Time Water Wheels of America,
By Joseph P. Frizell, 1893
In no way is the progress in any branch of art or industry more strikingly
illustrated than by a glance at the methods in use a century or more ago.
The annals of the American Society of Civil Engineers are supposed to be
a repository of American practice and American methods in engineering. The
writer, therefore, deems it no un worthy task to collect for preservation
therein such examples as he can of the motors and methods employed by our
ancestors of the preceding century in availing themselves of the power of
Had the task been attempted earlier, it would have been done better. The
old millwright was busy, practical man. He had little leisure, inclination
or ability for putting his knowledge on record in an intelligible manner.
He made little use of books in his vocation. He learned his trade as an
apprentice. He imparted his knowledge to others apprentices by example and
word of mouth. Of the crude results of his labors very few examples now
The sources of information available to the writer are: (1) Such isolated
examples as have come within his knowledge and recollection. (2) The drawings
and sketches extant in old books, which are usually mere hints and indications,
requiring to be carefully studied and pieced out in order to yield any intelligent
notion of methods of construction. Of these the book of Oliver Evans exceeds
all others in value. (3) Suits at law for the interpretation of old grants
of water, in which he has been employed as expert, sometimes involving an
injury into the methods and appliances in use at the date of the grant.,
From the testimony and traditional knowledge of old millwrights, and such
meager records as are available, some knowledge can be obtained.
The writer is quite well aware that this is a very hasty and incomplete
summary of the subject; in fact, little more than a beginning. It is expected,
however, that every man who takes it upon himself to comment on the subject
will contribute something of his own knowledge, and that, in this way, this
paper will result in putting on record a much larger body of information
than it contains.
Up to the beginning of the present century the chief and almost the only
application of water power was for grinding grain and sawing lumber. Occasionally
a wheel was set up, to create a blast for a foundry, or for fulling cloth,
or the carding wool, the spinning and weaving being carried on in the household.
A wooden water wheel consists in general of five principle parts, the
shaft, the arms, the shrouding, the soling, and the floats. It is only in
the latter element and in the mode of letting on the water that any distinction
exists between the overshot, the undershot and the breast wheel. The shaft
was an oak log 18 to 30 inches in diameter, dressed to a square, circular
or polygonal form, with iron bands and gudgeons. There were two methods
in use of attaching the arms to the shaft. The method of clasp arms was
not so common in this country as that although much used in Europe. The
shaft was squared at the place of application of the arms, or, if octagonal,
was made square by the insertion of corner blocks. The arms were halved
or locked together and set firmly upon the shaft by means of wedges. The
admitted of some adjustment of the wheel with reference to the shaft. This
was a very strong and durable arrangement for wheels up to 14 feet in diameter.
Beyond that size the four pieces forming the arms would not give sufficient
support to the shrouding, and the additional pieces were inserted, with
the blocks on which they were footed. These were notched to the timbers
and fastened with pins or keys.
The second method of inserting the arms is compass arms. Though the shaft
is here represented as round, it was more commonly dressed to a polygonal
form - six or eight sides, rounded at the ends to receive the bands. The
arms were inserted in mortises passing through the shaft. They were notched
and locked together in the center of the shaft. This method required the
length of the mortises to exceed the width of the arms by an amount sufficient
to admit of their insertion. The widest vacancy thus made was closed by
a heavy oak key solidly confining to the arms. A wheel might have six arms,
each of the three pieces forming the arms being cut away to the extent of
two thirds of its width. The same construction is used for a wheel of eight
arms, each of the four pieces being three fourths cut away. When the number
of arms exceeded eight, the additional arms were inserted into the arm next
to it diagonally in the middle.
All wheels run upon iron gudgeons inserted into the ends of the shafts.
These are cast iron gudgeons. The part inserted in the wood is a cross of
four feathers coinciding with a shaft in diameter. It is confined by heavy
bands, driven on hot. A wrought iron gudgeon it is inserted in a mortise
cut from the outside of the shaft to a depth sufficient to bring the gudgeon
to its proper position. After driving on the bands, the mortise is closed
with keys and wedges driven with great force. The wood is further compressed
by driving thin wedges into the end of the shaft.
The shrouding consists of segmental pieces of 3 or 4 inch plank, forming
a ring like the fellows of a wheel. They are attached to the arms in different
ways. Halved and pinned or forked which creates a wheel with radial floats.
With shrouding of sufficient thickness the arms are attached with mortise
and tenon, which simplifies the insertion of the floats. The shrouding pieces
are halved and pinned to one another at their ends, and where the soling
is wanting the joints are strengthened with splicing pieces.
The floats were sometimes inserted in grooves cut in the shrouding, sometimes
merely "sprigged" to the shrouding and confined at their ends
by blocks of one inch board, fitted in an nailed to the shrouding. In the
earlier forms of breast and overshot wheels, they were inserted as flat
floats, rather than elbow buckets. They were different in the later forms,
as will appear further on. The space contained between two consecutive floats
was called a bucket.
In the high breast wheel and the manner of letting on the water, the part
loaded with water is surrounded, at a distance of about 3/4 of an inch,
by a tight drum of planking called the apron. This prevents the water from
spilling freely out of the buckets as they descend. To transform this into
an overshot wheel it would only be necessary to remove the apron, reverse
the direction of the floats and carry the water over the summit. The overshot
wheel did not usually have an apron, as that forces the water to leave the
wheel in the same direction as it approaches. both the overshot and breast
wheel ran at a speed much in excess of what would be considered economical
in modern practice.
The undershot wheel in common use for grist mills, the floats being radial,
the arms are forked on the shrouding. The floats are set in dovetail grooves
cut in the shrouding and confined with keys, the expectation being that
they would yield instead of breaking when foreign matter was caught between
the float and the sill of the race. The whole structure is evidently made
light and elastic with the view. The floats moved with a velocity about
two thirds that due the head of water, instead of half, which is all that
modern views of hydraulics would allow.
These wheels never required more than two sets of arms. They were never
geared to more than two run of stones, and the width required was from 2
to 6 feet.
The counter gearing was the style of gearing in use for two run of millstones
is the "big face wheel" which gears with the two "wallowers."
On the same shaft with the wallower is the "little face wheel."
This gears with the "trundle," which is attached to the spindle
of the millstones. The velocity of millstones was about 100 turns per minute
for a 5 foot stone, more for a smaller stone, less for a larger, a rough
rule being to give the circumference of the stone a velocity of 1,500 feet
per minute. The gudgeon of the wallower shaft, next the main shaft, rested
on a sliding block, which enabled the wallower to be drawn out of engagement
with the big face gear, leaving one stone idle. It will surprise some readers
to learn that gudgeons of horizontal shafts ran on stone bearings. Oliver
Evans gives directions for the selection of these stones: "hard and
free from grit." He also enjoins great care, to prevent the heating
of gudgeons, as it is liable to crack the stone on which they run.
This was the traditional gearing for a two run mill. Mills, by this and
closely analogous methods, were fitted up and operated for generations and
centuries; and no man entertained a doubt that they embodied the perfection
Another wheel much used for grist mills, viz, the tub wheel, so called
because it ran within a circular enclosure of thick planking put together
in the form of a tub, without bottom. It runs upon a vertical shaft, and,
with a head sufficient to give the necessary velocity, drives the stone
without the intervention of gearing. More commonly a spur gear and trundle
were used as drive the millstones. The wheel is 7 feet external diameter,
with a head of 10 to 12 feet. The floats are radial and are fastened to
starts inserted in the shrouding pieces by means of dovetail tenons and
keys. The water was let on through a spout leading from the flume to the
A wheel of this construction was running at Lowell machine shop some 30
years ago, and isolated specimens are still extant in old mills. Sometimes
the circular rim was made lighter and attached to the floats forming a part
of the wheel and revolving with it. Often the floats were set in an inclined
position, more nearly perpendicular to the direction of the water issuing
from the spout. Ordinarily a small wooden pipe was inserted in the top of
the spout close to the flume, reaching to some height. The precaution was
necessary, to prevent the collapsing of the spout on the sudden closing
of the gate from the partial vacuum created by the momentum of water.
A flutter wheel was simply a tub wheel without the rim, or perhaps more
properly, a tub wheel without the tub. Two examples of this form of wheel
was used in saw mills, which was their principle application. The wheel
used for giving motion to the saw. This had to be small diameter in order
to give, under any ordinary head, the required speed of about 120 strokes
per minute to the saw. The body of the wheel is represented as a solid piece
of 27 inches in diameter. At one end a gudgeon in inserted, at the other
end a crank for giving a reciprocating motion to the saw frame, by means
of a long wooden rod called the pit man. The floats are secured in forked
starts set in the body of the wheel by dovetail tenons and keys. The weight
of the water was considered advantageous as giving a more equable motion
to the saw. The water was let on from an opening at the bottom of the flume.
Another arrangement of sluice for applying the water, allowing it to fall
down an inclined chute through a gate opening formed to give it the proper
initial direction. An old method of lifting and dropping a gate, the agate
stem passed through a guide. A chain for lifting the gate was attached at
a point near the top. The chains pass in opposite directions around a drum
and are fastened at the back. the operation of the gate by means of the
lever is obvious.
The flutter wheel used in a saw mill for running back the carriage. It is
4 1/2 feet in diameter to the extremity of the floats. Its construction
is the same as that of the tub wheel already described, except that the
shaft is squared where the arms on, and the latter are applied.
The mechanism of the primitive saw mill now as much an antiquity as the
old domestic spinning wheel. In the water wheel powered saw mill using a
flutter wheel, the saw stretched in its frame, ran in grooves in the fender
posts. These were attached by hook tenons to the beams of the mill and were
adjustable laterally by wedges. The carriage which supports the log it runs
upon the ways consisting of narrow plank set edgewise in notches by a trundle
on the same shaft with the rag wheel. This wheel has an iron ratchet ring
in its periphery, and is provided with teeth in the manner of a face wheel.
The vertical shaft of the flutter wheel carries at its upper end a lantern
which remains in gear with the face wheel, the flutter wheel conforming
to the movements of the carriage. A series of teeth is inserted in the bottom
of one of the side timbers of the carriage frame. They are alternately on
opposite sides of the way on which it run and fit the latter closely. The
way is interrupted at the point where the trundle engages with these teeth.
A lever worked by the top of the saw frame imparts a rocking movement to
a shaft with a projecting arm to which the hand pole is jointed. This pole
carries an iron "hand" at its opposite end resting on the rag
wheel. This wheel is provided with a pawl which prevents it from going backward,
advances the carriage slightly at each stroke of the saw. The hand pole
is attached to the arm by a pin, and by means of a series of holes in the
arm the feed can be varied according to necessarily. One end of the log
rests upon the head block, fixed upon the carriage, the other on the tail
block which is adjustable to suit the length of the log. "Dogs,"
attached to the blocks, are driven into the log to hold it in position.
When the tail block closely approaches the saw, a projection on the carriage
strikes a trigger which lifts the hand pole and pawl off the rag wheel,
turns a light stream of water on the flutter wheel and the carriage runs
gently back till the saw, which does not stop, enters the recess in the
head block. The attendant then releases the dogs, adjusts the log with his
mill bar to a new position, drives the dogs, drops the hand pole, and the
work goes on.
There are several forms which the breast wheel assumed early in the present
century when textile manufactures began to extend. It was at a later date
that iron shafts were introduced. In this form many specimens still survive,
though no new ones are constructed. It was not far from 1845, that the breast
wheel in New England began to be replaced by turbines, and as the substitution
is not yet complete, it is apparent that the wheel must possess merit in
order to hold its ground so long.
In this form the wheel has a width, or as we should say, length, greatly
exceeding the old grist mill wheels, the one wheel being something over
20 feet long. It has four sets of arms and shrouding. The shrouding at one
end is made much heavier than the others, and on this is bolted a series
of segments forming a toothed ring through which the power is transmitted
to a pinion called a jack gear. The arms radiate from cast iron hubs or
discs fixed upon the shaft. The larger ends of the arms enter recesses in
the hub and are strongly secured by hard wood wedges, then covered by a
plate united to the disc by bolts and nuts. The soling is secured to the
shrouding by wood screws or lag screws. The buckets differ from the older
form of wheel in being made on two parts - the start, which is radial to
the wheel, and the float, which is usually in a direction nearly tangent
to the soling. This bucket does not empty so soon as the older form. In
the construction the floats and starts can be inserted in grooves cut in
the shrouding, the soling being applied afterwards. In that the soling is
put in first, forming a continuous drum or barrel before inserting the buckets.
The admission of water is controlled by horizontal sliding gates, the wheel
having three sets of three gates each. Each gate is attached by two rods
to arms on the rocking shaft, which is controlled by a regular in the manner
indicated. In wheel of more recent construction, the method of admitting
the water is employed, a heavy web of rubber cloth or leather is wound upon
the iron cylinder and uncovering the orifices according the the requirements
of the work. This method, however, does not appear to offer any advantages
over the old arrangement of sliding gates. On the contrary, the latter appears
to offer more ready adjustment of the velocity to sudden variations of power.
Ancient Gearing. - Large face wheels were made of two thicknesses
of 4 inch plank, each wheel requiring 12 pieces called "cants."
The pieces were all circled to the proper radius. Six of them were scribed,
halved and pinned together into a continuous ring; the remaining six were
butted together, forming another ring; which was confined to the former
by a great number of wooden trenails. A face, some 6 inches wide and projecting
1/2 inch. was formed on this later ring, in which teeth were inserted. A
set of arms was inserted in the shaft as already described for water wheels.
To these the wheel was adjusted by suitable notches and shoulder and secured
The lantern was formed of two discs united by the rounds which served as
teeth. Each disc was made of two thickness of plank or boards pinned together
and strongly banded. These wheels were attached to their shafts or spindles
The Construction of the Spur Gear. - This is gathered from descriptions
of old gearing. A series of thick wooden staves or segments is bound together
by iron hoops at the ends, leaving the central part free for the mortises
in which the teeth are inserted. The segments are cut with the grain parallel
to the axis of the wheel The wheel thus formed is mounted on the arms in
the manner indicated. A mode of constructing large spur wheels more common
in Europe was the following; two sets of arms were mounted upon a square
shaft in the manner of a clasp arm. To these were applied shrouding pieces,
as in the case of the water wheel, forming two rims at a suitable distance
apart. Between these rims was inserted a series of segmental blocks, the
grain radial, a tooth being formed on the outer end of each block. The whole
was firmly bound together with bolts and straps.
Old time water powered grist mill, though not comprehended in the title
of this paper, will be of interest in this connection. It is borrowed from
Fairbairn's "Useful Information for Engineers," and represents
a corn mill erected in England in 1730. It is said by Mr. Fairbairn to correctly
represent the state of art with reference to mill gearing at that time.
It shows that our American millwrights were not much behind their fellow
craftsmen of England at that early date.
F. COLLINGWOOD, M. Am. Soc. C.E.- I would like to ask Mr. Frizell whether
he has paid any attention to cast iron wheels? As a lad, I remember there
was a foundry near where I lived where a great many cast iron wheels were
made. That was in 1845.
J. P. FRIZELL, M. Am. Soc. C.E.- Do you refer to breast wheels or overshot
wheels of cast iron?
Mr. COLLINGWOOD.- They were small wheels, not over 4 feet in diameter, with
curved buckets cast in iron.
Mr. FRIZELL.- Many turbines have been and still are made almost wholly of
cast iron. Some very efficient wheels are made in this manner. A preferable
method, in my opinion, is, to make the floats and guides of wrought iron
or steel. These are set up in the molds before casting, and on pouring in
the molten metal they become solidly united with the cast iron parts.
Mr. COLLINGWOOD.- I asked the question because it seemed to me that was
probably the beginning of the introduction of the turbine, which has finally
displaced the wheels that are described in this paper.
Mr. FRIZELL. - Water wheels wholly of cast iron are not now, and never have
been, to my knowledge, made. I use the term "water wheel," now,
in distinction from "turbine," to indicate a wheel running on
a horizontal shaft, with a diameter nearly or quite equal to the fall, sometimes
greatly exceeding the latter, the water acting on the wheel mainly by gravity.
Such wheels continued to be made after the substation of iron for wood in
construction. The shaft, and what we might call the "hubs," that
is, the members which unite the radial arms with the shaft, called rosettes
or centers, were of cast iron; the remaining parts of wrought iron or wood.
Wheels of this construction still continues to be made in Europe. As late
as 1886, as elaborate German work, by Bach, was issued, devoted mainly to
One of the largest and most elaborate wheels of this class, made wholly
of iron, was erected at Greenock, in Scotland, some time previous to 1850.
It was some 70 feet in height. Its construction was similar to that of the
Columbia bicycle, the weight of the water acting directly on the gears and
producing no torsional strain on the shaft. For aught I know, it may be
still in operation.
The extension of the iron industry, and the substitution of iron for timber,
was not, as had been suggested, the sole cause of the replacing of these
wheel by turbines. Other reasons contribute largely to this result. These
wheels were bulky and occupied valuable space. In our rigorous northern
climate they had to be housed in to protect them from frost. I imagine that
a turbine 5 feet in diameter would furnish as much power as the enormous
breast wheel that I have just referred to. There was still another reason
of great force. These wheels revolved with a slow speed. The Greenock wheel
referred to would not probably make more than two revolutions per minute,
and an ordinary 20 foot breast wheel not more than seven or eight. To bring
this speed up to the requirements of modern industry involved changes and
transformations consuming much power. So that although the old overshot
and breast wheel might, and no doubt often did realize quite as large a
percentage of the power of the water as the modern turbine, the losses incident
to the intermediate gearing made the net result materially less. When we
reflect that, as compared with the great wheel at Greenock, a modern turbine
of 5 feet diameter, would not only furnish the same power, but would run
with a speed of 200 or 300 revolutions a minute, we perceive the enormous
practical disadvantage under which the great wheel works.
For operations requiring a very slow movement, there is no question that
water wheels may be made to use water with an efficiency quite equal to
that of the turbine, or even greater. The Sagebein wheel, invented in Europe,
some 20 years ago, yielded by actual test, and under the hands of an experimenter
no less distinguished than M. Tresca, an efficiency of 93%, a result never
equaled or closely approached by any turbine. This wheel had a diameter
many times the fall, and revolved with a very low velocity, not more than
4 feet per second at the circumference - this wholly unsuited to most modern
There is one use to which a wheel of this kind may be applied in which it
will probably never be superseded by the turbine, viz, the raising of water
for irrigation or other purposes. A wheel of large diameter working on a
very low head, too low to be available for a turbine, is provided with a
series of small vessels, which fill when at the lowest point and at the
summit of the wheel, discharge into a spout leading to the irrigation ditches.
Many such wheels are found in great numbers in India, China, Southern Italy
and Spain. Being used exclusively in hot countries, they require no protection
from the frost. It is said that more water is raised by this means than
by any other device known to man.
T. C. CLARKE, M. Am. Soc. C.E. - I remember seeing the old fashioned wheels.
It may be of interest to call the attention of the Society to the reason,
which I do not see given in this paper, why all these wheels have gone out
of date and have been superseded by the turbine. It is because the age of
iron had suppressed the age of wood. In the old days they made what they
could in wood; the facilities of the country in working in iron were not
sufficient to make the modern turbine, even if it had been invented. After
the invention of modern machinery the wooden wheels were superseded by iron,
just as you have seen the iron plows take the place of wooden plows on the
Mr. FRIZELL. - I think the reason why these great wheels have been superseded
is not altogether on account of materials, etc.
Mr. CLARKE. - That confirms the view which I suggested. It would be impossible
to make a turbine or Pelton wheel out of wood; you have got to learn how
to work the iron before you can make these modern wheels at all.
JOSEPH T. DODGE, M. Am. Soc. C. E. - This being the first time I have happened
to attend a meeting here I feel a certain embarrassment about making any
remarks, but I remember the use of a wheel with cast iron curved buckets
placed inside of a rim, which was used specially to run the carriage back
where they were sawing lumber. The spout conveyed water down at an angle
of about 45 degrees, striking the curved buckets at a right angle as nearly
as possible. Those curved buckets were supported by an outer and an inner
rim. The overshot and undershot wheels were also in use, as has been described.
That, of course, refers to the time between 1830 and 1840. I recall how
the wheel was propelled in one saw mill. The saw was an up and down one;
the wheel consisted of two rims with straight floats between; the water
was delivered at an angle of 45 degrees, striking the flat faces and working
the crank which operated the saw. That is the way the country saw mills
were operated. The overshot was used too, in some cases.
O.F. NICHOLS, M. Am. Soc. C.E. - The Burden overshot wheel at Troy, which
has, I think, been mentioned, should rank among the old American wheels.
IT was constructed of metal, and its diameter was very great about 60 feet.
The conditions under which it was used are still maintained, and the wheel
is now in use. These conditions were very peculiar. There was a comparatively
slight flow of water reaching the site of the wheel through the tail race,
and falling 60 feet over the wheel. This was an uncertain flow, reliable
when you could have it, but you could not always have it, and at very low
stages of the river every drop of the water was allowed to pass over the
wheel. The wheel was 22 feet wide on the face. It had 36 buckets, each 6
feet deep, and, of course, every time it turned around there was a great
deal of work done with a relatively small amount of water.
Mr. Frizell has spoken about the necessity of housing in; as to the space
occupied by this Troy wheel, the wheel was placed in a niche or cranny of
the rock that could hardly be used for any other purpose. They simply housed
in a corner in the rock in which this wheel was placed. I doubt if the space
could have been as well utilized in any other way. Working under a continuous
supply of water, efficiency of this wheel would be very great indeed, perhaps
surpassing that of some of the modern turbines. The burden Iron Company
advise me that the wheel was built in 1851, is still in operation and is
rated at 200 horse power.
The turbine wheels of modern times are among the most efficient of prime
movers, and skill and ingenuity have so developed and improved these machines
that they have been introduced to the exclusion almost of all the other
forms which utilize peculiar surroundings, as at Troy. The modern wheel
is compact and convenient; you can use several smaller wheels in the same
space occupied by one large one.
I recall an instance in ordinary railway experience where it was desirable
to furnish the power for drills, etc., in tunnel boring, and where it would
have been very difficult to obtain or set up steam apparatus or indeed any
machinery, and very expensive to transport it to the tunnel sites. In one
or two instances we arranged to use a large overshot wheel. The reason for
selecting this character of wheel was that, having a pretty good grade in
the river, a mountain torrent, we could use a closed tube under very light
pressure, carried at very light grade, about one half of one per cent. This
was enough to carry the water along the bank of the river, and then we simply
extended this pipe or tube until we had reached the level of the top of
the wheel. The wheel was made in sections and could be moved from one place
to another and had no expensive foundations. A turbine wheel in such cases
would be out of the question. I wish Mr. Frizell had given us some instances
and illustrations of the more celebrated of the older wheels. I know that
the wheels he speaks of in Europe have been illustrated, and I wish the
Troy wheel might have been pictures, and its efficiency, present condition
and relative usefulness made know; it was certainly one of the older of
the powerful American wheels, and the forerunner of the more compact but
not more efficient turbine.
Mr. LEVALLE. - I hope I do not impose on the kindness of the members. In
France, especially on the Seine, where the current is very rapid, they have
large wheels made like those of a steamboat; these are in the front. The
paddles are very long and instead of moving the boat, the boat is anchored
and the wheel is let loose and it is moved simply by the velocity of the
current, and this is utilized for small industries, for turning a lathe,
or for grinding coffee, or probably chicory or whatever goes into the coffee.
I do not know whether these have come to the knowledge of Mr. Frizell. I
think it would be interesting to know. I think they would be of value to
small industries. I do not suppose they could be applied in this country
to large industries.
Mr. COLLINGWOOD. - In reference to those wheels, I saw a great many of them
in going down the Danube; these were quantities of those wheels. I think
they were used for all milling purposes, I think for grinding corn.
A. McC. PARKER, M. Am. Soc. C.E. - I have seen water wheel much of that
pattern used in Colorado for irrigating purposed. A wheel was put in a stream
so it could turn, and as it turned it had boxes on the outside of it, each
of which took up a small amount of water which was thrown into a flume and
used for irrigating purposes.
L. L TURNER, Assoc. M. Am. Soc. C.E. - I remember an instance of what might
be called an automatic water wheel running a small saw mill which I came
across in a small stream high up in a Swiss mountain district. I was curious,
also wanting to rest, so went into the saw mill, but could find no one operating
it. There was, however, a vertical saw industriously cutting through a log.
I waited a little while to see what would happen. Attached to the end of
the log was a small stick nailed at right angles to it; when the saw reached
this end, the stick of wood struck a lever, releasing a gate in the flume;
the gate fell and the mill stopped. When the proprietor returned from his
morning's farming he would turn over the log, lift the gate and the mill
would start in again and work until another board was sawed off. Two boards
a day or perhaps three.
*Additional discussions on this paper received before July 1st, 1893, will
be published in a subsequent number.
Return to HomePage