Bourne Archive: Brees: Water wheels        Latest edit 15 Jun 2010


The Bourne Archive


The Architectural and Engineering Glossary of S.C. Brees (1852)

Extracts Concerning Water Wheels.

pp. 470-7

WATER-WHEEL, an hydraulic machine employed in connexion with millwork, filling the situation of prime mover, and being the instrument whereby the motion of the water in a river of stream is brought into action. The use of the water-wheel was not unknown to the ancients, but its use does not appear to have been revived until late in the middle ages.1

There are four descriptions of water-wheels—viz., 1st, the undershot; 2nd, the overshot; 3rd, the breast wheel, each of which receives the impulse of the water vertically; and, 4th, the horizontal, upon which the water acts horizontally or bodily.2

The undershot water-wheel is the most simple in action, and was in use long before the others, being the cheapest and readiest for small streams when in their natural state. It may be used almost without any fall in the stream, provided there is plenty of water and a good current, as it acts principally by the momentum, and not by the weight of the stream; it also answers equally well either way, which renders it very suitable for tide rivers, and where the difference between the ebb and flood is not very great.3 The wheel should not be immersed in the water much beyond the width of the float-boards4, on account of the loss of power from the action of the water upon them in returning upwards, after having passed through the lower part of the wheel-course. The wheel should be made of sufficient diameter to allow of a small segment only being covered by the water5.

The overshot wheel is usually made in the shape of a drum, upon which a series of buckets are constructed, the water passing over the top of the wheel into them; it therefore acts by the gravity or weight of the water in the buckets, as well as by the momentum of the stream. This mode affords the greatest power with the least quantity of water, as the thickness of the stream is seldom more than half an inch, or an inch. A penstock6 is fixed at the head of the wheel, in a proper trough, to regulate the supply. The overshot wheel requires a fall on the stream equal to rather more than its own diameter, which renders it necessary to make it of grater length in proportion to its height than is usual with other wheels. Its power is calculated at double that of the undershot wheel.

The breast wheel is a medium between the two described, and is consequently much the most general; but, like overshot, it requires a considerable force in the stream, which is consequently rendered unfit for the purposes of navigation. The water usually strikes the wheel just below the axis, but it is sometimes situated above, and either floats of buckets are employed to receive it; the former being mostly adopted. The whole of the water which passes the mill-course operates upon them, there being no space left between by which it can escape7. The supply is regulated by a penstock, the same as in the last description. The breast wheel consumes about double the quantity of water of the overshot wheel in performing the same quantity of work; the diameter of the wheels, number of float-boards, &c., being similar in each case. This method is most suitable when the fall is from 4 to 10 feet8; when it exceeds the latter, it is best to divide it into two falls, and the supply of water must of course be ample in either case.

It is highly essential with every description of water-wheel to get rid of the tail water, as it is discharged at the bottom, since the power of the wheel is considerably reduced by it9. Two small culverts or drains are sometimes employed to effect this, which are made in the masonry, passing from the head of the wheel to the tail water, when the impetus of the stream rushing from the upper pond down these drains carries off the spent water very effectually. A penstock is required to be placed at the top of each of these culverts, in order to cut off escape of water in dry seasons, or when scarce.

In situations where the supply is large and the fall small, an undershot wheel may be used; if, on the contrary, the fall is large and the supply small, then the overshot is the most appropriate10; and in cases where the height of fall and quantity of water is but moderate, the breast wheel should be adopted. An undershot wheel works best when its circumference moves with between one-third and one-half the velocity of the stream. Some of the earlier engineers made it one-third, while the modern authorities, including Navier, Poncelet, and Morin, give it one half, whether the floats are on the line of the radius of the wheel or curved. Overshot wheels are not influenced by the current, as all the buckets are filled in succession11. Smeaton determined 3 feet per second to be the best velocity of fall for the latter, the distance from the spout to the receiving bucket being made two or three inches.

The full power of the stream should always be taken advantage of in the construction of mills; a wide wheel of a small diameter is best where great speed is required; in other cases, a large narrow wheel may be employed. Some consider that the diameter of the wheel should never exceed the height of the fall, while others, including Smeaton, affirm that the higher the wheel in proportion to the whole descent, the greater will be the effect. Much difference of opinion also exists respecting the proper number of floats. Smeaton gives from 24 to 40 for a wheel of 20 feet diameter; and Poncelet, 30 to 36 floats for a wheel of 16 to 18 feet. In an experiment made with two wheels by Mr. George Rennie, one having 40 and the other 48 floats, it was found that “the wheel with 40 floats ground in 31 hours 359 bushels of wheat, and the wheel with 48 floats ground in 32 hours 392 bushels; hence, the wheel with 40 floats ground the same quantity of wheat (by experiment), with 1.43 per cent. less water than the wheel with 48 floats. It was also proved that 58.62 cubic feet of water, falling one foot, ground one bushel of 60 lbs. weight of wheat for the the wheel with 40 floats.” Each of the wheels was 19 feet 4 inches diameter, and 6 feet in width, and the height of the fall was 12 feet 2 inches. The machinery consisted of two wheels and two pinions of cast iron, and two pairs of 4-feet French burr stones.

Little or no benefit is derived by inclining the floats, but M. Poncelet’s curved floats proved:—“1st, That the maximum velocity of the wheel with curved floats was 0.55 of the velocity of the stream; 2nd, That the dynamic effect for small falls and large openings, and generally speaking, the effect of the wheel with the curved float compared with the effect of the wheel with vertical floats was 0.60 to 0.50 of the power expended.” 12

We are indebted to the Franklin Institution13 for some valuable experiments upon the subject of water-wheels. They were made with wheels of 20, 16, 10, and 6 feet diameter respectively, all the various retarding forces, as friction, &c., being attended to. The results of these experiments prove that the maximum and mean efforts of large wheels are greatest with moderate velocities; the latter, however, is double what Smeaton assigned as the maximum velocity for the mean circumference of a water-wheel, and that these effects are both diminished with an increased velocity, as the wheels are diminished in diameter.










20 feet diameter wheel:

128 Experiments.





15 feet diameter wheel:

88 Experiments.





10 feet diameter wheel:

180 Experiments.





6 feet diameter wheel:

178 Experiments.





Water-wheels were usually constructed of wood in the time of Smeaton, but the employment of iron has led to sundry improvements, principally connected with the buckets. The arms are now made of wrought iron, with cast centre boss, and the buckets and shrouding of plate iron.

The breast-wheel is preferred to the overshot at the present time, on account of the facilities it presents for receiving the water, as well as emptying it afterwards: the height of the wheel being kept so far above the fall, neither backwater nor floods make much difference to it. It is generally found advantageous in every fall, however low, for the water to act by gravitation and not by impulse only; therefore every vertical water-wheel14 may be called a breast-wheel; and the grand desiderata is to admit the water rapidly, to retain it in the buckets as long as possible, and to allow the air displaced by it to escape readily. When this is not accomplished, the air becomes condensed15, and resists the entrance of the water, by which the power of the wheel is considerably reduced. This was remedied formerly, by boring holes in the buckets and sole boards supporting them, but it is now accomplished by having ventilating buckets (see cut). The air escapes through the apertures at the back, and passes upwards, while it makes its exit in the direction shown by the arrows a, b. This improvement increased the power nearly on-fourth, and presented no backwater whatever. Mr. Fairbairn’s ventilating water-wheel is somewhat similar in plan, and is a most effective remedy. It prevents the condensation of the air, and admits of its escape during the filling of the bucket with water, and readmits the air as the water is discharged at the lower level. The uniformity of the motion is not impeded, if the wheel is immersed in 5 or 6 feet of backwater in a fall of about 18 or 20 feet, the air escaping from bucket to bucket, until it rises above the surface of the water. The motion of this wheel is transmitted by means of toothed wheels fixed to the interior of the shrouding.


M. Poncelet’s water-wheel is of a most novel description, and has been much employed on the continent. The wheel represented in the following cut was erected by Mr. De Bergue, and is on the Poncelet principle. It is 16 feet 8 inches diameter, and 30 feet wide, with a 6 feet 6 inches fall.  It passed 120,000 cubic feet of water per minute, when the velocity at the periphery amounted to 11 or 12 feet per second. An ordinary breast wheel would require to be made about three times as wide to pass a similar quantity of water. It was ascertained that the velocity of the periphery required to be about 55 per cent. of that of water flowing through the sluice; and taking this as data, would make this wheel about 180 horse power. The buckets are curved, and quite open at the back, and formed of wrought iron, one-eighth of an inch thick, and are much more numerous that usual. The water strikes the buckets at a tangent with great rapidity. The main shaft consists of a cast iron cylinder 4 feet 6 inches diameter, bolted together in short lengths. The arms are formed of light wrought iron, and placed very close together, like those of a paddle-wheel. The strain is borne entirely by the main shaft, and the weight of the wheel amounts to no more than 30 tons. The sluice is formed of cast iron, in one large shuttle, the entire breadth of the wheel, and is adjusted by racks and pinions with the greatest of facility, and connected with the stone apron by hinged tie-rods.

The force of the water is expended at once on the wheel without the interposition of any wheel-race, and therefore acts with its full initial velocity. The back of the buckets is perfectly open, so that the water passes out directly, without encountering the slightest retardation from the air. The wheel is also less affected by tail water than any other, and therefore very suitable for all falls of less height than 8 feet.

The horizontal water-wheel is rarely met with, being very inferior to others, on account of its resistance which the float-boards offer in returning against the stream, and other defects. The floats are usually inclined or curved, but the results seldom exceed one-fourth of the power expended. Mr Beatson suggested the employment of suspended float-boards, which should present a surface for the stream to act upon in passing down, and allow the water to pass through them in returning upwards against it. He also applied the same principle to windmills. Barker’s mill and the turbine may also be instanced as specimens of the horizontal mill.

The propelling wheels of steam boats are termed paddles, and are employed in a totally different manner from the wheels before described. The paddle-wheel acts upon the water, using it as a resisting force, whereas the former are acted upon by the water, i. e., by the action of the stream.


In Bourne, at the time of the Domesday Survey (1086), there were four, probably five water mills. It is not possible to be sure owing to the various way in which the report is expressed. With the re-planning of Bourne in around 1140, the number of mill sites was reduced to two, with a third probably, at Dyke. The two in Bourne will have been respectively, those of the Castle and Abbey estates. The former site is that of Baldock’s Mill and the latter, of Notley’s mill. Much later, by the mid 18th century, a small bark-crushing mill had been inserted into the system, in West Street. Other mills were powered by wind, steam and oil engines.

Though the buildings are much later, the Baldock’s Mill site is that of the castle mill and the mill pool is part of the castle’s moat system. Its life as a water mill ended about 90 years ago when the breast-shot wheel began to break up but that has been replaced in recent years by a Poncelet wheel, though one with fewer buckets than is indicated in the drawing above.

1. ^    According to the Domesday Book, mills were fairly common and widespread in England at the time of the Norman Conquest: 1066. Since windmills had not then been introduced to England, it is generally assumed that they were water-powered. While it is conceivable that they might have been animal-powered, there are sites from shortly after which are clearly arranged to use water. In Bourne, both the principal 20th century water mill sites show every sign of having been contrived in the first half of the 12th century.

2. ^   To clear up any ambiguity, it is perhaps best to state that Brees’s horizontal wheel turns on a vertical axis, while his vertical wheels turn on horizontal axes.

3. ^   In the later 16th century, Peter Morice set such a wheel up under the northern arch of London Bridge in order to pump water to the citizens. Later, the southern arch was similarly used for grinding corn. In the 17th century, the water supply was increased by such wheels mounted on pontoons moored near the old bridge. They therefore rose and fell with the tide so that their wheels could dip to a constant depth in the passing waters of the ebb and flood. The energy in the passing water was by no means all employed but there was more than was needed.

4. ^     OED Float, n. 11. One of the boards of an undershot water-wheel or of a paddle-wheel; a float-board. The quoted uses of the term date from 1611, 1731, 1806 and 1856. That is, the depth of the water in the stream entering the buckets.

5. ^   In other words, the wheel would suffer from backwatering. In agitating the water in the tail race, the wheel would give up some of the energy that had just been put into it by the passage of the water.

6. ^   PENSTOCK, a sluice or floodgate employed to retain the water of a mill pond, water trough of a water wheel, &c., and to let it off when required (Breese p. 301). It may be pictured as a valve in the form of a sliding door.

7. ^   Shaped buckets may be employed but more usually flat boards (floats) are mounted radially. It is important to minimize the loss of the water‘s power by minimizing the clearance between the floats and the adjacent parts of the pit in which the wheel turns.

8. ^   That is the working head of the water input — the vertical distance through which the water falls in working the wheel.

9. ^   Water full of potential energy, enters the system at its head and having done its work, is exhausted at the tail. The extent to which the wheel causes turbulence in the exhausted water is an indication of the extent to which energy supplied by the falling water is then wasted in freeing the wheel from the tail water.

10. ^ An overshot wheel is most obviously appropriate in a mountainous region where the water supply is a relatively small stream. Water can be taken from upstream and fed artificially at a gentler gradient than that of the natural valley to feed a wheel of large diameter. In the case of the Laxey Wheel, the supply is piped in an inverted siphon.

11. ^  Their motive power relies almost entirely on the weight of the water in the buckets acting downwards on one side of the wheel. The momentum of the water entering the buckets is insignificant.

12. ^ The curved floats reduce the amount of energy wasted in turbulence. The principal feature of Poncelet’s system is the care with which the water is fed to the wheel both cleanly and at high speed. It makes use of the available head of water to produce a jet of water and the momentum of this is used to drive the wheel. In the drawing, note the step in the bed of the race below the wheel. This allows the spent water to escape readily so that it is not carried up by the wheel nor does the wheel have to push it clear. The wheel is not backwatered.

13. ^  Possibly the Franklin Institute.

14. ^ That is: working on a horizontal axis.

15. ^  Compressed.


Archive Contents