint. That, ordina- rily, they but imperfectly fulfill their purpose at full gate, and are a prolific source of waste at part gate. The first step in the improve- ment of the water-wheel, therefore, is, as it appears to me, to dis- pense with the guides, and to adopt a form of wheel that does not require them for constructive reasons. The second essential to the perfect action of the water is a gate 9 A More Perfect Form of Water - Wheel. 31 » i y wrevesssrnanpwaswawae sede PLD PPP PET I PPP PLD PI PIS a ee, ee Bor? rot | -__ N Ze az" COLI LILA “a t) \) \ rest Sacks ” 2 4 all! | iD RF oo nn LY RY sS nl a rn ! AS ina al is a \ | ; SEED | See | SRE f Ke tN oN : aa ee : N Nissi} N N : a ZZ tas "ee, Pe, 35 5 > ea. ‘a. ooo > \ \' \\ \ \\ \ ‘ x % 33 g y ree 3 Ka N N N N N 1 : N N : N f . ; ) ontal sh aft N N & y : N Vertical section. Scales! N : N i N \ i : | | : Draft- tube H H Draft tube . : i N N N N N N N \ . N L : N Fig. Pal ae : \ N N : . | N e that shall, in closing, contract not only the influx of the wheel, but the entire passages through it. The attainment of these conditions necessitates an abrupt and radical departure from all existing forms of water-wheel. The figures which follow illustrate my idea of a form of water- It must wheel calculated to obviate all the above described defects. fect Form of Water - Wheel. ev A More Per SSSR AS AAAS anennnnnnnn ny //74 COSSSSUS SSSI SG aioe v2 shaft on horizontal No 2 Wheel Sectional plan- half, looking up, and half. looking down. Fig. £3 A More Perfect Form of Water - Wheel. 33 be observed that these are, in no sense, working drawings. They are intended to exhibit the principles on which such a wheel could be constructed, and are sufficiently detailed to show that the construc- tion of such a wheel involves no mechanical impossibility, and is entirely within the resources of modern engineering. Fig. 25 shows the form of the floats and float passages, Figs. 22 and 23 the form of the gate. The gate g is a conoidal dise with a central hub, which slides on the shaft. A heavy hub, secured to the shaft, carries the flat dise or discs dd, to which the floats f are se- Wheel Na2 .on horizontal shaft Elevation, Fig. a4 cured. ‘The outer part of the gate is provided with openings through which the floats pass. ‘These openings admit of being packed as in- dicated at Fig. 28. Stout rims are attached to the outer ends of the floats in the manner to be described later. ‘The water approaches the wheel from the exterior with a revolving motion. Its rotatory velocity at its entrance to the wheel is equal to the velocity of the exterior circumference. If this was not so, the water would not enter the weeel. It is discharged in the reverse direction with a velocity nearly equal to that due the head. Its entrance to the mov- ing orifices of the wheel implies no greater shock or loss of head than would be involved in its entrance to rounded stationary orifices. It passes the wheel under conditions consistent with maximum efli- ciency. These conditions are in no way changed by the opening or closing of the gate. ‘The energy of the water in this wheel is devel- oped wholly by reaction. It is, as shown at page 10, nearly double 34 A More Perfect Form of Water - Wheel. fe y y ; on y y of fig 22 Fig. 25 Section m of Water - Wheel. A More Perfect For LA oe: is Fig. &6 36 A More Perfect Form of Water - Wheel. Pn ss eT ~ ~ ei y fO X-X UO UO!LI—aS gee =- = - - - — a - but half of this energy | 2 the whirling movement to the ) that due to the head and quantity of water ; is absorbed in the work of impartin water before or during its entrance to the wheel. A More Perfect Form of Water - Wheel. 37 We will now describe the construction of the wheel and its ad- juncts. Figs. 22 to 28 relate to a wheel on a horizontal shaft. The Gate is in two parts: — The first part is a dise of conoidal shape which slides horizontally upon the shaft; the second is a cylin- der with a broad internal flange, as shown at Fig. 28. Both parts are pierced with openings for the passage of the floats. At the junction of the two parts is a thick sheet of fibrous packing, pierced in like manner. ‘This being compressed by the screws which fasten the two parts together packs the floats. The cylindrical part of the gate has the packing ring 7, so that, in a single wheel, the space between the gate and disc, in a double wheel the space between the two gates, becomes a water-tight compartment. It would be a very simple matter, also, to introduce packing at the hub where it slides on the shaft. The Discs. — The wheel under consideration is a double wheel. The central hub which is fastened to the shaft carries two flat circular discs. These have openings, 0, which allow the water to pass freely. The rim of the dise is thickened and notched like a ratchet wheel, as shown at Fig. 27. The Floats. —'These are continuous through both wheels. The part traversed by the gate is of the form shown in section, Fig. 25. The part resting on the dises, and lying between them, has the sec- tion shown by the cross natching in Fig. 27, the blade of the float coming to a close shoulder against the disc, and preventing any ten- dency to endlong displacement. After being put in place and tem- porarily fastened, the wheel is put in a lathe, and the central part of the floats turned off smoothly, leaving the diameter of that part of the wheel a little greater than that of the outer ends of the floats. Then a heavy band, B, isshrunk on. Fig. 28 shows this band extend- ing oyer the whole space covered by the discs, but two narrower bands would do as well. The float may have a thickness of 14 inches near its outer edge, forming a very strong stiff bar. The outer end of the float is turned down to a cylindrical stud, 14 inches diameter, and 14 inches long, and threaded. Some 4 of the floats in a wheel are bored longitudinally with holes, 0, Figures 25, 26 and 28. These holes reach from the outer end to the disc, and are there met by holes cut from the outside of the float. These holes connect the interior of the wheel with the low pressure compartment of the case. The band, B, is thicker at one disc than the other to allow the cylindrical parts of the gates to telescope when open. A groove is turned in 38 A More Perfect Form of Water - Wheel. the band, at each disc, for the packing ring 7. After shrinking on and finishing the bands, the cylindrical part of the gate is put on, the flange straddling the floats. Then the sheet of fibrous packing is slipped on. Then the discoidal part of the gate is adjusted, and the screws turned up which connect the two parts and compress the = N SS LikiiiiiiiiiiiDieliisidiiViidildllldlde < \ : : By , yy ei \ | Wiss SSN FY Se Section on k-kK of fig 27 Fig. 28 packing. Then the rim, which is bored to receive the threaded studs on the outer ends of the floats, is put in place and solidly fastened by countersunk nuts. ; The Shaft has an internal bore-hole communicating with the inte- rior of the wheel. At the end of the shaft this bore-hole communicates with a pipe by a stuffing-box. A More Perfect Form of Wuter - Wheel. 39 The Case. — The wheel being set above the level of the canal of discharge, the case has high pressure and low pressure compartments. The case is shown in Figures 22, 23 and 24. It is a short cylinder of cast iron with flanges for the attachment of the penstock and draft tubes. The interior compartment is in communication, by means of the penstock, with the upper level of the mill site. Being exposed to a bursting pressure, it is formed by two conical diaphragms joining the large cylinder. The junction is marked by two broad ribs run- ning around the latter and widening into flanges for the attachment of the penstock and draft tubes. An annular portion of these dia- phragms, next the wheel, is made detachable for convenience of fin- ishing. Ribs also run around the ends of the exterior cylinder, and expand into flanges for the attachment of the draft tubes. The pres- sure on the ends of the cylinder is inward, and these ends are of dish shape, bending inward. ‘This form not only gives great strength to resist the pressure, but shortens the unsupported part of the shaft. The flattened portion of the penstock, at its junction with the case, can be strengthened by external ribs, if required. ‘The draft tubes being rectangular in section are divided by partitions into several dis- tinct passages, as shown in Fig. 24, for greater strength. As the rim w, of the wheel, Fig. 26, must revolve without touching the case c, an annular space must exist, allowing for wear and imperfection of workmanship, through which the escape of water would be objec- tionable. To alleviate this difficulty, the small ring, 7, is confined to a seat turned on the case so as to be capable of slight lateral movement. This ring can fit the rim of the wheel much closer than the fixed case, while yielding to any slight displacement of the wheel, from wear or other cause. The Bearing, }, of the shaft, is an undivided cylinder. It has a rib or flange around the middle by which it is riveted to a circular plate of wrought iron. This latter is fastened, at its outer circumference, to the end plate of the case, covering a circular opening in the same, some 30 or 36 inches in diameter. The bearing has an oil-cup and a packing gland not shown. This bearing, I presume, will be the subject of some criticism, but I think it a correct design. Bearings are usually made in two halves, for convenience ih setting up machin- ery rather than any inherent advantage in that method. The ring of boiler plate surrounding the bearing gives it a slight but sufficient degree of flexibility, which is favorable to uniform wear. The press- ure being from without inward favors the admission of oil. The 40 A More Perfect Form of Water - Wheel. admission of water to a bearing which is well lubricated is no disad- vantage, as is shown in the Westinghouse engine, where the bearings ~ run constantly in water covered with oil. Regulation. — Fig. 29 is a schematic sketch of the proposed regu- lator, and Fig. 380, a section of the valve for controlling the move- ment of water through the central bore-hole in the shaft. The plug of the valve carries an arm resting on the spindle of the regulator, Lib med LDL if 4 Sa Je Sketch of governor and section of valve to regulate Speed of wheel—— and weighted as indicated, so that it follows the movement of the spindle, under the action of the revolving balls, rising when the speed diminishes, and falling when it increases. An increase of speed opens the passage leading from the upper level to the wheel; a decrease closes it. Now, in the normal running of the wheel, water is escaping from the closed chamber through the small orifices o, in the floats, and entering through the valve. When the valve opens wider, an increased quantity of water enters the chamber, raises the pressure therein and moves the gate to close. When the speed sud- A More Perfect Form of Water - Wheel. 41 denly starts forward, the valve opens wide and closes the gate rapidly. When the valve closes to less than the normal opening, water escapes faster than it enters, the pressure in the chamber falls below that in the low pressure compartment, and the gate opens. The rapidity with which the gate will open when the valve is entirely closed will depend on the size and number of the openings 0, that is, upon the quantity of water constantly wasted. ‘The wheel under consideration has an external diameter of 5 feet and is supposed to draw some 200 eubic feet of water per second, 12,000 per minute, under a head of 20 feet. The cross-section of the closed chambers may be taken at 20 square feet. To move the gate at the rate of 12 inches in a minute would involve a loss of 20 cubic feet per minute, which in comparison with 12,000 is not worth considering. A movement at the rate of 12 inches per minute would suffice for any ordinary use of a water- wheel. A closing movement of any desired rapidity can be attained by suitable proportions of valves and passages without waste of water. To start the wheel from rest, the weighted arm is thrown into the position ‘* to start,” Fig. 29. Then the interior of the wheel is in communication with the lower level, and water from the upper level rushes through the nozzle into the pipe leading to the lower level, forming a jet pump which draws the water or air from the in- terior of the wheel, and opens the gate. ‘To stop, when running, the weighted lever is thrown into the position marked ‘‘ to stop.” Then the valve is free from control of the regulator, and the interior of the wheel is in full communication with the upper level; the gate closes. On a low head it might be doubtful whether the wheel could be started with certainty by this method. In that case it might be ad- visable to temporarily connect the pipe ‘‘to upper level,” with a municipal water-main, or with the fire-tank of the mill, which would, if required, make a complete vacuum in the wheel chamber. In start- ing the wheel after the water has been shut out of the penstock, the draft tubes and wheel chamber would be filled with air. The latter would remain filled with air after the starting of the wheel, and, as it filled with water, the air would collect at the center and obstruct the regulation of the wheel. In this case the gate would be opened far enough to admit a large volume of water and expel the air from the draft tubes, raising the water in the latter above the wheel. ‘Then admit water from the upper level, close the gate and expel the air, which escapes through the highest orifices 0. Then the gate can be opened without admitting air. Of course the air can be drawn out « ct Form of Water - Wheel. fe . A More Per 42 le S14 “UOIIDIG «—"4FFRYS |BIW4fysA UO jaoymM : nee Sy AAS \ 28 » ‘ x Note Rae - \ F ¢ : : ESQ 5 —= = SSS eee 7 eae } Fey, GEL EES STI EE TITIES EET EET eee ee | CILLA etter icdeitinitaraditanacce TZZ LLL 4 [ZZ deed hhedekehchehecechecheahokechchchechechtheck jawwaaw ben an f H |||! " " 0 eat K N N «IN =A Zz XS NdOYSUdY See 7 ‘an | N \\t ee 7 AVN WW Saeeeee’ oo AE ae N ; NN N | ! 4 K NA N i “NN \a N 4 ou oe beet ee -_-NA y 4 ————————— SN Rees =e y | ; N N |QAo, - 4ayERM 4QamMo7 N Fe i N \} ‘; | | | i : | i - | iN 2 \ \ A More Perfect Form of Water - Wheel. 43 ‘ by the jet pump, but this would involve opening the gate to its full width, which might not be desirable. Equations 11 to 15 were deduced with reference to the regulation of the wheel by the action of centrifugal force. We will inquire what amount of force we have at disposal for the movement of the gate. ‘The wheel under discussion is 5 feet in diameter on a head of 20 feet. This wheel takes a higher velocity than one of the common form, in which the head is partly expended in imparting the velocity through the guide passages. The exterior circumference would move Wheel on vertical shaft-— Plan Fig. 32. _ with very near the velocity due the head. When the influx of water to the wheel chamber is shut off, the internal pressure at the cireum- ference is equal to the external pressure at the center. The internal pressure at the center is less than this by the head due the velocity, which we may call the centrifugal head. This head, in the present case, would not be less than 16 feet. The force tending to open the gate is equal to 8 feet depth of water acting on a surface of some 20 square feet — over 10,000 pounds. The force to be overcome is the friction due to the weight and packing — nothing approaching the above. Consider the most unfavorable case that could occur, 12 feet may be taken as the lowest head that a horizontal wheel would be used on, and 36’/’ diameter would be about as small a wheel as would be used on such a head. The centrifugal head would not be less than 10 feet. The pressure tending to open the gate is, roughly, fect Form of Wuter- Wheel. A More Per 44 e¢ Sy “UOILIOS tFEYE [22f4en UO jeoym ajguis 2 ) Pa Fat kh I cn teerotge Fomehy iF, aes pF OFS 5 MAD. PE RL i GOI LEE vee ee Pe Si op Ly te Ss OS Uw BIR ORCS Lie ie Nira’ 0 ae ani ca 3 Nene fe Sigskt Vests # . eG ata em dato a s Pde Dini Niel 7 * r 3 ‘ “2 ¢,% phe ee AUN bs SEM pac Ea We ae | SS CLI SS > A) | SSS 0 | pe 4) a Y Hi! 4 5 | 7 YAO { 9 aay ~ —_———_— = arias canna aaaSSamaanoian ——————————————— = J@AZ] uayem uamory | tas A | i | Z A | —— SEES etd Ss ® A More Perfect Form of Water - Wheel. 45 yx 38x 3 XK 7854 XX 62.5 = 2209 lbs. For a head less than 12 feet we should employ a wheel on a yertical shaft. 6 feet may be regarded as the lowest head worth developing by water-wheels. On such a head there is seldom occasion to use a small wheel. » A 5-ft. wheel on a 6-ft. head would expose to pressure something over 20 square feet for moving the gate. The centrifugal head would be about five feet. The pressure would amount to about 20 K 25 x 62.5= 3125 lbs. Of course we may conceive of a head so low and a wheel so small that this mode of regulation would be inapplicable. To meet such causes I am prepared to say that a regulator may Single wheel on vertical shaft Plan. Fig. 34 be devised capable of creating a total or partial vacuum within the wheel chamber. There is never any question as to the amount of force available for closing the gate. With full communication between the wheel chamber and the upper level, the water enters the former much faster than it can be discharged. The centrifugal head acts in con- cert with the static head to close the gate. In every application of this method, the water issuing from the floats changes its direction by a right angle in leaving the wheel. This change of direction causes a pressure on the gate tending to open it. The water should not leave the wheel with a velocity of more than 6 or 7 feet per second. With 7 feet at full gate, the pressure would amount to a head of some 15 inches, diminishing to 46 AW More Perfect Form of Water - Wheel. nothing as the gate closes. No reliance can be placed on this force as aiding the opening of the gate. ; | The foregoing description contemplates a wheel on a horizontal shaft. Figures 31-2-3-4 show that the proposed construction is just as applicable to a wheel on a vertical shaft; 31-2 show a double wheel; 33-4 a single wheel. Neither of these presents any peculiar difficulty, or requires any special description. For a small quantity of water on a low head, a single wheel would be preferable to a double one of smaller diameter. In such case it would probably be better to let the wheel discharge upward instead of downward as indicated, in order that the weight of the gate might assist in open- ing. Wheels often run under conditions to which the foregoing mode of regulation is inapplicable, as, for instance, in connection with a steam-engine, which controls the speed, and meets all variations in the demand for power. In this case, the wheel, when under no limitation as to quantity of water, runs at full gate, the regulator disconnected, and the valve set to keep the gate open. Sometimes, however, the wheel is under limitations as to the quantity of water. Natural limitations resulting from diminished flow of the stream. Legal limitation of leases and grants. In such case the gate cannot be set at any unalterable opening. It can only be left in control of the valve. It is obvious that the latter must be controlled by condi- tions other than the speed of the wheel. It would be easy to point out modes of regulating the discharge, without reference to the velocity, to conform to the varying flow of a stream, to a uniform draft of water or to a uniform output of power. But such inquiries would carry us far beyond the contemplated limits of this paper. i bl -. i . ‘’ sd > «gal? me t q i . 1 : x i, . ‘ ios i Metin? | . ‘ « ’ 9 ‘ - b i: b a 3 20 73213 . a v " ; Fe = Ser be by ne ‘ Fs i Fah