Sunday, February 16, 2014

Sandown, New Hampshire, Meeting House, 1773

I substantially revised this post on April 18, 2014.

Timothy Palmer is listed as the master builder of this meeting house. 12 years later he built the Rocky Hill Meeting House in Amesbury, MA., which I wrote about in my last post -

I was curious to find out if the geometric language used to design the Rocky Hill Meeting House could be seen here as well.
I knew of the Sandown Meeting House but had never seen it.
Luckily HABS drawings of the Meeting House are on-line. This photograph is part of that record. I borrowed it through wikimedia. A detailed history with photographs can be found at

I thought I'd just see what Timothy Palmer built, and how he evolved as a master-builder - simple.
Instead I found a larger fascinating history.
 But first: the geometry of the Sandown Meeting House.
 I know this is dense. Start with the drawings. Click on the images to enlarge them. Use my words to help explain what is there, why I added my lines. This would be so much easier to explain in person!  

I start with the simple geometry, the square.
Here is the front elevation - 2 squares side by side. I've crossed the one on the left side with both diagonals, but only drawn one diagonal for the square on the right side. The squares determine the size of the front and the structural dimensions: they include the sill and the top plate

The window placement is also simple:
On the right side I have divided the square - see 'a-a-a-a' - in half vertically and then drawn the diagonals for those rectangles. The intersections are at the edges of the windows - see 'c'.  The windows themselves are 2 squares. See the diagonals in the window upper left corner.
Then I've rotated the square to show how the line determines the location and width of the entrance - see 'b-b-b-b'.

This is quite basic design, balanced, stark - except for the door.

The side elevations have similar window placements, though here the squares cross each other. See 'a-a-a-a' and 'b-b-b-b'. The placement of the windows is noted by the arrows at the bottom. The attic window follows the pattern - its placement and size determined by squares which grow out of the crossed squares below.
I have not yet figured out why the door is off-set. Maybe the space needed for the stairs to the balcony threw things off.
Note how the symmetry of the windows is so strong and the triangular pediment such an 'eye catcher' that one has to pay attention to see that the door really isn't centered! 

Where did Palmer begin his design? The church committee overseeing construction would have told him approximately how big to make the building. How did he determine his layout?

I think he began with the pulpit.
Here, as in the Rocky Hill Meeting House the pulpit with its sounding board is the centerpiece of the church. Not only is it ornate and dramatic, it surrounds, protects and presents the preacher and the Bible. The book would be laid out, open, on the lectern. And the lectern is precisely at the crossing of the squares.   (Here I think I need to add a picture of the pulpit.)

 One square begins on the right of the pulpit and extends to the left; the other begins on the left and extends to the right.

I have blacked in the columns as the design locates them. Here the lower edge of the squares runs through the columns which support the balcony. The center the squares locates the columns in the open space  - both, however, in only one direction.
The left and right sides of the squares are the locations of the posts on the north and south walls and the roof trusses above. See red lines on either side of the main aisle. 

Next I added the diagonals for the 'left over' rectangles created by the crossed squares  - see the dotted red lines. Where the new diagonals cross the diagonals of the squares is the location for the second set of posts in the exterior walls and the post holding up the balcony.

The length of the diagonal of the square is also the width of the meeting house itself.  I have used that diagonal to draw a square inside the frame of the meeting house. It extends outside the building at the entrance and the pulpit. Where it crossed the outside walls is the line which determines the posts on each side of the central aisle. See the green square set as a diamond and its left and right green triangles.
The final columns are those on either side of the east and west entrances. The triangles noted in green also mark off the 2 squares on either side of the central aisle. When those squares are subdivided, the subsequent intersections determine the columns - see the small squares and their diagonals on either side of the left (west) door.

 The West Elevation - Interior shows Palmer again using the crossed squares. The diagonals are also the position of mortises in the roof trusses. The extension of either interior diagonal  lands on the top of the brace.

The roof pitch is an 8/12. I have drawn it as an array of squares. I think the trusses may also be determined by the same geometry, but so far I've not worked it out.

Many more parts of the building are determined by this geometry: the other braces, the balcony, the railings, the pulpit. If I draw them all the drawings becomes a mess - illegible.

Instead I have posted a diagram for the entrance.

The diagonals with solid lines, the perpendicular and horizontal divisions in dashed lines: 8 equal sections, 8 spokes.

Timothy Palmer did not use circles with 6 points (the daisy wheel). He did use the square 'flat' and 'rotated', and its 8 points to design the entrance. I just needed to determine his basic length: the radius of the circle.

The first dimension I saw was the distance between the center of one pilaster and the other - the horizontal line in the middle of the circle. I deliberately white washed the extensions of that line to show the circle divided into 8 segments and the distance between the pilasters.
Then I realized that that length is also the height of the columns from the pedestal to the capital. Hmmm....
That length is also the radius of the circle that encloses the entry from the peak of the pediment to the bottom of the door.
The width of the pediment is determined by the circle's upper spokes. The height of the pedestals is set by the lower spokes.
The proportion of the door itself is the Golden Section.

The use of geometry here in the entrances is much more sophisticated than that used in the facades of the meeting house. I thought the door and its surround might use the Golden Section.  A circle fits much better. For a more complete discussion of the entrance geometry see my post on the Rockingham, VT, meeting house.

Perhaps Timothy Palmer's own history accounts for this.
He was apprenticed to Daniel Spofford, an architect and millwright in Newburyport, MA. Some records say he also learned shipbuilding  from Ambrose Spofford. He seems to have been responsible, at age 23, for building this meeting house. Perhaps he also built the pulpit.
His other responsibility was to span the open interior space of the church. He would have know what was necessary to frame a boat hull - which is in some ways an upside down roof.

At the Rocky Hill Meetinghouse he used other geometry: a circle with 6 spokes or petals: a daisy wheel, a much more complex double square, and the 3-4-5 triangle.
In 1792 he quit building churches and became a bridgewright. He built many bridges including 4 across the Merrimack River, and the Market Street Bridge in Philadelphia. For more information see Frank Griggs' excellent well illustrated article:

Tuesday, February 4, 2014

Rocky Hill Meeting House, Amesbury, MA. - 1785

The Rocky Hill Meeting House was built in 1785 by the people of the West Parish of Salisbury. It was placed on ledge at the top of Rocky Hill on the road to Portsmouth, NH. By 1886 it  belonged to the West Parish Society who, in 1942, gave it to SPNEA, now known as Historic New England. The church is now open occasionally for gatherings and concerts.
Here is the link for the meeting house:

In the summer of 2013, Historic New England offered a tour of the building led by their staff. A happy group assembled, delighted to look carefully at the building with knowledgeable guides. We were able to sit in the pews, speak from the pulpit, explore the balcony and climb the ladder into the attic to see and touch the trusses that clear span the sanctuary: 48'-9" outside to outside. We visitors came with many specific interests which led to excellent wide ranging discussions.

The drawings were available through HABS, so I, of course, began to consider how the meeting house had been designed.

Here is how I think the building was planned, the dimensions and spacing determined.

The committee knew about how big the building would be, that it would have a balcony and be set on ledge with no crawl space. The first dimensions would have been two  3-4-5 triangles, shown here, with the long sides ('a' and 'b') the location of the interior columns to support the balcony.

For simplicity I have shown the balcony plan here. Remember that all diagrams can be enlarged by clicking on them.

2 squares were laid out, based on the sides.  Here I have labeled the diagonals 'a' and 'b'. and drawn one square with solid lines, the other with dashed. 

These overlapping squares determined the floor plan.

I have drawn the multitude of diagonals that grow from those 2 interlocking squares. The side doors were centered, easily. The pulpit itself, is centered on the space, with the lectern - where the Bible sat - at the center of its square.
Most of the columns fall on the crossing of diagonals. I think where some do not it is because of logistics.
The columns which should be directly in line with the side doors would have made an awkward view and passage for someone entering there. The aisle would have needed to detour around the column - visually and physically unappealing.

Similarly the columns near the back of the church, if placed on the crossings would have created box pews almost too small to be useful. This I know from personal experience, having sat in similar box pews.

The entry 'porch' is also almost a square, a continuation of the pattern. It is about a foot too long. Again, I think the stair- which is quite steep with very tight landings - did not fit so the porch was extended.

  The section is the daisy wheel - circle geometry. Here the 4 side petals determine bottom side of the bottom plate, the top side of top plate (bottom chord of the roof truss), and the bottom of the ridge. The roof trusses were built to sit above the daisy wheel.
The roof pitch is determined by the daisy wheel.

I have noted the top and bottom plates with a  star: *. (No, it's not an asterisk - no footnotes!) 

 The attic trusses are divided in thirds. The center of the determining circle is starred (*). It is the midpoint of the length of the daisy petal. It is also the top of the bottom chord of the roof truss. The radius of the circle I have marked in a thick red line. I did not draw the whole circle - you can project it. 
Note that the central tie beam falls on the radius of the circle.

The front elevation follows the Golden Section layout that seems to have been popular at the time which I drawn in several previous posts.

Here drawn in red, are 2 squares with their diagonals (The Golden Section) crossing in the center.

It could also be 2 squares whose surrounding circles meet in the center, but it's not quite as neat.

Note how the eaves of the porch bump into the windows. The porch is squeezed in... Maybe it was something the house-wright didn't think about very much until he had to build it. I think he is saved by the strength of the proportions - the eye doesn't catch it. I looked twice when I stood at the window in the balcony and saw the eave right there in front of the sash. If nothing else it is an invitation to rot.

I was curious to see if the window placement might follow the layout of the overlapping squares used in the floor plan. At least on this elevation it seems as if it does. I have only drawn the left side. The center of the square on the left of the entry determines the left side of the center window. The square with its right side in the center of the porch  - and center of the meeting house facade - lands at the left side of the farthest window.
The right side matches.

The master builders for the Rocky Hill Meeting House were Timothy Palmer and Ambrose Spofford. Palmer is also listed as the master builder for the Sandown, NH, Meeting House, built  in 1773. He is thought to have built that pulpit which is quite similar to the one here.
I am curious to see how that meeting house frame compares to this.