Oronoque Saltbox Historic Home Survey
by John Poole. Published 10 June 2013. All photos and drawings by the author, except where otherwise noted.
This report describes an eighteenth century saltbox home of Stratford, Connecticut, which I’d surveyed throughout October 16-22, 2012. A large and overly built home, with pronounced Georgian features, it was constructed in 1772, by Reverend Nathan Birdseye. The home possessed a number of interesting nuances of colonial domestic architecture and framing, including: 1) a main cornice defined by a cantilevered front plate; 2) a style of tying joint previously found in Connecticut’s lower Naugatuck Valley; 3) a wide lean-to roof supported by an elevated rear plate; 4) precisely one joist system comprising a summer beam; and 5) a single, central tying joist, as the basis for each of the other joist systems. Those features are examined here in considerable detail. Unfortunately, the home is no longer available for further study, as it was torn down by commercial developers during the week of October 26th-31st, 2012.
- General Characteristics
- Significant Architectural Features
- Timber Frame Overview
- Cantilevered Front Plate
- T-Style Tying Joint
- Elevated Rear Plate
- Summer Beam
- Center Tying Joist
- Summary, Interpretations, and Conclusions
- Related Pages
The Oronoque saltbox was constructed in 1772, by Reverend Nathan Birdseye, ostensibly for his son, Captain Joseph Birdseye. It was located on a property owned by Rev. Birdseye called Prayer Spring Farm, and the local history of this home is largely represented in the article referenced by the preceding link. Tax records for the home, which had been located at 7296 Main Street, state that it was built in 1775.
The Oronoque home was more substantially built than typical earlier saltboxes. It had also been infused with a number of mid-to-late Georgian period features, including widely spaced (non-paired) front fenestration (note the traces of missing shutters in the photograph of Fig. 1), a heavily molded front cornice with returns, molded rakes, and a pilastered front entry with a denticulated pediment, more generally representative of the period 1730-1750. The exterior window trim was kept relatively plain, most likely with the intent of shifting elaboration to the shutters themselves.
The Oronoque home followed the general plan of a typical New England saltbox, of course, but what stood out most about this home were its large size, wider-than-usual lean-to, and the various Georgian details described previously.
I’d also discovered that the present day lean-to wasn’t in its original position, but had been raised and extended farther back, at some point in the home’s more recent history. This was evidenced by the presence of sawn, dimensional rafters which had now supported the lean-to roof, elevating it above the hewn common rafters. Nonetheless, the original lean-to had still been quite wide, and it’s this earlier roof system that’s modeled in this survey.
Raising and extending the lean-to in this manner enabled the rear kitchen wall to be pushed back about eight feet, and also provided enough additional garret space to allow a longitudinal partition wall to be erected on the second floor, defining an end-to-end hallway separating the front sleeping chambers from the garrets.
A small, single story ell had also been attached to the rear of the expanded kitchen, and most likely served as a summer kitchen. This made cooking in the large kitchen fireplace during the summer months unnecessary, thereby keeping the main house cooler.
An unexpected characteristic of the Oronoque saltbox was that its timber frame was completely exposed to the interior, even though it’d been shaped to accommodate plaster. Traces of lath plaster were, in fact, plainly visible on the clear timbers of the second floor chambers. But most of this earlier treatment had been removed in more recent times. It’s also possible much of it had been destroyed during a house fire that occurred in the 1940s.
The home also suffered from interior damage of a more recent nature: some due to neglect, and some the result of numerous DIY projects that were never completed. The previous owners also left much trash and personal effects behind. In fact, I’d found it hard to believe that someone had been living here until just a few months prior to this survey.
Yet, despite all this, the house frame itself was still in remarkably good shape. It was mostly plumb and true, and its joinery still tight, with no significant withdrawal to be seen anywhere. The Oronoque saltbox had never been moved, and was situated on a foundation that apparently had remained quite stable, with no obvious bows or sags.
Significant Architectural Features
The Oronoque saltbox possessed a number of historically significant architectural features, the most prominent of which were the following:
- A main cornice defined by a cantilevered front plate. The cantilevers were shaped from the tie-beam ends, which extended beyond the posts and projected the plate outward.
- A unique style of tying joint, previously discovered in several eighteenth century homes of Connecticut’s lower Naugatuck Valley. Its simple geometry made it ideal for tying a post to a cantilevering tie-beam, the only context in which I’ve encountered this joint, so far.
- A wide lean-to roof supported by an elevated rear plate. The plate was secured to tall post tops that raised it 18″ above the tie-beams.
- A single joist system comprising a traditional summer beam.
- A single, wide, central tying joist as the basis for each of the remaining joist systems.
All of these architectural nuances were realized, of course, by very specific design and joinery details of the Oronoque saltbox’s timber frame, which is the main focus of this report. The next section gives an overview of the frame.
Timber Frame Overview
Fig. 7 is a perspective view of the Oronoque saltbox’s timber frame. The frame had been cut from white oak, scribe fitted, and sheathed plank-on-frame with vertical oak boards, sawn 5/4″ x 12″-15″.
The frame comprised four bents, which I’ve labeled I-IV, beginning at the south end of the house. Examples of the five major architectural features previously described are pointed out in this diagram.
The plan of the Oronoque saltbox frame is shown in Fig. 8. Note that this is not a floor plan in the traditional sense, but rather a projection of the timber frame as viewed from directly above, and with rafters and joists removed.
The plan emphasizes the footprint and geographic orientation of the frame, the reference faces of the bents, bay spacing and depth, and the interpreted use of each bay and areas of the lean-to.
Regarding bay use, I’ve concluded the south bay had served as the hall, given its location at the warmer (south) end of the house, as well as the presence of a partial cellar beneath it. The north bay, being at the cooler end of the house, had accommodated the parlor. This is also confirmed by the fact that the front second floor stair’s initial ascent was situated adjacent to the north bay. These were typical patterns of early Connecticut house plans.
The lean-to included the kitchen, of course, but possibly had been partitioned further into both a pantry and bedroom, as was typical in most saltboxes. Had this been case, the bedroom most likely would’ve been at the warmer (south) end of the lean-to. However, I’d found no physical evidence in the present day home for the existence of either room.
The front elevation of the Oronoque saltbox frame is shown in Fig. 9.
Key measurements provided here include the height above ground sill of the rear plate, attic floor plane (that is, the upper faces of the tie-beams, taken collectively), and second floor plane (likewise equivalent to the upper faces of the wall girts).
Note that the height of the top of the front plate above sill was actually 1/2″ lower than the attic floor plane, or 184″.
This diagram also gives the on-center spacing of the common rafters, which was 45″, except for the rafter pairs flanking the chimney. The Oronoque saltbox had no “chimney rafters” butted directly up against the chimney stack, as often are found in early homes.
The rear elevation of the Oronoque saltbox frame is shown in Fig. 10. Most significant here are the framing members of the rear lean-to area, which are generally aligned with the bent system of the main house.
The upper plane of the garret joist system was about 6-1/2″ lower than that of the second floor joist system of the main house (see Fig. 11). The heights above sill of both the top of the lean-to plate, and the garret joist tops, respectively, are also shown.
Fig. 11 provides a sectional representation of the Oronoque saltbox frame, as viewed from the north end (Bent IV). This diagram clearly illustrates the relative positions of the front, rear, and lean-to plates, and how they’d supported the common rafters, thereby defining the profile of the roof.
In particular, it’s obvious from this diagram how the elevated rear plate enhanced the width of the lean-to roof, and how that, in turn, provided the first floor lean-to area with a depth comparable to that of the front rooms of the main house; that is, 12′, versus 14′-6″, respectively.
Finally, Fig. 12 gives a sectional view of Bent III, from the perspective of the north end. What’s interesting about Bents III and IV is that both embody elements of each of the major architectural features this survey is focused on.
Of particular note in Fig. 12 is the contrast between the summer beam above, and the wide tying joist almost directly below it (also see Fig. 7). These central tying joists were reduced at both ends, and framed into adjacent girts using half-dovetails, thus preventing the girts from drifting apart. This allowed the surrounding joists, which were of varying widths, to be dropped into simpler, butt cog pockets.
The next five sections delve further into the timber framing and joinery details of each of the five significant architectural features of the Oronoque saltbox frame, respectively, that are being examined in this report.
Cantilevered Front Plate
Fig. 13 shows the front plate at Bent II. The cantilever, cut from the lower end of the tie-beam, supported the plate while projecting it outward. The plate seated the rafters, and also provided an infrastructure for the main cornice, which was boxed over the plate and cantilever. The plate’s interior face also served as a fastening surface for the vertical plank sheathing of the front wall.
In Fig. 14, the front plate has been pulled away from the tie-beam, revealing details of the cantilever and tenon. The tenon is modeled as 4″ long, with an 1/8″ chamfer, and cut 2″-to-2″, relative to the upper face of the tie-beam. A blind mortise had been cut into the plate to receive the tenon, and the joint was tightened with two 1″ diameter pegs.
An oddity here was the 5″ x 5″ front joist, which ran alongside the plate with little more than 5/4″ of clearance to accommodate the sheathing planks. Probing its joist pocket revealed that its end had been lapped, rather than butt cogged, most likely to avoid removing too much wood from the tie-beam end.
Another oddity was the height of the front plate itself. Why 5-1/2″, when a 6″ high plate would’ve been flush with the tops of the tie-beams? One possibility is this was simply a construction error, and the front joist might’ve been added to provide an even bearing surface for the attic floor board ends.
Fig. 15 is a photo of this area, taken from the interior of the hall chamber. The front joist appears in the upper right corner of the photo. The plate, of course, is on the other side of the wall, and not visible.
Opening a portion of the front wall provided me access to the front plate and joinery just beyond the vertical planks, so I could determine their shapes and dimensions (albeit, not without some effort). This is shown in Fig. 16.
Interestingly enough, during the final phase of the Oronoque saltbox’s destruction, the very same joinery I’d been investigating earlier became almost perfectly exposed, and just before the demo crew quit for the day.
This enabled me to get photos like the one shown in Fig. 17. Here, the front plate, cantilever, and tenon are clearly visible, as well as an object that looks like one of the two pegs securing the plate to the tie-beam. The plate appeared to have split apart right at its mortise.
T-Style Tying Joint
Fig. 18 illustrates the tying joint used at the tie-beam/post/front plate nexus of Bent II. Note this is a view of the underside of the drawing of Figs. 13 and 14, and a model of the joinery partly captured by the photographs of Figs. 15-17.
I’ve termed this particular joint a “T-Style Tying Joint”, because it comprises a large teasel tenon spanning the depth of the post top (Fig. 19). I’ve encountered it in several 18th century homes in the lower Naugatuck Valley of Connecticut, and believe it might share a common ancestry with a similar joint of colonial Dutch origin found in the mid-Hudson Valley region of New York (Fig. 37). Several variations of this configuration are also recorded in the historic timber framing literature, but with no specific information on their provenances.
Tying joints have been characterized as being cut either at, or below, the plate [Sobon 2002]. When cut below the plate, the joinery is usually quite simple; the through tenon with diminished haunch (Fig. 12) is a good example of this. However, when the plate is also secured at the tie-beam/post connection, the joinery is usually more complicated. For example, the classic English tying joint of many 17th century New England frames is a complex shape that joins tie-beam, post, and plate together at the post top.
But the T-style tying joint is simpler and more flexible in design. It separates the tie-beam/post connection from the tie-beam/plate connection, while still allowing post and plate to be arbitrarily positioned, relative to one another. For example, looking at Fig. 18, it’s easy to imagine the front plate being pushed backwards, until its exterior face is flush with the exterior face of the front post. Then, the cantilever would be eliminated, since it’d no longer be necessary. Such a configuration could be used, for example, in a building whose eaves were flush with its exterior walls, such as the barn described in Fig. 37. In this case, the same basic joinery design would’ve been used to realize two very different architectural styles.
To see how these variations of the same basic design are possible, consider Fig. 19, which shows the front post separated from the tie-beam. The post top has a single teasel tenon as wide as the post itself, and, therefore, reasonably resistant to shear in that dimension. But the tenon is relatively short in length (only 4″ long in this particular example), so the blind mortise in the tie-beam can clear the notch for the plate, as long as the tie-beam is of sufficient depth. This is generally the case, whether the tie-beam cantilevers, or not. And since the dominant forces at the tie-beam/post connection are downward compression and rearward shear only, tenon relish is not a primary concern here, so the teasel tenon could be cut relatively short.
Fig. 20 shows the T-style tying joint of Bent I of the Oronoque saltbox. Here, the two pegs are clearly visible, as well as the other end of the front joist, which had likewise been lapped into the tie-beam.
Elevated Rear Plate
Fig. 21 shows the rear plate, in the vicinity of Bents II and III. The plate was secured to the tops of the rear posts via notched shoulders and tenons, and the posts elevated the plate top 18″ above the attic floor plane. The joints shown in Fig. 21 were double-pegged, but those of Bents I and IV singly pegged, so as to allow greater relish at the plate ends.
Fig. 22 provides a view of this assembly from behind. The plate had been of substantial size (6-3/4″ x 7-3/4″), so as to support the weight of the large lean-to roof.
In Fig. 23, the rear plate has been pulled back to show details of the joinery, which consisted of the notched shoulder and a 4″ blind tenon, laid-out 2″-to-2″ to the exterior face of the post. Of particular interest here is the front side of the notch, which extended upward and slightly above the plate, and had been shaped to a graceful taper.
I believe the taper was an attempt to evenly distribute any component forces exerted inwardly by the rear plate, and thereby minimize the possibility of tear-away. The small dashed lines on the side of the taper in Fig. 23 portray the scribe lines used in laying-out its shape, and they’re likewise the sections I’d defined in SketchUp to model the taper itself. These scribe marks were plainly visible on the actual timber.
The photograph of Fig. 24 shows the front face of the rear post of Bent III. When I initially found it, it was surrounded by much construction paper, acoustical paneling, and fiberglass batt insulation.
Fig. 25 shows the same tapered top, as well as the rear plate, after I’d removed much of the material obscuring it. Note the aforementioned scribe lines defining the shape of the taper, as well as its hewn surfaces.
Also visible in Fig. 25 is the rear post’s tenon, just where it’s entering the mortise in the bottom of the plate. The timber pegs had been driven from the opposite side, and their tips were cut flush. The cut tips can be seen in the peg holes, whose centers were 1-1/2″ above the shoulder.
The Oronoque saltbox had but one summer beam, which supported the joist system of the attic floor above the parlor chamber. As in most early Connecticut homes, the summer ran longitudinally, having been framed into the chimney and end girts. The photo in Fig. 26 shows the summer beam and its joists, viewed in the direction of the chimney girt (south).
Fig. 27 shows the summer and joists, as seen facing the opposite direction (north).
The summer beam was 12″ wide, 7″ deep, and 14′ long, and supported six pairs of 3″ x 7″ transverse joists. Its bottom edges were softened by 2″ wide chamfers that ran full length, and each end was framed into its girt using a 4-1/2″ deep, shouldered dovetail. Fig. 28 shows the summer beam in my model of the Oronoque saltbox frame, as viewed from above while facing south.
I had no immediate way of determining if the joists were original, or perhaps framed-in at some later time, but it was obvious that the framing system I’d discovered was shaped to accommodate a plaster ceiling, and indeed had once been concealed by one.
Plainly obvious (Fig. 29) were traces of lath plaster across the joist bottoms, all of which had been hand planed flat, and set nearly flush with the bottom of the summer beam. A plaster line was also visible along the length of the chimney girt, and running just beneath the summer beam.
There were a number of interesting oddities in the detailing of the joists and summer beam, as well. For example, the joists had been cogged into the summer beam, and the rear joists were cut with angled reductions beneath their tenons. But the front joists were notched, and in some cases, where the notch wasn’t flush with the summer, small filler pieces had been shaped and inserted into the notches. This is shown in Fig. 30.
The summer beam also had stops at the ends of its chamfered bottom edges, as well as one positioned under each joist pocket (see Fig. 30). While this might not be an unusual detail of a carved summer beam, what was odd here was that these stops were small isosceles reveals that had been cut separately and then nailed to the summer beam. Fig. 31 shows the underside of the modeled summer beam near the end girt, including the stops and their dimensions.
My conclusion is these “stops” weren’t decorative at all, but possibly had served to level-out the angled chamfer when attaching plaster lath near the bottom of the summer beam.
Center Tying Joist
The remaining major joist systems of the Oronoque saltbox’s main house frame — parlor, hall, and hall chamber — were not based on traditional summer beams, but rather a simpler system of longitudinal joists that were run between the chimney and end girts. These joists were far more substantial than the relatively thin, transverse joists used with the summer beam.
An advantage of this alternative style of joist system is that the dead load of the floor is evenly distributed along the lengths of the girts, the floor’s primary bearers, whereas a longitudinal summer beam concentrates that all that floor weight precisely at the unsupported midpoints of the girts, just where they’re most likely to sag. This alternative strategy also meant less pieces to cut, and less complex joinery, as most of the joists could simply be butt-cogged.
In the actual examples I’ve encountered here in Connecticut, both in the Oronoque saltbox and elsewhere, the central joist of this alternative system is a tying joist, meaning that it holds the opposing girts together, so there’s no chance of them drifting apart over time. This enabled the remaining joists to be framed-in with simple butt cog joints, where a plain joist pocket received a reduced joist end. In all the cases I’ve found, the tying joist was invariably framed using a half-dovetail joint.
Another observation I’ve made is that these central tying joists were invariably wider than their surrounding joists. Furthermore, the joists immediately on either side of the center joist often were likewise of greater width than the remaining joists in the floor system, but still not quite as wide as the center joist (see Fig. 34).
Within the Oronoque saltbox, I’d also found that, even though the joists decreased in width farther away from the center joist, their pockets were sometimes cut to uniform widths (e.g., in the hall chamber), with the wider joist ends reduced to fit them. And in other cases (e.g., parlor and hall), the pockets were scribed to fit the varying widths of the joist ends, instead (see Fig. 35).
Finally, these longitudinal joist systems of the Oronoque saltbox likewise were shaped to accommodate plaster, just as the parlor chamber’s summer beam and transverse joists had been. These joists had been hand planed on their lower faces so as to collectively achieve a level surface. And as shown by the photos, many were hewn from highly irregular timbers, since they were never intended to be exposed in the completed house.
Summary, Interpretations, and Conclusions
The Oronoque saltbox uniquely combined a number of novel architectural features, and this survey report has described them at length, including details of the timber frame construction and joinery realizing them. Here, I’ll compare these against other recorded examples. I’ll also provide some more speculative thoughts on the evolution of regional timber framing practices in this particular area of Connecticut, of which the discovered state of the Oronoque home is perhaps a snapshot.
Cantilevered Front Plate
In Early Domestic Architecture of Connecticut, J. Frederick Kelly recorded a number of early Connecticut homes with main cornices defined by extended framing members, usually the tie-beams (upper girts), or attic floor joists. Here are three particular examples of plates cantilevered on tie-beams:
- Older Williams House, c. 1690 (known today as the Buttolph-Williams House, c. 1711), of Wethersfield, Connecticut — This home has a pronounced main cornice defined by a secondary plate, which is projected outward by the tie-beam ends. The primary plate ties the bents, while the secondary plate seats the rafters. Both secondary plate and cantilevers define and support the cornice box. The cantilevers are lapped over the primary plate. [pp. 33-34, 124, 127]
- Linsley House, c. 1770, in North Branford, Connecticut — This home included a single, large front plate, projected outward by cantilevering tie-beams, the exposed ends of which had been decoratively carved. The bottom of the plate was left exposed, with no other significant elaboration beyond a simple weather board covering the front face of the plate. [pp. 123, 126-127]
- Evarts Tavern, c. 1710, in Northford, Connecticut — This house was very similar in construction to the Oronoque saltbox. It had a single, large front plate (7-1/2″ x 16″) cantilevered on extended tie-beam ends, which in turn were exposed to the outside as decoratively carved brackets. The plate tied the bents, seated the rafters, and supported a built-up cornice box. [pp. 122-123, 126-127]
Where the Oronoque saltbox differed from Evarts Tavern was that its front plate was of relatively smaller dimensions (5-1/2″ x 6″), and both plate and tie-beam cantilevers were completely enclosed by the cornice box.
Elevated Rear Plate
Kelly likewise reported many Connecticut saltboxes where extended framing members supported wide lean-to roofs. In some cases, tie-beams or attic floor joists where cut well beyond the rear plate to directly support the lean-to rafters. In other cases, blocking of some kind, or a secondary rear plate (structural purlin), was elevated above the main rear plate to support the lean-to rafters. Kelly noted two particular examples of elevated plates:
- Cyrus Hawley House, c. 1740, in Monroe, Connecticut – Here, a secondary rear plate had been positioned above the primary rear plate, serving as a purlin for the lean-to rafters. [p. 189]
- Evarts Tavern, c. 1710, in Northford, Connecticut — This home likewise employed a secondary rear plate supporting the lean-to rafters. This plate reportedly was suspended about 18-24″ above the primary rear plate. [p. 52]
The Oronoque saltbox differed from either of these two examples in that it leveraged a single, elevated rear plate to tie the frame, and also support the lean-to roof, rather than relying on two distinct plates to perform these functions. It should be noted that Evarts Tavern was the only 18th century Connecticut home of the many surveyed by Kelly that included both a cantilevered front plate and an elevated rear plate, together, in the same structure, and thereby reinforcing, at least in part, my case for the historical-architectural uniqueness of the Oronoque saltbox.
T-Style Tying Joint
Here are a few examples of joints very similar to the T-style tying joint, mostly coming from the traditional timber framing literature:
- Jack Sobon’s Historic American Timber Joinery: A Graphic Guide (2002-2010), p. 11, Fig. 11, illustrates a plate/post/tie-beam configuration from a c. 1791 house in Cheshire, Massachusetts that’s geometrically very similar to the T-style tying joint, as illustrated in Figs. 13, 14, and 18 of this report, except the roles of tie-beam and plate are effectively switched.
- Historic American Timber Joinery, p. 14, Fig. 22, diagrams a corner post/plate/tie-beam configuration from a c. 1810+ house in Canaan, Connecticut that appears identical to the T-style tying joint. Even the dimensions are nearly the same. Sobon’s drawing also shows vertical sheathing planks hung from the inside face of the plate, as was typical in the plank-on-frame homes of Connecticut.
- Historic American Timber Joinery, p. 15, Figs. 25-27, illustrates a plate/post/tie-beam configuration from a c. 1810+ house in Windsor, Massachusetts that’s essentially a square-rule equivalent of the T-style tying joint, except for the singly pegged post/tie-beam connection, and the fact that the plate was not cantilevered, and had been lapped over the bottom end of the tie-beam.
- The Society for the Preservation of Hudson Valley Vernacular Architecture’s Hudson Valley Vernacular Architecture Newsletter, January, 2003, describes a plate/post/tie-beam connection from a c. 1850 barn in Amsterdam, New York, that’s likewise nearly identical to the T-style tying joint, mainly differing in the fact that its singly pegged, the plate is not cantilevered, and its teasel tenon is of significantly less width (see Fig. 37).
- Finally, Craig Chartier, of the Plymouth Archaeological Rediscovery Project, mentioned to me in a personal communication some time ago that he’d found similar joinery in a barn at the Taylor-Bray Farm in Yarmouth, Massachusetts, and suspected it had also been used in the c. 1640 Richard Taylor House.
What’s significant about the HVVA find (illustrated in Fig. 37) was its geographic location: an area where timber framing practices had descended mainly from Dutch, rather than English, traditions.
Was the T-style tying joint a Dutch design that became infused in English colonial timber framing in Connecticut and Massachusetts? Or was it a simplification independently arrived at by various regional timber framers? Further investigation into its provenance is clearly warranted, but I suspect it’s likely to contain at least some Dutch DNA.
Why did the Oronoque saltbox have just one summer beam?
Part of the answer might be found in Kelly’s account in Early Domestic Architecture of Connecticut of the early diminution, or even outright deprecation, of the summer beam in this particular region of Connecticut [pp. 65-69]. He noted how the early adoption of plaster finishing, especially in the nearby New Haven Colony, inspired local housewrights, from early on, to significantly reduce the depths of summer beams, or even eliminate them altogether.
Kelly also described how priority for plastering was often given to the front rooms of the first floor. Often, these rooms either had sharply reduced summers which could be concealed behind plaster, or none at all, presumably resorting to longitudinal joists. But the second floor rooms were often left unfinished, and summer beams were still found exposed on the second floor. But by about 1750, summers were either concealed, or altogether eliminated, from new homes throughout much of Connecticut.
So, given this regional-historical context, two possible explanations for the Oronoque saltbox’s lone summer beam emerge. Clearly, parlor, hall, and hall chamber all were intended to be plastered. As the hall chamber was often used to accommodate guests, it’s understandable the owners might’ve wanted to finish it, as well. But the parlor chamber might not had been intended to be finished, so the floor above it was framed with a summer beam. Then at some later point, a plaster ceiling was added, and the original transverse joists might’ve been replaced with the deeper ones found in the present day house. The rather odd detailing of the summer beam and joist connections reported here might be attributable to this.
On the other hand, there’s also the possibility that these deep joists, cut flush with the bottom of the summer, were original, and that the summer beam was intended to be concealed from the start. If that’s the case, why use a summer? Why not a system of simpler, longitudinal joists, instead?
My only explanation for this is, given that the summer beam was steadily being de-emphasized in this region of Connecticut, framing a single summer beam like this might’ve represented a trade-off between ingrained tradition and the newer methods. Although the builders were well aware of the advantages of a simpler joist system, their psychology still might’ve required them to hold on, to some degree, to a past framing tradition they were comfortable with. But, of course, this is purely conjecture on my part.
Center Tying Joist
Finally, I have some thoughts about the dimensions I’d found in the systems of longitudinal joists. As mentioned earlier, the center tying joist was invariably the widest, and also was flanked on either side by two or more joists that likewise were wider than the remaining joists, though not as wide as the center joist itself. Why were they cut this way?
Part of this might’ve been influenced by the perceived enhanced tensile strength of a wide center tying joist, which was solely relied upon to tie the opposing girts. However, this makes less sense in light of the reductions found near the ends of each center joist. On the other hand, early timber framers usually relied on increasing the width (rather than the depth) of a spanning floor joist to mitigate its potential to sag. And, at least in this particular situation, increasing the depth of the longitudinal joists would’ve run counter to the reason for eliminating the summer in the first place; that is, accommodating a continuous plaster ceiling.
So, it seems likely that these 18th century Connecticut framers, who were transitioning away from the summer in favor of simpler joist systems, well understood the need to continue to mitigate sag in the very center of the floor, something the summer beam fundamentally achieved (even though, as noted earlier, it still had the potential to cause the grits themselves to sag at their midpoints). Therefore, they cut the center tying joist, as well as its flanking joists, with relatively greater widths.
[Kelly 1924] J. Frederick Kelly, AIA, Early Domestic Architecture of Connecticut, Dover Publications, New York, 1924.
[Poole 2012a] John Poole, A Preservationist’s Technical Notebook: Imminent Historic Teardown, September, 2012.
[Poole 2012b] John Poole, A Preservationist’s Technical Notebook: Oronoque Saltbox Requiem, November, 2012.
[Sobon 2002] Jack Sobon, Historic American Timber Joinery: A Graphic Guide (2002-2010), National Center for Preservation Technology and Training, and The Timber Framers Guild, 2002-2010.
[Sobon 1984] Jack Sobon and Roger Schroeder, Timber Frame Construction: All About Post-And-Beam Building, Storey Publishing, 1984.
Timber Framers Guild, Glossary of Timber Framing Terms (PDF).