Monday, November 20, 2023

The Practical Geometry of the Parson Barnard House: the Bents

 

This is the second part of my study of the use of geometry in the layout and design of the Parson Barnard House, in North Andover, Massachusetts.

Here's the house. The left section, the 'front', was built in 1715; the saltbox added in 1720. The rear extension to the right came in the 1950's. 

 

For my first post on the Parson Barnard House please see the link at the bottom of this page.*

 


 

 

The first floor plan shows the 2 back wings. The middle section, labeled kitchen, back hall, and study is the 1720 saltbox addition. The laundry and entry spaces were added 230 years later.



 

 

The geometry of the first and second floor plans match. The posts and beams of the second floor match those of the first floor.




The plan shows a step down into the saltbox addition.





 

 

 

The side elevation also shows this step down. The dashed lines locates the floors and the step down.



 

 

The layout Lines locate the inside edge of the posts and the inside width between the posts on either corner of the house. Together they outline 3 sides of a rectangle, the layout for the bents of the frame.  The radii of the arcs is the width of the frame. They cross at the placement for the lower side of the beam which supports the attic floor joists. It is the upper beam of the bents that frame the house.  

This use of geometry matches the way the parlor width was determined. The thin black Line marks the rectangle which is the inside of the bent.

 

 

 

 

 

The rectangle from that layout crossed with its diagonals. Where the Lines cross is the height of the beam which carries the second floor joists. 

*See note below for more information.

For reference here is the geometry of the first floor plan with the Hall on the right, a square; the Parlor on the left, a rectangle whose width is laid out by the arcs. This geometry is laid out from the inside edges of the posts. 

 


 

 

When the same geometry is used to lay out the bent using the width from the outside of one post to the other as the dimension, the crossing of the arcs is at the upper side of the attic beam.

**Look back at the elevation of the house with the rectangle derived from the interior dimensions. Note that the second floor and attic windows - which appear to be original - are centered on the space. Using the outer dimensions the windows are no longer centered.



 

 

 

 

Here is the design for the 6 bents for the Parson Barnard House (without the braces). This simple layout requires only basic knowledge of Practical Geometry. 

The bottom red line is the sill.

 

 

Do the layout differences I've shown here - between the use of the inside and the outside widths - provide evidence that the interior widths were used? Or do they simply mean the drawing is too small, my lines too thick or not perfectly placed, that the attic framing might yield other information for an accurate assessment?

 

 *  https://www.jgrarchitect.com/2023/11/the-practical-geometry-of-parson.html

Tuesday, November 7, 2023

The Practical Geometry of the Parson Barnard House: the Floor Plan


The Parson Barnard House, North Andover, Massachusetts, built in 1715. This picture was taken in 2022.


The original house is the front (left) section: 2 rooms up and down, each with a fireplace. The chimney in the middle served all the fireplaces and acted as a radiator.

The saltbox extension was added c.1720. The rear wing dates to the 1950's.

 

 

 

 

 

John Abbott measured and drew the floor plans and elevations of the house for HABS in 1934. At that time it was thought that Simon Bradstreet had lived here. Now we know the families of the Reverend Thomas Barnard and his son, the Reverend John Barnard, were the first residents from 1715  to 1757.
 

 

Here are the tools a  Massachusetts Bay Colony carpenter had in 1715 for planning and laying out buildings*. He used a compass, a square, horizontal and vertical levels, a straight edge marked in regular increments (which might or might not be inches) and a line with a spool on one end and a plumb bob on the other. The builder also had an awl, chalk, charcoal, and an 8 or 10 ft. rod.

His square was small and might not have a true 90* corner. His inch, foot, rod, varied from those of other carpenters. He would have learned his skills as an apprentice to a master carpenter, become a journeyman, then a full-fledged carpenter. His training would have included practical geometry. Tape measures had not been invented, paper was precious. He drew plans on framing floors, on sheathing, on dirt. He did not need to be literate.

Using a compass, a straight edge and a scribe a carpenter could layout the plan on a board, then step off the plan on the site with his rod or his compass, set his lines, and true them. This is the same order in which we layout buildings today; we simply use more modern tools.


This is the floor plan of the original house.
Noted in black are the sills, the posts and beams for the first floor frame . 

 The carpenter knew the first floor would have a Hall - on the right labeled 'living room', and a Parlor - on the left called the 'dining room'. 

 

Between them, in the center of the house would be the chimney stack with a separate flue for each of 4 fireplaces.

Here is the chimney above the roof with articulated flues.  

These spaces add up to a house about 18 ft. wide by 42 ft. long.

 

 

The width and length of the house were stepped off with a compass or marked on a length of twine. The exterior of the house would have been staked.  3 foot units could have been stepped off 6 times for the width and 7 times for the length. Layout with a rod marked in feet would probably have been faster. A chalk line might have been snapped. Lines could have been tied to stakes for the men digging the foundation. Diagonals would have trued the foundation. All of this is similar to how we layout buildings today.

Stone foundations of  pre-1900 houses tend to be vertical on the inside of the wall, the basement side, battened into the soil on the exterior. 2 lines would be required to accurately set the top of the foundation - where the sill would sit, where the outside wall of the house would stand.   


The length of the beams of the frame, approximately 18 ft., probably determined the house width.  

The Hall, the biggest room, was a multi- purpose room, used for cooking, chores, gathering. Often set in the southeast corner of the house, it had sunshine from early morning to late afternoon.  Here the room is square with a beam to support the 2nd floor across its center. The arc of the 18' width locates the posts.

Note that the dimensions appear to begin at the inner side of the right hand posts, indicating that while the exterior of the foundation was laid 18 ft. x 42 ft., the layout for the timber frame appears to have been set from the sill and those first posts. 

Laying out the geometry from the inside of the frame would have been a practical choice. The sills would not cover the line, but located beside it. Truing a rectangle by checking its diagonals to the outside corner of a post would be tricky, especially after the posts were in place. The framing timbers also needed to set to the line. And, the outside of a stone foundation could be irregular - as field stones are - without compromising the bearing of the frame on the foundation. 

This length, this inside dimension, would also have been the one the framers used to lay out the mortise and tenon joints on the beams.

Next to the Hall is the chimney stack. In plan it is a 3/4/5 rectangle often used by masons to keep the bricks plumb and level during construction. The fireplaces fit within, their fires creating a massive heat sink.  

The entry and staircase fit into the leftover space in front of the stack.

 

 


The Parlor, the room to the left, was used for business and formal occasions, to welcome and entertain visitors. Sometimes it was also the master bedroom.

It was smaller than the Hall and also had a centered beam.  The arcs cross at the outside edge of the wall, setting the width of the room. 

The length of that width could easily be measured from the layout of the Hall. Here the arcs cross, giving 2 points for drawing a line which, extended, is the width of the Parlor. Note the black line with arrows.


Yes, if the line which measured the width of the square of the Hall  had been folded in half and marked, it also would have given the point needed to determine the width of the parlor. In either instance the framer needed to understand the geometry. Geometry was a tool. It was practical. It is also why the proportions of these buildings are graceful; why they speak to us.


 

 

The Parson Barnard House, seem from its front garden in 2022.

 Note that the geometry of the house is so strong that the front door of the house seems to be in the middle of the facade. Actually the right side is wider than the left side. The windows on the left  are closer together than those on the right and the wall spaces between the door and the windows on either side are not equal. 

The geometry of the frame and the elevations will be another post.  


* the image is the frontispiece  for  Giancomo Barozzi Da Vignola, *Canon of the Five Orders  of Architecture, translated by John Leeke, published by William Sherwin, 1669. 


** A square can be laid out by a compass. Square corners can be determined and proved by a daisy wheel. Here is the visual explanation: the width A-B as the radius of the circle and locates the 6 points of the circumference A, B, C, D, E, F, G. Then: Lines A-F and B-E are perpendicular to A-B. Line G-C locates the end (west) wall  of the Parlor.



For more information and a tutorial see: https://www.jgrarchitect.com/2023/01/geometry-in-construction-practical.html