Itchen Navigation: The Locks and Barges


This page considers the design and appearance of the locks on the Itchen Navigation when the waterway was in use and the boats or barges that used them. The ravages of about 150 years of disuse make it difficult to appreciate the locks today and there are no known examples of the barges. Few images exist of the waterway from the days of commercial use and so some features cannot be positively detailed. In addition, the names of some of the features of a lock and of boats can vary from waterway to waterway. In this description, the terms commonly used today on the English canal system are used.

The Locks

The Barges

We must remember that in the 17th Century what we would today call “civil engineering” was in its infancy - indeed the term was not used until the second half of the 18th Century. Waterway construction tended to be a case of trial and error: if a structure you had built failed, then you would rebuild it more substantially.

Construction of the Navigation seems to have started soon after the passing of the 1665 Act of Parliament although it appears that disputes with landowners and millers led to the passing of about 45 years before regular trade between Northam and Winchester commenced. Thus the structures of the Navigation were built over a considerable period although basic designs are unlikely to have changed significantly over this period.

On rivers, dams and weirs were often constructed by millers to provide a head of water to drive their millwheels and weirs across rivers were also constructed to trap fish in nets or baskets. In effect, dams and weirs divide a waterway into a series of steps to climb or descend a hill.

A lock is a device, usually near a dam or weir, allowing water in a river or canal, with any craft on it, to be held back or released under more-or-less controlled conditions, allowing boats to travel up or down hill. There are two main types of locks:

Pound locks

A pound lock is a reasonably simple structure to allow boats to proceed up or downhill. In essence, it is a box that can be filled with water or emptied at will using gravity. At each end are a pair of gates that can be opened and closed (but only when the water level is the same on either side of a pair of gates) to allow boats to enter and leave.

Simple diagram of operating a lock

Simple diagram of operating a lock to pass a boat going uphill.
Image: © TES Global Ltd.

Flash Locks

There were two flash locks on the Itchen Navigation: Brambridge Single Gates and Shawford Single Gates. It appears that the primary function of these locks was not intended to enable barges to be raised or lowered but rather to maintain the water level upstream for the benefit of the mills at Brambridge (demolished between 1810 and 1869) and Shawford. It would seem that these half locks had a single pair of lock gates and the rise/fall at these structures was normally quite small; maybe only a foot (0.3m) or so. Almost certainly, there would have been one or two capstans or winches upstream to open the gates against any head of water and probably a further capstan to pull a barge travelling upstream against the flash of water that would be released when the gates were opened.

Turf-sided Locks

As was usual with river navigation construction at this period, all of the pound locks (with the possible exception of Woodmill Lock) were originally built as ‘turf-sided’. This means that the sides of the lock chamber between the top and bottom sets of gates would be constructed with sloping earthen banks, at an angle of about 45°, rather than of the brick or stone walls that were usual in later times. When constructed, these sloping chamber sides were usually lined with a thick layer of turfs where the grass and other plants would be encouraged to grow to help bind the soil together.

St Catherine Lock c1850

Pencil sketch of St Catherine Lock with the saw mill on the left. Wooden fendering protects the mill’s waterwheel on the left and the foot of the sloping side on the right.
Image: © Winchester College. Image date: c1850.

It is also very likely that the foot of each turf side would be protected with either a timber or masonry wall to just above the lower water level so that, when filling the lock, the not inconsiderable force of the incoming water did not erode the foot of the chamber side. At Conegar Lock just south of Bishopstoke, such walls made of brick can still be seen.

Some locks on other early river navigations were spanned by one or more “galley beams” or “lintels” that prevented the walls and/or gates of the locks from collapsing inwards (see picture of Spellbrook Lock on the River Stort). It is quite possible that some existed on the Itchen but there is no known evidence that any lintels were ever installed on the Navigation. They were more common where gates were hung on hooks and rides (see below).

On many waterways with turf-sided locks, vertical timbers would have been inserted at intervals along the sides of the lock chamber to prevent boats descending in the lock from settling on the sloping bank although there is no evidence, either documentary or visible on the ground, that locks on the Itchen Navigation had these. Indeed, there is some evidence that members of the crew were accustomed to use shafts (poles) from the lock side to manually keep boats away from these slopes.

Only two turf-sided locks can be still be navigated today in this country and both are on the Kennet & Avon Canal: Garston Lock east of Theale and Monkey Marsh Lock near Thatcham, Berkshire.

South Mill Lock

South Mill Lock on the River Stort was turf-sided until rebuilt in 1923. Note the vertical timbers and the protecting timber walls just above the water level.
Image: Image date: before 1923.

Spellbrook Lock

Spellbrook Lock on the River Stort was turf-sided until rebuilt in 1921. Note the “galley beam” or “lintel” spanning the lock.
Image: Image date: c1908.

Garston Lock

Looking upstream at Garston Lock on the Kennet & Avon Canal. This view clearly shows the retaining timbers at the (high) toe of the sloping turf banks. The iron rails were installed whilst the canal was owned by the Great Western Railway.
Image: © Traveler100 (cc-by-sa/3.0). Image from
Image date: 1 Oct 2007.

Garston Lock

A downstream view of Garston Lock, Kennet & Avon Canal showing the sloping turf banks flooded over when the lock is full of water. This lock was originally built in c1720.
Image: © Graham Horn (cc-by-sa/2.0). Image from
Image date: 3 Feb 2007.

Monkey Marsh Lock

A picture taken from a boat in the chamber of Monkey Marsh Lock, Thatcham, Kennet & Avon Canal. Ignoring the modern metal railings and safety ladders, Itchen locks may have looked something like this. Originally built c1720 and restored/rebuilt 1989.
Image: © Pierre Terre (cc-by-sa/2.0). Image from
Image date: 22 May 2004.

Gates and cills

The lock chambers were enclosed by two gates at each end which were constructed of timber which was probably oak. The gates at the lower end or “tail” of a lock are always larger and heavier as they extend to the full depth of the lock whereas those at the upper end or “head” need only be the depth of the waterway upstream. Each gate would be slightly wider than half of the lock’s width so that the pair would form a shallow ‘V’ pointing upstream. This enabled the gates to withstand the water pressure when water levels on either side were not equal and also helped the gates to seal.

At each end of such a lock, the gates would be hung from a substantial framework as, of course, it is important that the gates form a reasonably watertight seal. It is likely that, in common with other 17th Century navigations, this framework would originally have been made of wood but that later, when the wood deteriorated, this was replaced in brick or stone. For example, the brickwork at the head of Twyford Lane End Lock incorporates the date 1749 (about 40 years after this part of the Navigation was built) - probably too soon to be replacing existing masonry.

Parts of a lock gate

The basic parts of a lock gate.

The vertical timber of a gate furthest from the side of the lock is usually known as the “mitre post”. This must fit snugly together with the same timber of the other gate of the pair to prevent leakage. The vertical timber nearest the lock wall is called the “heel post”. Besides providing the leverage for opening or closing the gate the nearly horizontal “balance beam” also acts as a counter-balance for the weight of the gate. Contemporary pictures of St Catherine Lock on the Itchen show the gates with balance beams (but unusually gates on the River Stort had none - they were originally opened and closed using chains attached to the mitre posts).

Both upper and lower gates, when closed, bed against a “sill” or “cill” at the bottom of the waterway. On the Itchen Navigation, this was sometimes called an “apron”. The inside of the upper cill would be uncovered when the lock is empty, but the whole of the lower one would be permanently under water. The cills on the Itchen were of massive construction in an attempt to ensure that they and their gates were as stable as possible to prevent leaks from developing.

The old type of cill was made of wood. Although the details of construction varied from waterway to waterway, often the foundations of a cill might originally be formed with vertical wooden piles, rammed earth and rubble. Subsequent maintenance may have led to partial or total replacement with brick and/or stone. Over the top of this was often laid a base cill consisting of a block of elm stretching right across the lock and keyed into the lock walls on both sides. Over this would be oak boarding spiked down onto the base cill. This boarding was found at Mansbridge Lock in 1992 during archaeological investigations. On top of this boarding was the structure sometimes known as the “clap cill” against which the bottom beam of the gates would bear when the gates were shut to stop water passing under them.

Wansford Lock, Driffield Navigation

Wansford Lock on the Driffield Navigation, Yorkshire during restoration in 2008-9. This lock has a long top cill rather like locks on the Itchen. A hole in each gate (covered by a paddle on the upstream side) can be seen at the bottom next to the lock side.
Image: Hargreaves Lock Gates.

Clap cill on Lock 58, Kennet & Avon Canal

“Clap cill” for the upper gates of Lock 58 at Crofton on the Kennet & Avon Canal seen before the lock was restored.
Image: © Martin Addison (cc-by-sa/2.0). Image from
Image date: 12 May 1979.

On early navigations, lock gates were hung like field gates on “hooks and rides” (hinges) from a post attached to the lock side but with this system it would have been difficult to maintain a watertight seal. The use of galley beams or lintels (see above) was more common on waterways that utilised hooks and rides.

Locks in use today use a system whereby the face of the heel post nearest the lock side is rounded to fit a “hollow quoin” in the masonry within which it partially rotates. Inserted in the bottom of the heel post is a round “tan pin” which can rotate in a cup or “pot” set into the cill. The top of the gate is kept in position by a metal “collar” which is fixed to an “anchor” fastened into the masonry at the top of the lock wall. Looking at the remains of the turf-sided locks, it seems highly likely that the lock gates on the Itchen Navigation were hung using hooks and rides - at least originally.

Hook and ride (hinge)

Hook and ride hinge. Also known by
various names such as pintle or
gudgeon hook and band or strap.

Tan pin in heel post

Tan pin in the bottom of a rounded heel
post on the Rideau Canal, Canada.
Image: © 2001 Ken W Watson.

Fitting a collar

Fitting a collar round a heel
post on the Rideau Canal, Canada.
Image: © 2007 Ken W Watson.

Anchor holding gate

An anchor with collar disappearing around
the heel post on the Leeds & Liverpool Canal.
Image: © 2014 Mike Clarke.

Hollow quoin in lock masonry

Hollow quoin in lock masonry on the
Rideau Canal, Canada, viewed from above.
Image: © 2009 Ken W Watson.

The distance between the top and bottom sets of gates seems to have varied between about 90 and 100 feet but the top cill was often 10 or more feet long. The top cill is in effect the step in the waterway that the boat overcomes by means of the lock and stability of this part of a lock is crucial. In the 17th Century with little understanding of the forces involved, a massive structure would have been built in an attempt to ensure stability.

Other fittings

To make a lock functional, there must be means of getting water into and out of the lock chamber. On early waterways, this was achieved by means of a sluice, mounted on the upstream side near the bottom of a gate. This sluice could be raised to allow water to pass through a hole in the gate. Today this would normally be called a “gate paddle” but many waterways had their own name for this device. The paddle was connected to an arm or “rod” passing up the gate to a simple, vertical piece of wood with handles and a series of holes which with a peg could be secured in a raised position. Filling the lock would involve raising the paddles in the top gates to allow water from the higher level to pour into the lock using gravity. Similarly, the lock would be emptied into the lower level downstream of the lock. It seems that the top gates of the locks had an open aperture or vent at the top of the gate through which excess water in the pound above could discharge into the lock to help regulate the water levels. To aid this, the normal method of operation on the Navigation was to leave a lock with the top gates closed and the bottom gates open.

Finally, to protect the masonry at the head and tail of the lock from the barges entering or leaving the lock, wooden fendering would often have been fixed to the brick or stonework at about water level. This was achieved with very large iron nails driven through the wood into the wall behind. There is an example of such a nail remaining at the head of Mansbridge Lock on the offside.

An old paddle board

An old paddle board.

A modern paddle board

A modern, raised paddle showing the uncovered hole in the gate.
Image: © 2007 Neath & Tennant Canals Trust Ltd.

Bar to operate paddle

Simple paddle bar at Worsfold Flood Gates on the River Wey. When the paddle is raised (opened) a peg secures it in position.
Image: © Graham Horn (cc-by-sa/2.0).
Image from
Image date: 25 Aug 2008.

These methods of construction had an major advantage: they were relatively cheap. A turf-sided lock tends to leak water through the earthen sides but, although a matter of concern on an artificial canal where water supplies may be limited, on a river navigation where river water was usually fairly plentiful this was a minor consideration. Whilst cheap to build, the lock will use more water than a traditional brick lock of the same size. In the 17th Century, bricks were expensive and in southern England suitable building stone was not readily available at reasonable cost. Transporting heavy goods by horse and cart on poor roads was expensive - indeed, this is a major reason why water transport started to be developed at this period.

Thus we find that river navigations such as the River Wey (opened in 1653 from Weybridge to Guildford), the River Kennet (opened in 1723 from Reading to Newbury) and the River Stort (opened from the River Lee to Bishops Stortford in 1769), amongst others, were all originally built with turf-sided locks. In almost all cases these locks were rebuilt over the years with masonry chamber walls. There are just two examples of turf-sided locks still in use in England: at Garston Lock and Monkey Marsh Lock on the River Kennet section of the Kennet & Avon Canal between Reading and Newbury.

Allbrook Lock

The Allbrook Lock that we see today north of Eastleigh is not turf-sided. The original turf-sided lock was replaced on a new site when the London and Southampton Railway (later to be renamed the London and South Western Railway) was being constructed in about 1837-8. The railway between Northam and Winchester was opened in 1839. The railway embankment came so close to Allbrook Lock that this had to be rebuilt in a new position by the railway company at its own expense which partly explains the new lock’s comparatively good condition today.

The lock was constructed in the style of the day with brick chamber walls and only just long enough for a seventy feet long boat. The gates were hung using hollow quoins and collars (see above). Although the collars are now missing, the anchors for the upper gates remain fastened into the stonework and the hollow quoins at the head of the lock are now obscured by modifications to the weir made during the summer of 2003. The widened road bridge at the tail of the lock obscures the gate fixings there. Pictures of the lock can be seen here.

Woodmill Lock

Woodmill Lock is no longer visible having been filled in during road works towards the end of the 19th Century. It was a sea lock with a masonry chamber running east-west. It is possible that this lock was originally turf sided but it was certainly reconstructed in 1829 with bricks. In this form, it had the usual two pairs of lock gates pointing upstream to retain water in the non-tidal river, plus a third pair, pointing downstream, to prevent very high tides from flooding the Navigation with sea water. A wooden bridge across the lock was also built at around this time, but by the middle of the century was in poor repair. Pictures of Wood Mill and the site of the lock can be viewed here.

The Barges

The Locks

There are no known boats or barges remaining in existence from the days of commercial use of the Navigation. The author is aware of only four contemporary images of the barges but these, although not inconsistent with each other, are not as clear as one might wish. Also it cannot be told how much artistic licence is involved in any of these pictures as none of them is anything like a detailed technical drawing. Previously, very little has been written about the barges on the Itchen Navigation - certainly not in detail.

The description given below these pictures is derived partly from them but also from a few documentary sources and comparison with cargo carrying craft on other waterways. The website manager would be grateful for any information that might improve this description, particularly any more pictures of these intriguing boats.

How many barges?

Records indicate that no more than six barges were ever registered for use at the same time on the Navigation and this was during the wars with France in the early 1800s. It is quite possible that there were one or two unused barges in addition. The useful lifespan of a wooden, commercial craft may be limited to about 30 years after which replacement might be desirable. The design of any replacement might evolve slightly but radical changes were unlikely on a small, isolated and not very properous waterway. It is likely that the barges were built and maintained at the dock that once existed on the west bank of the Navigation almost opposite Domum Wharf, Winchester.

Painting of barge below Tun Bridge
An extract from an oil painting showing Tun Bridge and a barge pulled by two horses.
Painted by Frederick Waters Watts in 1843.
Water colour of barge near Domum Wharf
An extract from a water colour of a barge near Domum Wharf.
Painted by Richard Baigent in 1837.
Picture of a barge at Blackbridge Wharf
An extract from a picture of a barge at Blackbridge Wharf.
By an unknown artist and probably dating from before 1795.
Stylised etching dating from 1767
An extract from a stylised etching.
Image dating from 1767.


Written accounts say that the barges were about 70ft (21.5m) long and around 13ft (4.0m) wide. No details of draught are available but they are likely to have been in the order two and a half feet (0.75m) when laden and probably less than 1ft (0.3m) when empty. Loads carried seem to have normally been in the region of 20 to 30 tons as the waterway was often shallow through lack of dredging. Some of the Acts of Parliament regulating the waterway state that vessels capable of carrying between 20 and 45 tons were permitted. Most of the bridges spanning the Navigation were wooden. Headroom was very limited at all the bridges as evidenced by the two remaining masonry bridges at Mans Bridge near Swaythling and at Wharf Bridge at Winchester. This website also has photographs of two of the original wooden bridges of which one at Brambridge can be seen here and the other at Tun Bridge near Winchester on this page.

The journey via the Navigation from Winchester to Wood Mill was 10.4 miles (16.7km) with 15 locks. This would probably take 8 or 9 hours. A trip upstream would probably take a bit longer against the flow of the river. The tidal river down to Northam Quay is 2 miles (3.2km) and passage would likely be made within an hour with the tide. However, it might have been necessary to wait for the tide to turn before attempting passage (see also the section on The crew below).

Construction details

All of the four pictures above show the barges as rather simple, punt-like craft with no permanent cabin. The hulls had a square-cut, sloping bow or ‘swim’ and a square, vertical transom at the stern which supported a long, external rudder. The sides of the barges rise at right angles to the bottom. This form of hull is not unlike river boats from an earlier age. For example, ‘Western Barges’ trading on the River Thames upstream of Oxford were built on very similar lines dating back to medieval times although the stern was the same shape as the bow and they stepped a single pole mast to raise a square sail. However, from the evidence given by the master of one barge (see below), a sail was sometimes used on the tideway - presumably this would have been jury-rigged as no permanent mast is evident.

Shelter appears to have been provided by a canvas awning or ‘tilt’ at the stern supported by semi-circular hoops and this cover seems to have been removable. It is possible that this shelter would be used as living accomodation if the crew were not mooring near home. It is not known whether any covering to protect the cargo was available if required.

Punt with no kelson or thwartsThe inside of a punt with no kelson or thwarts

More comprehensive details of the barges’ construction are not known. Because they were trading in shallow water, it is extremely likely that they had no keel and were flat bottomed. Possibly there would have been a kelson (or keelson) that ran along the centreline inside the boat from bow to stern. This would have been crossed from side to side by beams at right angles joining to the chines which run the length of the vessel where the bottom meets the side at right angles. Alternatively, there might have been no kelson, in which case the bottom planks would have been a bit thicker and have run along the length of the barge rather than across the boat. In this case, the planks would then have been attached to the beams running across the bottom of the vessel.

Attached to this framework would be knees to hold the sides of the boat to the bottom. The bottom would likely be lined with wide boards several inches thick, probably elm, nailed to the outside of the framework along the length of the boat. Very likely, at about gunnel level, there would have been a number of thwarts (boards maybe a foot wide) fastened across the boat at maybe 12 or 15 feet (3.5m or 4.5m) intervals to help prevent the sides being pushed outwards by a loose cargo such as coal or even inwards when empty.

Finally, there would almost certainly have been a sectional and removable wooden false floor in the hold a few inches above the bottom planks, usually called shutts or shoots. These kept the cargo out of any water in the bottom of the barge. They would also protect the bottom planks of the boat when shovelling out loose cargo such as coal or chalk.

Picture appears to depict the top of some knees on the far side of the boat a little aft of the bow. This picture also seems to show a modestly refined bow shape with the gunwales rising slightly towards the front and the bow narrowing a little.

Picture shows one of the boats with a form of foredeck across the width of the bow. A second barge shows a much narrower feature at its bow. It is possible that such a foredeck is just visible in picture . This picture also shows a mast a little astern of the bow which is not shown in any of the other images. Presumably this was a towing mast. Ideally, when towing a boat from the towing path at the side of the channel, the rope should not be fastened at the very front of the vessel as this will tend to pull the boat towards the bank unless the steerer is constantly steering away. To overcome this problem, the tow rope should be fastened to a point on the boat’s centreline maybe a fifth or a quarter of the boat’s length back from the bow - the exact distance largely depends upon the shape of the hull underwater. A mast will also help lift the rope to minimise it fouling on the boat, it’s contents or vegetation on the bank.

The rudder on an Itchen barge would have been quite long as evidenced by the pictures. With the barges having a shallow draught, it would have been necessary to increase the length to maintain the underwater area of the rudder needed to effectively steer the vessel. Although not shown on the pictures, the tiller would also be quite long to provide greater leverage.

Picture also shows a smaller boat near the wharf about the size of a modern leisure punt or slightly bigger. It seems to be carrying some cargo and have a crew of three. No further information about this sort of boat on the Itchen Navigation is available.

The animals

An often overlooked part of the system was the motive power to take barges along the waterway. The animals usually used for towing boats were horses but on some waterways mules or even donkeys were used. It is not known which animals were used on the Itchen. From pictures and it appears that two animals were used in single file as the towpath was rather too narrow to be used with them abreast. Going upstream on some of the river sections the current can get quite strong, particularly in winter, and when fully laden, the load might prove too heavy for a single animal. It would have been usual for someone to walk and control the animals but it was not unknown on some waterways, where the same route was often repeated, for some horses to be capable of walking by themselves knowing when to stop and start. Animals were not used on the tidal section of the Navigation between Northam and Wood Mill as there was no towing path.

The crew

On the non-tidal sections of the Navigation it would have been possible to travel with a crew of just two: one to see to the animals and one to steer the barge. But more hands would have made life much easier, quicker and safer, especially at locks. On the tidal section, however, more crew would have been essential. When the Southampton to Netley Railway was under consideration in the early 1860s, Charles Penton, captain of a barge for many years, gave evidence to Parliament about the bridge crossing at St Denys. He stated that the barges never went upstream against the tide but sometimes went down against it. “I and the other captains of the barges sometimes used a Sail on the tidal portion of the river but it appears to me that the projected embankment of the Railway will interfere with our doing so for the future.” It would seem that usually they used the tide for propulsion and shafts to control and steer the barges. It also appears that crews of four or even five were not unusual.