Wednesday, 29 June 2016


Today marks HMS Terror’s two-hundred and third birthday - the anniversary of her launch in Topsham, Devon on June 29th, 1813. Within a year of her launch, Terror would be harassing American cities along the Eastern Seaboard, and would even have an epic poem written about her exploits by Francis Scott Key. Today, when Americans sing their national anthem, they reference the history of our favorite exploration vessel.

This date also marks the third anniversary of the start of the Building HMS Terror blog, and today we’ll be celebrating our respective anniversaries by discussing Terror’s pumps.

By 1845, Terror had at least five pumping mechanisms installed, though it likely had other moveable “fire pumps” as well. For her size, Terror had a relatively large number of water management devices. Comparing the 1836 upper deck and profile plans of Terror to the 1839 plans shows that an additional two common pumps were installed just behind the forehatch, while the main pumps flanking the main mast were upgraded and replaced. 

These changes were likely a response to the near sinking of Terror during George Back’s arctic expedition of 1836-1837. In the spring of 1837, ice damage to Terror’s sternpost and keel created leaks so severe that five feet of water gushed into her hold every hour. Back ordered his men to work the pumps continuously during Terror’s return voyage across the Atlantic. The crew became so exhausted that they had to beach the vessel at the closest landfall, at Loch Swilly in Ireland.

Rice, the shipwright responsible for Terror’s 1839 refit, responded to this near disaster by upgrading all of Terror’s pumping systems. He also introduced penstocks into the limber board system in the hold, which allowed the crew to manage the flow of bilge water into her well.

Below, I’ll discuss each of Terror’s new pump systems in turn.

Massey’s (Massies) Patent Pumps (Bilge):

Massey’s patent pumps were a reliable flywheel pump system of a type that became very popular on civilian and navy vessels in the latter half of the 19th century. Testing of Massey’s pumps began in 1833 and timed trials on board HMS Thunderer showed clear advantages over traditional chain pumps. Water discharge rates over short durations were similar to the chain pumps, but at greater periods of time, the Massey pumps outpaced the chain pumps significantly (1). The main advantage of the flywheel design seems to be that “it [did] equal work with less fatigue to the men” (1). The trial was so impressive to the Admiralty that they installed Massey’s pumps on HMS Vestal in 1834 (2), and increasingly on Royal Navy vessels thereafter.

Massey’s pumps were a double action “lift and force” pump, consisting of a camshaft driven by two crank handles. The camshaft drove two 18 inch piston rods that powered the pumps. A heavy iron flywheel was mounted on the fore end of the camshaft, and, once in motion, it assisted in maintaining the momentum of rotation (thereby making the crank handles easier to turn). Unlike chain pumps, Massey’s pumps were very difficult to  clog, and could “…discharge a block of wood 9 or 10 inches in diameter”(1).  This was obviously the perfect pump for an arctic expedition vessel.

My plans for the Massey pumps on board Terror are derived from measurements shown on the 1839 plans of Terror and Erebus, with additional information gleaned from the somewhat more detailed HMS Investigator plans. Cross sectional details, especially of the fly wheels and piston rods, were derived from historical images of similar flywheel pumps, as I was unable to locate the patent for Massey’s 1833 pump design.

Plans for Massey's Patent Pumps, as installed on HMS Terror in 1839.

Common Pumps (Bilge):

Terror was fitted with two common, or suction pumps, also known as "elm tree pumps" due to the use of a single bored-out elm trunk as their barrel or tube.  Elm was used because of its general resistance to water, though other water-resistant woods could be employed. The pumps were extremely simple, consisting of a brake (or handle), a spear (or piston) and two valves, and were thus very easy to make and repair. The advantage of placing two common pumps in the fore of the ship was that the fore hold could be pumped out independently of the rest of the ship; which could be critical in a situation such as Back found himself during the 1837 Atlantic crossing.

Plans for HMS Terror's common pumps, as installed on HMS Terror in 1839.

Truscott’s Pump (Fresh Water):

Truscott’s pump was a simple invention that revolutionized the way water was stored and retrieved on Royal Navy vessels. Inspired by a visit to an ale house in 1812 (3), Truscott designed a relatively simple iron pump attached to a small diameter pipe system that led to the hold. Just like in the ale houses, he attached a flexible leather hose to the end of the pipe and fed it into the water casks. This obviated the need to move the casks to retrieve water. This was a major boon on sailing vessels, because moving casks to retrieve water was time consuming, dangerous, and inevitably impacted the ship’s trim, requiring periodic rearrangement of the ship’s casks and ballast. This simple invention eventually led to the use of permanent iron water tanks on sailing vessels, which ultimately resulted in the abandonment of shingle and iron ballast.

A plan of Truscott’s pump (ZAZ6848), dated 18th September 1814, is held by the National Maritime Museum.  The 1839 lower deck plan for Terror and Erebus indicates that one Truscott pump was located at the rear of the ship’s stove, on the starboard side, close to the door to the sickbay.  This pump permitted the crew and the ship’s cook to access fresh water as it was need. Since the water tanks in Terror’s hold surrounded the Sylvester stove/furnace, it is possible that the device could have be used throughout the winter months.  However, it is likely that the fresh water tank above the Fraser stove provided enough for the ship’s needs without using the hold tanks (at least during the winter months when fresh water ice was available).  

Truscott's pump (in red) as installed on HMS Terror in 1839.
The ship's water tanks and Fraser's stove appear in blue. 


1. The Nautical Magazine: A Journal of Papers on Subjects Connected with Maritime Affairs in General. 1833. Page 292.

2. Sharp, James. 1858. Memoirs of the Life and Services of Rear-Admiral Sir William Symonds, Surveyor of the Navy from 1832 to 1847. Longman Brown Green Longmans & Roberts, London. Page 153.

3. Transaction of the Institution of Architects. 1865. Page 191.

Monday, 30 May 2016


“Fairing” is an important step in constructing an accurate hull shape on a ship model. It involves sanding and beveling the bulkheads to create an even surface for the planking to sit flat against the hull frame.

For my build, this was an extended process that allowed me to learn much about the construction of Terror and Erebus, especially at the stern and the bow.

In this post, I will present a photo essay documenting the steps I took to prepare the hull for planking. As I progress through the steps, I'll discuss what I learned about Franklin’s exploration vessels. 

Before I could begin faring the hull, it was necessary to fill in the
stern and bow of the model using filling blocks. I created these
from 1/4" plywood, laser cut using measurements from the
ship's plans. This image displays the three starboard filling
blocks used at the bow. 

The filling blocks were carved to shape using card
guides cut to match the lines of the half breadth plan. 

Placing the filling blocks side-by-side as they were carved ensured that they
were symmetrical. The lamination in the plywood was also helpful
 in this regard. However, plywood is a poor carving material,
and I would think twice about using it again. 

In 1839, the solid chock (ice) channels on Terror were extended around the bow. I
constructed these from several layers of basswood. 

These chocks were then shaped to match the proper cross section of the channels. 

The ice channels were glued in place on the bow and scrap wood 
was used to rough out the bulwark shape. The excessive
use of glue didn't escape Mini-Crozier's critical eye. 

The gaps in the bulwarks were filled using basswood strips of appropriate thickness. 

Rather than filling and sanding seams and gaps, basswood leveling strips
were applied to the upper surface of the ice channels. The channels
were then filed to shape using card stock templates. Scrap wood was
used to fill in any large gaps in the bulwarks. 

The completed bow just prior to sanding. I checked the symmetry
and level of each side of the ice channel obsessively with a height
gauge while the model was still on its building board. The
asymmetry of the filling stock used to shape the bulwarks
is a product of the odds and ends in my spoil bin, and while 

unsightly, it won't be visible when the model is planked. 

This image shows the faired forward bulkheads and bow filling blocks, just prior to final
sanding. The merchant-like shape of Terror's bow and the imposing nature of the ice 
channel grafted to it can be seen in this view. Note how far the ice channel 
overhangs the bow relative to the port side of the ship; this is because it 
sits on three layers of planking, including a layer of 3" lower planks,
 a second layer of 8" planks, and a third layer 
of even thicker reinforcing planks. 

An image of the faired stern, detailing the single filling block used
in this area. The stern rabbet is in the process of being finalized

in this image. 

With the hull faired, the stern timbers could be installed. These
were laser cut from Swiss pear.  The outermost stern timbers,
on the left, were cut in two sections, as they form an angle
when installed correctly. 

Prior to describing how the stern timbers were installed, it is important to note how this area of the ship was designed by Oliver Lang, the shipwright who refit Erebus and Terror for the Franklin Expedition. Because of  the massive size of the propeller well and the rudder post which formed its aft wall, Lang had little room left to fit the six stern timbers and four stern lights (windows) in the counter. His solution can be seen in the 1845 Erebus and Terror stern model in the collection of the National Maritime Museum, in Chatham. 

Inspection of the propeller well in that model shows that the stern timbers were actually used to form side walls of the well. However, they could not be fayed directly to the sides of the rudder post as this wouldn't leave enough space for the stern lights (windows). This meant that the stern timbers had to be inset into the sides of the rudder post by three inches to form the side walls of the propeller well. Remarkably, Lang achieved all of this with almost no modification of Terror's existing stern framing. With the rudder post locked directly into the two central stern timbers, the whole structure was incredibly robust.  It is important to note that the inset stern timbers may not have been needed on Erebus, which had a slightly wider counter than Terror

Simplified plan of Terror's counter architecture. Note how the stern timbers overlap
the rudder post. Also noteworthy is the position of the upper deck transom,
which could be fayed directly to the aft side of the rudder post in this configuration.  

In this image, the slot/inset for the stern timber has been cut into the rudder
post. Note how it is level with the interior sides of the propeller well. 

A view from the aft side of the rudder post showing the insets for the stern timbers. 
Note the square slots in the stern filling blocks cut to accept the heels of the stern timbers 
(no wing transom was required for construction for this stage).

Checking the fit with a stern timber. This won't be visible in the finished model. 

Checking alignment. 

The stern timbers were fitted with the help of a jig. The jig was designed to be
clamped to the bulwarks, using the station lines printed on it as guides. 

Detail of the aft part of the jig. 

A height gauge was necessary to ensure that the jig was properly aligned along its aft margin.
This gauge slid tightly over the aft support of the building board, using the tracks on the left. 

The jig and height gauge in place, with the center two stern timbers
installed and clamped. 

"Wing transom" filling pieces. These are not entirely accurate architecturally 
(they are more like half-transoms), but were carved and sanded to shape 
to provide a platform for planking the stern. 

The "wing transom" in place. Note the very slight curve in the transom. As confirmed
by the 1845 stern model and the 1839 model of Erebus, Terror's
stern was very square indeed.

Rough transverse framing was installed to support the stern timbers. This 
framing is not accurate to plan or scale but rather simply supports the structure
and will not be visible when the model is planked. See the above plan for the
 correct framing. As with the bow, I obsessively relied on a height gauge to
ensure the entire structure was level and square. 

A port side view, detailing the stern architecture. Note how the stern
timbers adjoined the propeller well and rudder post.  

The completed stern. 

Completing the construction and fairing of the model's stern was a milestone for my project. Not only is the model now ready for planking, finishing this stage of the build revealed a minor mystery surrounding how Lang planked Terror's stern . Lang's 1845 stern refit plan stated that an "....additional part of the wale [was] added to the after end of the ship to form the well or trunk..." for the propeller. Unfortunately, his plan does not reveal if both layers of planking were extended to accomplish this (Terror was double planked against the ice). However, with the construction of this part of the model, his solution became clear to me.

If my model is correct, then it shows that the first layer of Terror's hull planking did not need to be modified in any way by Lang.  In fact, it could simply be left in place, terminating at the edge of the lower counter, as was typical of bomb vessels. Again, if my model architecture is correct, then it shows that Lang could have just extended the second layer of planking to the rudder post. The 1845 stern model shows that this planking rose straight up the rudder post and, when it hit the counter, turned to trace a graceful arc, running from the upper end of the stern rabbet to the lowest portion of the counter at the sides (these planks were fayed directly to the previously planked counter). Lang's stern plan shows that the second layer abutted a beveled margin plank on the counter, although this isn't detailed on his stern model. 

My planking plan for Terror's stern. The red lines show the lower planking, while
the white lines show the upper level of planking. The overlap of the planks
accords well with the 1839 midships section for Erebus and Terror

The current condition of Terror. She's just about ready for planking. 

A view from the upper deck. 

Sunday, 27 March 2016


In early May 1845, just prior to the departure of HMS Erebus and Terror on their final voyage, a number of the Franklin Expedition officers sat for a remarkable series of daguerreotype portraits. Taken by the renowned London daguerreotypist Richard Beard on the deck of HMS Erebus, most of the images include a cloth backdrop which obscures the ship’s architecture. However, one portrait, of Lieutenant Henry T. D. Le Vesconte, did not use a backdrop; this omission resulted in the only known contemporary “photographic” image of Erebus or Terror.

In the image, Le Vesconte appears to be sitting on the starboard quarterdeck of HMS Erebus, just in front of the ship’s mizzen mast, skylight, and wheel (note the original daguerreotype is a reversed image). Out of frame, in front of his left elbow, would have been a small charting table bolted to the deck. Le Vesconte may have used a stool or chair associated with that table to sit for his photograph.

I believe this location was carefully chosen for the photograph. In the Royal Navy, the starboard quarterdeck was considered to be an almost sacred location - reserved only for the ship’s master and his officers. On Erebus, it was the location where Franklin and his senior officers would have issued orders, considered routes, and charted coastlines. It was the nerve-centre of the entire Franklin expedition. It is astonishing that such an image has survived to the present day.   

Components in the Le Vesconte Scene

I have been modelling 1/48 scale versions of the ship's wheel and skylights for my HMS Terror project, and realized that I had the opportunity to recreate the famous Le Vesconte image. To model these fittings, I began with the original ship’s plans and used the Le Vesconte daguerreotype to provided key information. Further information was gleaned from the 1839 Erebus model in the collections of the National Maritime Museum. As I will describe below, one of these structures was recently discovered by Parks Canada near the wreck of HMS Erebus and the publicly released image of that artifact provided critical information.

The Skylights

HMS Erebus and Terror each had two skylights located on the aft of the upper deck. Both follow a design originally adopted by HMS Terror in 1836.  The sides of the skylights were rimmed with panes of glass to allow sunlight into the captain’s cabin and officer’s mess. Unusually, their tops had no peak or even a slight camber – attributes confirmed by the daguerreotype and the NMM model of Erebus. Interestingly, the daguerreotype indicates that the panes were not protected by brass rods as was standard on many Royal Navy vessels of the era – an indication of the peaceful aims of  the expedition. In addition, the aft skylight on HMS Trincomalee, which is very similar to those used on HMS Erebus and Terror, provided important details for my recreation. 

The 1836/1837 and 1839 plans for Terror/Erebus indicate the skylights may have had collapsible sides, at least on the starboard. This would have permitted ventilation, if necessary, though the plans seem to indicate the ultimate aim was to enable the skylights to be used as makeshift companionways.  In this scenario, it is likely the roofs of the skylights would have been removable. I opted not to model these aspects of the skylights, as I had no information on how these features were designed.

Laser cutting the parts for the skylights. The wood is swiss pear. 

Parts compared to the plans.

The completed skylights. 

Mini-Crozier stands next to the largest skylight. The window panes have been
sanded on their interior sides to simulate frost.

The Ship’s Wheel

The 1836/1837 and 1839 plans of HMS Terror and Erebus show that their wheels were slightly smaller than those typically used on Royal Navy ships, being more consistent with those used on merchant vessels of similar size. The recently recovered portion of  Erebus’ wheel, as well as the Le Vesconte daguerreotype, reveal that the wheel itself was relatively plain and was held together by copper alloy screws or nails. Fine examples of this type of  wheel can be found today on HMS Unicorn and HMS Trincomalee. Trincomalee’s wheel is perhaps closest in design, with distinctive grooved felloes nearly identical to those on the HMS Erebus wheel found by Parks Canada.

Like most large Royal Navy vessels, Erebus and Terror’s wheels had ten spokes, each radiating at 36 degrees from the barrel. Usually, survey vessels ships of this size would have had an eight spoke wheel; the extra spokes would have substantially increased the strength of such a small wheel, and perhaps it was deemed necessary for arctic exploration.

The Le Vesconte daguerreotype indicates that the top of the wheel pedestals were protected by a very heavy moulded brass guard plate. This can be confirmed by the subtle reflections seen on the plate in the daguerreotype. HMS Trincomalee has similar heavy copper guard plates on its wheel pedestals.

Laser cutting the wheel components.

Construction was aided by the use of a jig. 

The completed wheel; cleanup and sanding is still required. The spokes were shaped by
hand, without the benefit of a lathe. 

The completed pedestals, with the brass guards in place. Limited information
on the pedestal hubs was available, so here they are made as simple brass tubes.  

The completed ship's wheel, after applying a coat of Minwax poly.  

Another view of the completed wheel. 

Recreating in the Le Vesconte Image

The mocked-up components of the Le Vesconte scene. 

Mini-Le Vesconte sits on Erebus' sacred quarterdeck. 

Comparing the two images is rather satisfying. While it appears that the pedestal on the model wheel is larger than the wheel on Erebus, I believe this is a matter of perspective and differences between camera lenses, as the dimensions of the wheel barrel are based directly on the plans. It's also possible that Mini-Le Vesconte is placed too close to the mast in my reconstruction. However, it appears that I may have misjudged the size of the panes of glass in the skylights on my model which are somewhat taller on the daguerreotype. I’m okay with that small difference.

Mini-Crozier stands at his post on Terror's quarterdeck. 

Thursday, 3 March 2016