Program in Maritime Studies: Fall Research Semester 2000

Bermuda 2000 - Iron Knee Wreck Site

This fall’s fieldwork began with all participants making an initial reconnaissance dive to acquaint themselves with the wreck and diving conditions, The latter proved most favorable for accurate mapping and complete photographic documentation. Visibility throughout the project was virtually limitless, depth was only 15 feet, water temperature remained a comfortable 78 degrees, and save for a day or two of poor weather, work progressed without interruption. Indeed, by the second dive students found themselves removing loose ballast stone and chiseling apart concreted stone.

Within a couple of days all of the extant keelson, (approximately 15 meters long and running from the apron to the pump well) had been uncovered. The limber boards and several adjacent ceiling planks were also uncovered. Much of this was accomplished simply by muscling ballast stone out from the centerline of the wreck and then following up with a water induction dredge to clear away sand and sediment.

Keelson cleared of ballast, the mapping begins

As mapping was to be done with a grid system, a method new to most participants, students first spent a day assembling the six interlocking 2 x 2 meter aluminum grids on shore before attempting the task underwater. Two-meter long tapes were secured to each Y-axis of the individual grids and a third two meter tape was secured to a sliding bar that represented the X-axis. With the parts for the six interlocking grids labeled and bundled, participants headed back to the site and reassembled the apparatus over the wreck. Each of the individual grids were attached end to end, resulting in a 12 x 2 meter grid that ran from the pump well forward and covered the entire exposed area along the keelson. Using the graduated, two-meter long sliding bar and a plumb bob, each student plotted the features within their assigned grid and recorded these points on mylar graph paper. By simply connecting all of these plotted points an accurate scaled rendition of the wreck was easily produced. After a short practice period, students quickly became proficient with this new method of mapping and several significant design and construction features promptly came to light.

Just forward of the pump well three drift pins protruded from the keelson, suggesting that a mast (in this case probably the mainmast, due to the location of the pump box immediately aft) was stepped into a saddle rather than a mast step cut into the keelson. Several notches for deck stanchions were also evident along the top of the keelson, as were two 2 meter long rectangular pieces fastened side by side and to the top of the keelson. Interestingly, these two pieces, located approximately four meters forward of the pump well, covered a deck stanchion notch and perhaps prevented the keelson from being marred by ballast stone.

Stephen Hammack records the ceiling planking from the site

A number of “scab” pieces (boards hewn only on one side) were fastened to some of the ceiling planks, perhaps also to protect the ceiling from the ballast stone. An interesting 5 x 30 centimeter “patch”, laid into a ceiling plank approximately 1½ meters forward of the pump well on the starboard side, was also recorded. Finally, the stout framing pattern and lack of space between frames suggests 18^th century English construction.

Curiously, not a single material culture item was found on the site, significantly complicating the process of dating the wreck. Two more cast iron knees were recovered, however, as these currently represent the wreck’s most notable construction feature. Once back at the Museum Conservation lab, the knees were placed in a fresh water tank until conservation began. Follow this link to the iron knee conservation process.

Department of History