Plans for Falling Creek
What we thought we knew
Conventional wisdom had the Falling Creek Ironworks on the south or right bank of Falling Creek and Archibald Cary’s Chesterfield Forge on the north or left bank of Falling Creek. This was uniformly accepted from the 1885 Brock investigation to the 1994 William & Mary excavation. However, later investigations showed that interaction between the two banks had occurred and that what are now known to be erroneous assumptions were made.
The dam across the channel that today hugs the north side of the floodplain was a ripple except at extreme low tides when the character was more evident. It contained skulls and mossers from Cary’s forge operation. The floodplain was also littered with forge slag. Therefore, if the north bank belongs only to Cary hypothesis was true, there should have been no Cary material on the south bank. The biggest problem with this is that it flies in the face of standard practice. Cary owned both banks for a considerable distance on either side and there was no sound reason for him to have avoided the south side of the creek. In light of the very large pile of slag reportedly removed by the Highway Department in the 1930’s for road fill being on the south bank on the lower floodplain, the left bank supposition was built on no real evidence. Cleanouts of a culvert under Marina Drive just below the promontory were about half full of forge slag. The excavations of Hatch and Gregory as well as MacCord all had copious amounts of forge slag in them.
Cary operated his forge which produced copious amounts of slag as an industrial waste product. At the end of each day, each forge hearth was allowed to cool. In the next morning, the cooled slag lumps called skulls or mossers (so named for their resemblance to the human skull) were gathered and dumped off-site. This material was dumped on the floodplain to build up the land from floods and to provide a firm base for the buildings that Benedict Arnold burned in 1781.
Hatch and Gregory, MacCord and the W&M investigations all mentioned a charcoal pile. Accepted wisdom was that it was the charcoal pile for the blast furnace. A radiocarbon date of 1570 was obtained which seemed to validate the proposition.
There is but one large problem with that hypothesis. It is in the wrong location for a blast furnace. Blast furnaces are fed (charged) from the top via a charging bridge. Where the charcoal patch was located was on the floodplain and was at the toe of the slope. That is decidedly the wrong place to put charcoal which will absorb moister and become useless.
As the furnace had to be fed from the top, either the promontory was the “top” or the higher ground where the apartments are located was the “top”. Bensley in his construction has altered the back or south end of the promontory such that the original contours cannot be determined. If the original slope was steep, it is more likely that the higher elevation by the apartments was used. But, as the height of the stack is not known, and 17th century furnaces were not as high as later ones (up to the 25’ limit), it is possible that the promontory was where the charging bridge was located. Another factor is that the elevation of the base of the furnace over sea level is not now known. Given eustatic sea level rises of 4 feet, the possible seat might have been that much lower. The promontory might have been quite high enough.
The charcoal pile was in the correct location to feed a set of forge hearths that Cary had set up. But, that would go against conventional wisdom and was not considered likely until Hurricane Fran, followed by Hurricane Gaston, cut an “L” shaped channel out of the floodplain and removed all of the alluvial sands on the upper floodplain.
The Falling Creek Ironworks Foundation engaged Browning & Associates, Ltd. to do a set of strata cuts in the bank of the cut to determine the sequence. It was then noted that a thick layer of clay had been deposited down into the cut and that it contained 20th century debris, notably 3 hole molded bricks.
At first glance, the channel opened by Fran had the appearance of a major flood event by cutting a channel in the floodplain, but nothing indicated it was anything other than a single event. When Gaston came through, water was high enough to cover the promontory, and in fact nearly demolished the Falling Creek bridge upstream. It also peeled back the banks, removed Bensley’s retaining walls, removed the floodplain sands and widened the cut. The cut banks then revealed several tree stumps that were apparently in situ but at the bottom of the cut at the water level. These stumps all angled out into the water entirely consistent with trees that had once grown along the banks of the stream.
Subsequent to that, two storms in successive weeks in early 2007 dropped 4 inches of rain each into the Falling Creek drainage and the resultant floodwaters again altered the landscape. Their chief result was to widen and deepen the cut. The result was that Ralph Lovern noted that large timbers were protruding from the south bank of the new cut on the south bank of Falling Creek.
Enough material had been removed by then to provide information to interpret the sequence of events with some accuracy. This also necessitated completely revisiting and discarding the binary idea of settlement between the two separate sequences of iron operations there. It was abundantly clear that Cary had operated on the south bank and if subsequent excavations are done, will certainly prove that his forge operated there. The thick clay layer was shown to be deposited by Bensley in order to infill the cut channel and tailrace channel for Cary’s forge operation. As conventional wisdom has also held that the stone mill remnant on the north bank was William Byrd’s gristmill, the seating of the forge on the opposite side of the creek would make more sense from a physical spacing viewpoint. Also, low tides have shown that a wall lining the north bank of the north cut of Falling Creek east of the mill is in fact the edging for a roadway that is underpinned by planks and which is oriented to the dam that is largely composed of skulls and mossers. This road and dam would offer access to the floodplain from the north bank of the creek. We presume that there was a bridge crossing the cut on the south bank to allow crossing of that water body.
So rather than having one stream course hugging the north bank of Falling Creek, it appears that at for part of the time, there were two watercourses in operation and that Cary’s Chesterfield Forge was in operation on the south bank while Byrd’s Mill (later Cary’s Mill) was in operation on the north bank.
What's Planned for today
The magnetometer survey (Jones & Maki 1999) showed a large magnetic anomaly consistent with a blast furnace under the road to the marina downstream. Once the road was moved, the plan was for the roadfill to be removed in order to see what physical traces of the blast furnace site could be identified.
A trackhoe was brought in and the road fill was removed down to a charcoal layer. As charcoal was a component of the blast furnace, machine work was stopped. Stratigraphic archaeological excavation of two trenches showed that the charcoal contained barbed wire (post-1874) and clear glass (20th century) and was therefore part of Bensley’s road fill. The charcoal itself had no “structural integrity” and was in fact smashed into microscopic pieces by Bensley’s work. The charcoal layer in turn overlay another fill layer.
A mini-excavator was then brought in to remove the fill and yet another charcoal layer was discerned below the gravelly fill. Excavation was again done by natural layers and yet more smashed charcoal and barbed wire was discovered.
The mini-excavator was again brought in and more material was removed down to what appears to be undisturbed archaeological layers.
We are now at the stage where the next step will be to go into the cribwork that Archibald Cary built for his forge from 1750-1781. To do that, we are going to have to remove the crib infill and expose the timbers. The timbers are massive, some as large as 36x44x120” while others are just big and long (18”x33’). To let these pieces of history just decay to nothing is too terrible to contemplate. We will need to get them up after they’re recorded, then preserve them in chemicals, store them until we can get a museum built and finally put them into a display of very large and early timbers. To preserve the timbers, we need to have a storage space, to build a tank long enough and deep enough to put the timbers in it and to purchase the chemicals to preserve the timbers. All that takes time and it is expensive.
Once that part is done, we want to see how Cary renovated the 17th century Virginia Company cribwork as the magnetic anomaly the survey showed hasn’t been found yet and is presumably down under all the Cary built cribwork.
The short terms goals are to finance the excavation and preservation of the Cary Forge timbers and to investigate the 17th century ironworks remnants below them.