When the Kingdom of Tet Satou heard very believable rumors that a longtime rival, the Empire of Elbonia, was constructing new, 4000-ton, twelve-knot ironclad warships, it put the government in a bit of a quandry. While they knew that they did not yet have the industrial base or facilities to support ships that could defend against the rumored new ships, at the same time, the threat that they posed was too great to ignore. The Army proposed a viable solution: defend the nation's major coastal targets with coastal artillery located in shoreline fortresses. Almost all vital locations and harbors could be easily covered by such... with one major exception.
The largest, most important seaport for the entire Kingdom, Taulaga Muamua Sami, was located on a natural harbor with a mouth three miles wide... but the best coastal guns available, though able to fire such distances, couldn't be
accurately fired further than about one mile. Even if coastal fortresses were installed at either end of the harbor mouth, they couldn't effectively cover the entire width of the harbor, much less provide an interlocking field of fire for its entire width. Fortunately, the Army also had a solution to this issue. As the harbor mouth was about 25 feet deep, they proposed, it would be entirely feasible to build coastal fortresses on a pair of large, specially-constructed barges, which could then be towed out to the harbor mouth and, using ballast tanks, be sunk to sit on the harbor bottom, essentially forming artificial island fortifications that would provide the entire harbor mouth with interlocking fields of fire from "coastal" guns. Should major repairs of the barges become necessary, the ballast tanks could be pumped out to refloat the fortresses, and they could then be towed back to land to be careened; resupply of the soldiers stationed at the fortresses could easily be accomplished by boat.
A number of people in the fledgling Navy had doubts about the feasibility of such a scheme, arguing that it would be better to use floating batteries or monitors in the role, but the Navy was still young, and wasn't going to be able to provide such a solution for at least seven years, by its own estimates. Thus, the Army proposal carried the day, largely by virtue of being the only option that could be ready in time for the delivery of the anticipated Elbonian ironclads.
The final design of the fortress barges proved to be quite large indeed, with both length and width in excess of 100 feet, and extending twenty feet below the waterline when afloat, with a designed "emplaced" draft of 25 feet as it sat on the bottom. In fairness to the Army, the fortress itself was quite well designed, with iron armor eleven inches thick over the vital parts of the structure, backed up by eighteen inches of old-growth oak, with two feet of oak on the unarmored areas; the armor on all forward-facing surfaces (and the sides of the spotting tower) being heavily sloped to attempt to deflect cannonballs up and over the fort. Much of the volume enclosed below the waterline was taken up by the ballast tanks used to settle the barge onto the bottom, but there was still a significant amount of space below the waterline available for storage, particularly of shell, powder, and coal.
The fortress itself was heavily armed: two fifteen-inch Rodman columbiads in casemate mounts atop the structure, four ten-inch Rodmans on a gun deck below that, and eight Whitworth 70-pounder rifles on a second lower gun deck. All of these guns were designed to have a field of fire of about 135 degrees, ranging from directly abeam the fortress on their side to about 45 degrees to the opposite side of dead ahead. In addition, two Ericsson-style turrets were mounted on either side of the spotting tower. Each of these turrets had 360 degrees of traverse (and about 270 degrees of clear firing arc) for the two 11" Dahlgren guns installed in them; the turrets were suitable for upgrade to 15" Dahlgrens should such be necessary.
A pair of boilers at the level of the lowest gun deck provided steam to power both the turrets and an ammunition hoist that was the primary method of supplying the 15" Rodmans; the hoist also could stop at either of the lower gun decks to transfer ammunition for the smaller guns to and from bulk stowage below, but in combat, the smaller guns would primarily be supplied by muscle-powered block and tackle hoist systems that resupplied the ready storage racks. The spotting tower was also equipped to control command-detonated torpedoes (mines), as a supplement to control sites on land, on the grounds that visibility from the forts at sea could be better under certain conditions. Provisions were also made to later equip the fortresses as anchor points for harbor defense chains or booms, but these were not initially fitted.
A significant dock area was provided at the rear of the fortress, to accommodate the boats that would be used to ferry personnel and supplies to it. The Army's plan was for these fortresses to remain emplaced for extended periods of time, being careened perhaps once per decade; thus, they specified that the entire barge be coppered below the waterline, to protect it from being destroyed by borer worms. Quarters for the officers assigned to the fort were provided below the spotting tower, as was a mess hall for all personnel; the enlisted men were quartered in the two lower gun decks.
The first of the two fortresses is shown, along with a typical example of the harbor tugs used to position the fortresses in the harbor mouth before the ballast tanks were flooded through the large vents on the sides of the barge.
The project, naturally, ran over budget and behind schedule, and the Navy continued to call for killing the barge-based fortresses in favor of "cheaper" (but later) ship-based defenses. However, despite this, and despite some questioning whether the purported Elbonian ironclads would ever materialize, a combination of bureaucratic momentum and a feeling that
something had to be done to counter the perceived threat saw the project completed, with the two fortresses being initially emplaced in early November 1887. Beyond a quick repainting of their exposed iron surfaces from black to white, following the discovery that a black iron box would heat up like an oven in the sun of a tropical summer, the fortresses were found to be entirely serviceable and required no major changes for the first few years of operation.
However, this would be changed by the Great Typhoon of 1871. While the structures themselves weathered the storm well, storm-driven waves blew in some of the shutters over the lower deck gunports, resulting in large amounts of water being shipped aboard and causing internal damage, including to the gun mountings. However, this wasn't the worst problem from the storm; after it had passed, it was found that both fortresses had been pushed more than fifty feet towards shore from their originally planned positions. A number of people familiar with copper-bottomed vessels expressed concern that this could have damaged the sheathing as it slid across the seabed, and the Army decided that both fortresses should be careened for bottom inspection.
The worries only became greater when divers were sent to close the ballast tank vents in preparation for refloating the barges; the divers reported that there was now significant undercutting of the harbor bottom extending beneath the two fortresses. Both were refloated and taken ashore, but inspection and repair was delayed while the exact nature of the undercutting was studied. It was soon determined that the undercutting was the result of scour by the currents produced by the typhoon as it came ashore; whatever solution was undertaken to ensure there would be no further slippage of the fortresses, it would also have to protect against storm scour undercutting them to the point that they either sank or toppled.
One proposed solution involved driving a large number of iron pilings into the seabed until they reached bedrock, then placing the repaired barges on those pilings with interlocking pins holding them in place. However, this was deemed to be too expensive a solution. Instead, the Army chose a different route: the holes in the harbor floor would be filled in, the fortresses placed back in their original positions, and then large piles of fill and riprap would be used to permanently anchor the fortresses in place, effectively turning them into a pair of artificial islands. The Army proposed using rubble and debris left over from the typhoon as a large part of the fill material, with the rest coming from operations to dredge a deeper channel between the two fortress locations; this plan was approved in October 1871.
With this, a number of changes had to be made to the fortress barges. No longer could they be refloated and careened on a regular basis; they would become permanent parts of the harbor and thus would have to be made able to last indefinitely. The most visible change would be the removal of the valuable copper sheathing, to be melted down and reused, instead replaced with multiple thick layers of a paint based on natural rubber found in the nation, intended to permanently waterproof the wood planking and thus prevent rot. Other changes included reinforcing the structural framework supporting the fortress, and removal of the dock structure from the rear of the fortress, as this would no longer be of value in docking supply vessels. Additionally, a lighthouse structure was mounted atop the spotting tower, to warn ships (during peacetime) away from the new artificial islands.
Following the modifications, the two barges were returned to the water and towed back to their now-restored positions, then sunk via flooding the ballast tanks to emplace the fortresses again in July 1872. Following this, there was an extended period where fill dirt from the channel dredging was poured into the ballast tanks, to more firmly weight down the barges, followed by the dumping of typhoon debris and fill material to produce the anchoring artificial islands, which was then followed by dumping riprap around most of the islands to protect the islands from the possibility of storm scour. (A small area at the "rear" of the island was left clear of riprap, to provide a safe space for docking supply vessels.) The fortresses were officially returned to operation in January 1873, and would remain so for many years.
A new generation of coastal guns would, in 1899, render the artificial fortresses obsolete as gun emplacements, as the new guns could effectively cover the entire width of the harbor mouth. However, the old fortresses remained in use as forward spotting and fire control locations for the coastal guns, with armored long-base rangefinders replacing the old 15" Rodman guns and the old spotting tower's armor modified to allow a smaller rangefinder to be installed in it. In 1947, the Army would install fire control radar systems for its coastal guns, allowing it to declare the old fortresses obsolete and abandon them as military posts, though their lighthouses, marking the channel entrance, would remain manually operated until 1956, when they were automated and the keeper positions abolished.
Today, the fortresses remain in place, somewhat deteriorated but still serving as navigational beacons for the harbor of Taulaga Muamua Sami; efforts are being made by the Army Historical Command to restore one to its 1873 configuration, and the other to its 1945 configuration, so that both could be reopened as museums in the future. Ironically, the Elbonian ironclads the fortresses had been built to counter had been nothing more than rumors and disinformation; instead, the Elbonian Empire collapsed in 1877 and was absorbed into the Philippines.
(Hey, folks. When judging, please note that the decision to include the tug in the drawing was based on discussions in the challenge channel on the Discord server, where the general agreement was that I would have to include whatever method was used to transport the barge to its final destination in the drawing itself. Please don't ding me for that, as it's not meant as an attempt to get around the rules!)