Great Squares

rblood38's picture

I'm in the midst of what seems like a never ending research project on some of the great Georgian squares in London. I am trying hard to understand as much as I can about the design of squares, and the Georgian squares of London are more of a vehicle as opposed to an endorsement. I am working on several actual projects currently that include “squares” and so my interest is particularly keen with regard to dimensions that work and those that do not. But the other issue I am exploring is the connectedness of squares in predominately mixed use districts, and that too makes the London squares in Belgravia, etc. potentially helpful since you can “kick the tire” on twenty differently shaped squares in roughy a half mile radius. Plus, the average walking distance between the London squares surveyed is only 586 feet - I think a very remarkable frequency for an area very much developed by thoughtful speculators. Noteworthy.

My primary focus is to create a little notebook that would inventory and comment upon the dimensions and arrangements of the squares that are within a walk distance of Harrods – a notebook that a member of CNU could take with them on a walk through the series of small and large squares. there are over 20 squares in a very short walking distance and I am slowly analyzing each one, determining the developmental history, closure raios, etc. But Chuck Bohl thought the study could be much more useful if it also included some statistical analysis (closure ratio, width, breath and the nature of the architectural treatment of the enclosed space) of a few other squares from other places picked by trustworthy designers from around the world.

My hope has been to include two specific choices from Andreas Duany, Lizz Plater-Zyberk, Leon Krier, Victor Dover, Robert Davis, Jacqueline Robertson, Stephanos Polyzoides and Joe Molinaro - one recent and one antique. I have in hand special places chosen by both Lizz, Andreas, Victor and Stephanos, and Joe - half of which are in the States and half in Europe. Leon Krier has sent some preliminary thoughts, and I hope for his final picks soon. To get everyone to actually do this, i.e. make a pick, Lizz suggested a slight twist: Don’t place so much emphasis on the very best and beloved square of all time done or experienced by the designer, but the square that comes to mind today when the question is asked. This means that one might have a different pick on another day. That attitude relaxes the tension one might feel when asked to identify a favorite; it gives one permission to be a little lighter with the answer. Otherwise, I probably wouldn’t have gotten many picks!

Lizz picked the DPZ designed very small public space that was near where the CNU soiree was held in Providence. She couldn’t remember the name of the little space on the corner of two streets. I am researching that now. For an “antique” square (older than a century), she picked Place Furstenberg in Paris. My friend Lewis Nix has measured it for me on a recent trip.

Andreas and Robert Davis picked Ruskin Place in Seaside for their DPZ favorite, and Andreas picked Place des Vosges in Paris for his “antique” favorite. The latter is helpful to me partially because it violates some of Andreas’ rules-of-thumb (the closure ratio is very low), and thus is interesting to include. I imagine most CNU members have visited Ruskin Place. It is quite complex, and I am looking forward to completing a section that focuses on Ruskin.

Stephanos picked their new plaza at Aldea outside of Santa Fe, and plus two interconnected squares in Cortona and one in San Giminiano.

Victor Dover picked a Dover-Kohl planned (together with many hands, include Andres’ and Xavier Iglesias’ and Vince Graham’s) an elongated square in a quiet part of I’On that is intriguing. It is called Perseverance Square. His antique picks were an incrediably beautiful square in Bruges, Belgium plus a square in Damme. I still need dimensions on all three of these. I have good Google earth data on I'On but so far nothing measurable for the Belgium squares. The fine photo of the square aside the Beguine Convent of the Vine (Begijnhof)was taken by Brian james McMorro.

With regard to the London Squares. the following are included in the effort: Wilton Crescent, Belgrave Square, Eaton, Chesham, Cadogan Gardens, Lowndes, Hans Place, Cadogan Square, Lennox Gardens, Ovington, Beaufort Gardens, Brompton, Egerton Place, Egerton Crescent, Thurloe, Pelham Crescent, Brompton, Ennismore, Rutland Gate, Trevor, and Montpelier.

I have been measuring the ones I could get to using multiple techniques. Rick Hall was kind to lend me his "Road Runner" to measure Ruskin Place. I have used a laser on some and a tape on others. Some I have just "google earthed" which has actually proved fairly reliable. It is the vertical height that is often difficult to obtain. If anyone happens to have any independent measurements of any of the squares or a comment or two concerning your thoughts about any of these, please do not hesitate to forward them to me at

Place Royale.JPG848.59 KB
Palais Royale.JPG1.13 MB
Wilton Crescent.JPG694.23 KB
kensingtonbromptonbelgrave1.jpg93.39 KB
Brugge-Begijnhof.jpg133.32 KB
Newcourt 1666 london plan.jpg90.16 KB
Pietro de Giacomo Cataneo ideal city 1567.jpg106.03 KB
Proximate600 copy.jpg108.71 KB
1100 ft radius.jpg76.4 KB
ProxiAlmostPerfect.jpg131.4 KB
proxiLinearfinal.jpg135.37 KB
Proximate Park Alt 3 & 4_Page_1.jpg267.93 KB
Proximate Park Alt 3 & 4_Page_2.jpg126.78 KB


Cathedral Square

I lived in MIlwaukee for 36 years and Cathedral Square was my favorite place to spend time. It's MIlwaukee version of Philadelphia's Rittenhouse Square, my other favorite urban square. Like any good urban square, both are completely surrounded by buildings with windows.

paytonc's picture

My favorite squares

I was very impressed with Jardin St. Roch in Quebec City, part of that city's downtown revitalization efforts. It's on the large side for a square, but it hits all the bases: a hardscape plaza (over a parking garage) at the more urban end, an ampitheater transition, and a lush landscape at the quieter end of the park, terminating in a giant waterfall underneath one of the dramatic bluffs that define the city. Plus, the (mostly new) buildings along it all respect a common build-to line and cornice height. It also helps that they're mostly university buildings, which helps to keep the square populated.

I was recently in Savannah and found its squares (example) charming more for their proximity (every 500 feet!) than their proportions. It was hard to judge the proportions, actually, given the height and spread of the immense live oak trees. The powerful presence of civic uses on every single square -- originally, two sides of each square were granted as "trust lots" for civic buildings -- was also very compelling.

Here's a photo of Lafayette Square in Savannah:


Thanks for the Savannah link on the Project for Public Spaces web site. It offers some excellent insight. The area in London around Belgrave Square has a similar walk distance from square to square that you find in Savannah. Both layouts are rich in meaningful public spaces, happening with a frequency one doesn't normally find. we we look at this from the theoretical point of view of the 5 minute walk, we have a diagram with a central square and four sub squares about 550 feet removed from the center and each other. I will try to post a visualization of this later.

It is also worth contemplating that in Savannah we have a very rigid grid system and in Belgravia/Kensington a very irregular layout, but a frequency and walk distance that is almost identical.

Newport London and Pietro de Giacomo Cataneo

I will add the image of the 5 minute walk and the idealized square frequency from Belgravia plus an image of Newport's proposed plan for London in 1666 (rejected) as well as Pietro Giacomo Cataneo's ideal city from 1567. You will note the similarities. I am not sure how to add these images so they may show up in the original post.

London Squares

My favorite Square, and one I hope you will add to your list, is Bedford Square. It is in Bloomsbury and home to the Architectural Association school. It is noteworthy for its four identical facades that surround the Square giving it a very formal Georgian presence.

Bedford Square

James, thanks for the comment about Bedford. I believe Bedford Square was developed out of the great Russell estate between 1775 and 1780 though the Survey of London says 1804. I am not sure which is right. There is a smashing panorama of Bedford at

The square itself is quite long when one considers the overall height of the three story buildings (which I estimate to be on average about 40 feet tall to the top of the balustrade. This gives an enclosure ratio of about 1:9 on the width of 375 feet and a ratio of 1:12.5 on the length of 500 feet. The ratio of the width is close to the Place des Vosges in Paris which is quite wide (actually almost a 100 feet wider)but with taller buildings. Obviously, these ratios are beyond the Renaissance ideals, but mature greenery changes the game and makes much of it work well. Though large, Bedford is considerably smaller than Russell Square a few blocks away which measures a very large 665 feet by 680 feet. These dimensions make any recognition of people across the square impossible to recognize. I lose recognition of facial features at about 100 feet on a cloudy day (though I can still spot individual body characteritics at 250 feet). But 665 feet! No way. This means that one feels that "over there" is almost somewhere else. You have a very different feeling in a small square like Ruskin Place at Seaside (at least on an east/west basis).

Bedford was refurbished about a decade ago to reflect the original oval geometry of its green. The green itself measures about 285 at its greatest width and 375 feet longitudinally. The horizontal proportions of the square are 1 to 1.333, i.e. a square and a third. I will upload later a picture from another site that would give a good feel for this historic square. It does have that consistency that makes the overall volume "read." Bedford and about 10 squares before it were the precursors to the Belgravia squares that generally were a bit later. I need to go visit Bedford again because it has been some time since I was there, and it was a visit of no more than 5 minutes!


Over the weekend, I tried to unravel something that had been worrying me about the valuation assumptions. There is a point (from a developer’s viewpoint) where you have included too much green and your overall economics begin to crumble. As I have been working along on my London Squares project, this question has been in the back of my mind. I reviewed Andrew Miller’s price curve graph again from his MIT thesis and find it very believable, and similar to my experience in the sense that Andrew illustrates that value increases for being on or near an amenity drop rather quickly as one moves away from the amenity. Further, his suggested increase of 22% for frontage on a Square or park seems within reason. As one goes up the economic scale in terms of the market you are trying to hit, the increase for being on an amenity can be significantly greater than the 22% number, but what I did yesterday is create a little diagram to check the loss of standard lots in a grid plan against the increase of value that accrues to the lots near the created amenity. To do this, I created a grid 260 feet by 260. The grid need not be a perfect dimension from a planning point of view, but I needed a module that would allow me to check the loss of small lots against the increase in value. There are multiple reasons focused on the issue of a relatively small green space in the form of a square, but the math should be reflected of this conceptual approach to urban squares. I will upload the base diagram on my original Great Squares post. It will be the one with a green square in the middle and lot designations from A to D.

Bottom line, if you assume the value averages outlined below, there is a 32% increase in net revenues as calculated against the loss of revenues for the loss of lots in this 25 block analysis. I asked myself several other questions, one of which is also dealt with below in Alternative Two. The actual projected market increase that one would accrue in a specific location can vary a lot. So what percentage increase in value for a lot sitting on the amenity is the LOWEST percentage increase to at least cover the loss of the lots removed from the grid. The answer I came up with is about 17%. In other words, if you expect only to get a 10% increase in value for having a lot sitting on the amenity, you will have lost money for including the amenity. You need to get at least a 17% increase to warrant the introduction of the square. At least in this simulation. The introduction of the green hurt your economics rather than helped you.

There are two other simulations that I ran, but we will need some additional graphics to explain them. Briefly, I looked at what happens when another Square is introduced into the grid system. If you look a bit at the diagram above, you will note that there are lots in the 25 block area that have no increase in value (these are the D lots). Bottom line, we can put other squares into the grid system to improve the numbers overall. But how close is optimum?
I think the answer (assuming the 22% increase for A lots Andrew came up with) is not much closer diagonally than 900 feet. I ran two scenarios. One Square was 500 feet away diagonally and one was 950 feet away. The 950 feet diagonal spacing eliminated all “D” lots. Though the 500 spacing eliminated all D lots and upgraded others, the overall effect caused a net loss in value. I will try to upload an additional diagram later.

I assumed a grid system centered around a square that is created out of a full block, eliminating 26 townhome lots that are 20 feet wide and 120 feet deep. The square is 260 ft by 260 feet. "A" lots front directly on the Square. "B" lots are within 300 feet and the increase is priced at half the percentage that the"A" lots increase. "C" lots are between 300 and 600 feet from the Square and are half the "B" lot increase. "D" Lots are at the base price, being more than 600 feet from the Square.

Input base lot value: $50,000

Alternative One: No other Squares near by.
Input per lot increase in value of facing directly on Square: 22.00%
Figure a zone of 25 blocks with 13 lots of 20 feet wide each block face in one direction

Lot Type Projected Increase Increase/lot End Lot value Number of lots Total increases

A 22.00% $11,000 $61,000 52 $572,000

B 11.00% $5,500 $55,500 64 $352,000

C 5.50% $2,750 $52,750 290 $797,500

D 0 $0 $50,000 not counted $0

Total Increase: $1,721,500

For having created the Square, assume
that the loss of lots is 26 lots times $50,000 per lot -$1,300,000

Differential $421,500

Differential compared to loss of lots 132.42%

Alternative Two: At what increase is overall net a break even?
Input per lot increase in value of facing directly on Square: 17.00%
Figure a zone of 25 blocks with 13 lots of 20 feet wide each block face in one direction

Lot Type Projected Increase Increase/lot End Lot value Number of lots Total increases

A 17.00% $8,500 $58,500 52 $442,000

B 8.50% $4,250 $54,250 64 $272,000

C 4.25% $2,125 $52,125 290 $616,250

D 0 $0 $50,000 not counted $0

Total Increase: $1,330,250

For having created the Square, assume
that the loss of lots is 26 lots times $50,000 per lot -$1,300,000

Differential $30,250

Differential compared to loss of lots 102.33%
(Basically about even)

I have added to the images on the original post an excel analysis of Alternative 3 and Alternative 4 where Alternative 3 shows that a spacing of squares approxiamtely 950 feet apart on the diagonal appears to yield the highest possible economics.

Alternative 4 is also analyzed where a continuous linear park is considered. Before I looked at the numbers, I would have assumed that this configuration (linear) would have yielded decent results, but as calculated the results are quite negative. Instead of picking up $388,000 in Alternative 3 (the 950 foot spacing), we lose $836,000). The reason is that such a layout eliminates many more lots than the introduction of a modest square on a 950 foot spacing. It might be possible to improve the outcome by a tweak to the configuration of lots by introducing an additional alley mid block behind the A lots that would only run 120 feet to the main alley (a T configuration). This would allow the lots behind to front the streets perpendicular to the linear park and increase the number of B and C lots on these streets. However, a quick calculation makes me think the overall numbers will not improve - perhaps be even worse. I will also include the linear layout I used to compare to Alternative 3. The strange area chosen to calculate was so in order to have approximately the same number of developable lots.

Frequency and return

Lee Sobel, Andrew Miller and I have been having an offline conversation concerning frequency of squares and the layout issue as initially discussed above. My conclusion after doing more testing is that Andrew's prediction that having as much as frontage as possible on the square is true so long as the elongation is no more than about 500 feet and the square is halved in depth. Another way to say this is that the premium gained by the introduction of a square will be further enhanced if the shape of the square is modified to be longer and less deep so that the area consumed is approximately the same size as the original "square" square but provides more fronting lots. This means that the lots theoretically lost are no greater than in the original square assumption. It appears that instead of a premium gained to lot lost ratio of 130 percent, we increase the premium gained to about 140%. The gain falls off rapidly as we continue to elongate because more lots are being lost until the premium is negative as in the Alternative 4 above. The ideal frequency of the elongated version appears to be about 900 feet apart when measuring from the narrow side and 1200 feet when measuring from the ends of the long side. THis compares to an ideal spacing of 950 feet diagonally on a pure square shape.

There are many side issues - an open space network pattern may have justification apart from this theoretical economic analysis because of other considerations, but this effort has been to think about spacing from a purely mathmatical point of view when the average increase for square fronting lots is 22%.

Egerton Crescent

My favorite from the Belgravia area is Egerton Crescent (for plan see There are many reasons it is noteworthy. For one, it is extremely effiecient from the point of view of maximizing frontages on green while not only minimizing the area of green but also the loss of developable lots. What follows is some preliminary data on Egerton Crescent.

Long Closure Ratio (B): 9.7:1
Across Middle Closure Ratio (B): 4.5:1
Green 110 ft. by 375 ft. at apexes.
Face to Face: longitudinally 475 ft.
Face to Face : middle 220 ft.
Radius to Face: 160 ft
Height A: 44 ft.
Height B: 49 ft.
24 houses on crescent proper plus 5 to each side.

Egerton Crescent (for midblock photo see arose during the mid 1840’s under the hand of James Bonnin, Sr. who had begun the slightly larger Pelham Crescent a decade before. The original mansion was demolished by the trustees of the Smith Charity Estate, and Bonnin was granted an 85 year sublease on what was then known as Brompton Grange in 1843 from his financier and speculator Stephen Phillips who has also been involved in the Pelham Crescent development. It is almost certain that Basevi provided design direction as he had at Pelham. The development was renamed at the latter part of the 19th century for Francis Egerton, one of the Charity’s trustees. The development was rapidly occupied, owing both to the economy of the time and the excellence of conception and execution.

The stuccoed crescent here is continuous, unlike Pelham Crescent, and has no break at its midpoint. Fifteen percent smaller than Pelham, the overall assemblage is more sophisticated – one of my favorite compositions in the district. There are many details to admire.

The architect solution to provide both differentiation and slight variations in square footage and height creates an overall rhythm of DAAAABAAAABCCBAAAABAAAAD. Additionally, the balconies run continuously for two to four units and then break at the units of emphasis.

Story height also varies, from the two one story end masses to the four story primary points of emphasis at the center. Windows distinctly change from level to level vertically and from unit type horizontally. This is done with great skill. The most elaborate windows are reserved for the four story quoined townhomes (the B unit s in the rhythm series). Vertically, the fenestration decreases in scale and pane size at each progressively higher level. Surface differentiation occurs as well, with the entry level fully channeled, whereas the upper levels are smooth surfaced.

Unity and continuity are achieved by a simple white palette throughout, a continuous cornice at the third story, and the repetition of consistent townhome types. Additionally, all entries repeat delicately scaled Ionic capitals atop fully dimensioned square pilasters which flank each entryway. An elegant and discreet awning system has been added for additional doorway protection which add a pleasant randomness since some are extended and some not.

The entry level plan for each unit is similar to the rest, being approximately 750 sf in area and differentiated primarily by door placement and window type. Each unit is approximately 22 feet wide. Ceiling heights are greatest on the second floor, which partially accounts for the particularly fine proportion of each individual façade.

Egerton Crescent

I'll add some comments on Egerton Crescent in Belgravia to original post about Great Squares.


Write your comments in the box below and share on your Facebook!