A reconstruction of the second Globe as sketched by Wenzel Hollar Tim Fitzpatrick and Russell Emerson Department of Performance Studies University of Sydney.

A reconstruction of the second

Globe as sketched by Wenzel Hollar

Tim Fitzpatrick and Russell EmersonDepartment of Performance StudiesUniversity of Sydney

Click anywhere to continue…

This project was initiated and developed by Tim Fitzpatrick in dialogue with Russell Emerson, who created a CAD (computer-aided design) of the reconstruction, and then made a 1:50 card model which he photographed extensively.

The project has been in train for a number of years, and various stages of its results published in print and online journals. Those publications provide a full and detailed explanation of the principles and assumptions that guided our reconstruction (details on next slide).

This slideshow summarises the work, and is designed for individual viewing, not for the purposes of a slideshow-lecture (it’s too complicated for that).

References

Tim Fitzpatrick, “The Fortune Contract and Hollar’s Original Drawing of Southwark: Indications of a Smaller First Globe”, Shakespeare Bulletin, 14: 4 (Fall 1996): pp. 5-10

Tim Fitzpatrick with Russell Emerson, “Reconstructing the spatial dynamics of ‘lost’ theatre spaces; Shakespeare’s second, first and third Globe Theatres”, PaPER (People and Physical Environment Research) 53-54 (1999): pp. 42-57

Tim Fitzpatrick, “Reconstructing Shakespeare’s Second Globe using CAD design tools” Early Modern Literature Studies March 2004 (special edition, ed. Gabriel Egan, of online refereed journal http://www.shu.ac.uk/emls/emlshome.html)

Tim Fitzpatrick, “The Visual Semiotics of Elizabethan Public Playhouses.” International Yearbook of Aesthetics 10 (2006): pp. 29-44

Tim Fitzpatrick, “Patronage and Theatre Design: The First Globe and its modern Reconstruction”, Patronage, Spectacle and the Stage, eds. Irene Eynat-Confino and Eva Sormova, Prague, Prague Theatre Institute, 2006 (refereed proceedings of International Federation for Theatre Research Conference on Scenography: June 18-22, 2003)

Hollar’s sketch

In the 1630s Wenzel Hollar climbed the tower of Southwark Cathedral to do some preparatory sketches for an engraved panorama of London.

One sketch included the second Globe (rebuilt in 1614 on the site of the first).

It is the prominent building in the middle distance, about 400 yards away.

A close-up of

Hollar’s

sketch This detail is only an inch across in the original. It’s literally

a ‘thumbnail’ image. It is amazing that Hollar could cram so much detail into such a

small scale.

Some scholars think this is a ‘topographic drawing’, which would be highly accurate. But it’s not a drawing, it’s a sketch.

A sketch, not a drawing

Look closely at the multiple pencil lines under (and many of them not under) the inking.

Note also where lines have been sketched beyond their point of intersection: a sketching technique, not a drawing based on sightings of intersections.

>

>>

>

>> >

>

>>>

>

>>

>

>

>

>

>

>>

>>

>>

>

>

>

>>

Is Hollar’s sketch accurate enough to help us

get under its skin? If Hollar was sketching rather than doing a

calibrated drawing, it is nevertheless possible that an accurate sketch might still provide invaluable information about the structure of the building he was looking at.

Is it possible to deduce from the sketch the structural principles that governed the building, to work out the ‘skeleton’ which supports the ‘skin’ of Hollar’s sketch?

Could we theorise and build a coherent structure that would look like the building in the sketch — and in the process validate the sketch?

We think the building which Hollar saw and sketched from the tower of Southwark Cathedral might have been this structure.

It is seen here as if it were correctly oriented on the archaeological remains, and viewed from Hollar’s point of view.

We did: first in

CAD, then as a 1:50

card model

This is a sixteen-sided ad quadratum polygonal building.

It has a complex, M-shaped stage cover with two parallel roof ridges ending in a double gable.

And this superstructure is somehow integrated with the polygon.

Between the ridges of the roof is a ‘lantern’ structure. A hole in the roof lets in light, and the open structure above

it is crowned by an onion dome to keep the rain out. There are two external stair-turrets for audience access to the

galleries.

Our main task was to work out how the stage cover, an M-shaped double-gabled structure on a rectangular base, had been integrated with the polygonal structure of the playhouse.

The fact that a rectangle seemed to be somehow anchored on the polygon suggested that the playhouse might have been ‘square’ in some fundamental way.

This in turn indicated that it might have been designed according to the principles of ad quadratum geometry.

How did we arrive at this particular

structural interpretation ofHollar’s sketch?

A brief excursus into ad quadratum

The relationship between the square and the circle has been of fascination to geometricians for hundreds of years.

A simple mathematical relationship flows from a sequence of squares and circles nested within each other: the second circle is twice as wide as the first square.

Ad quadratum in theory

The diameter of the second circle is twice the width of the first square.

You can see this if you rotate one of the squares through 45 degrees, and draw some diagonals: the first square is two triangles across, the second circle is 4 triangles across.

[It is fascinating that this simple geometrical

relationship between the circle and square is proved by introducing the other member of the geometrical trinity, the triangle.]

An ad quadratum sequence wraps a circle around a square, then a square around that circle, then a circle around the second square, etc.

So if you start, say, with a square 43’ across, then the second circle will be 86’ across; the difference between the first and second circles will be 12.6’, or 12’7”.

Ad quadratum and the Globe?

43’

60’10”

86’

11’10”

16’9”12’7”

21’6”

30’5”

So if the second Globe was an ad quadratum structure with a 43’ base, its stage would have been 43’ wide, its galleries 12’7” deep. These numbers generated by an ad quadratum sequence are particularly interesting, given another set of numbers…

It is easy to see why scholars have been tempted to suggest the Globe was based on ad quadratum: the stage would be half the initial square, the gallery would be formed by the two circles.

There is a second reason why we might suspect there was something ‘square’ about the first Globe: the Fortune playhouse was based on it.

In 1599 Peter Street built the first Globe on the Bankside by re-erecting the timbers of the Theatre.

A year later Street was engaged by Richard Henslowe to build a new playhouse, the Fortune.

The building contract has survived, and it was for a square playhouse…

Another interesting set of numbers

The Fortune Contract

Building 80’ square

It is a courtyard playhouse 80’ square.

The courtyard is 55’ square, surrounded by galleries 12’6” deep.

Its stage is to be 43’ wide and to extend into the middle of the yard (so 27’6” deep).

Yard 55’ square

Stage 43’ in length, and half depth of yard

Galleries 12’6” deep

states explicitly that the Fortune

is to be modelled on the first Globe

“…, and in all other respects to be modelled on

the recently completed Globe on the Bankside.”

But the contract has some funny

numbers…

Building 80’ square

The galleries are to be 12’6” deep. Is that needed in the contract? Anyone can subtract 55 from 80 and divide by two to get 12.5, i.e. 12’6”.

The stage is to be 43’ in length. If the stage depth is half of 55, why not have its length half of 80, i.e. 40’?

Could these two numbers then reflect measurements brought across from its model, the first Globe?

Yard 55’ square

Stage 43’ in length, and half depth of yard

Galleries 12’6” deep

one unnecessary number:

and one strange number:

What if Peter Street measured relevant bits of the Globe (stage width, depth of galleries) as the basis for the plan of a square playhouse?

The Fortune contract’s combination of measurements (stage 43’ wide and galleries 12’6” deep) is very close to the combination generated by ad quadratum (a 43’ base gives 12’7”).

If these numbers are indeed relics of the Globe, they might point to an ad quadratum first Globe.

It would have been 86’ across, having a much smaller yard than the 100’-wide modern Globe reconstruction.

Superimposing this notional first Globe onto the Fortune contract certainly makes the Fortune look like a ‘square version’ of it…

Peter Street and a tape measure at the Globe?

The 80’ Fortune superimposed on an 86’ ad quadratum first Globe

The second Globe was built quickly, on the same site as the first.

Its size was in all likelihood constrained by some royal ordinances designed to limit building density.

So it may have been the same size, on the same foundations, as the first Globe.

So if the first was an ad quadratum design, the second one might have been too.

Has Hollar accidentally sketched an ad quadratum second Globe?

We could try and find out, by building an 86’ ad quadratum structure to see if it matches Hollar. This is where the fun starts…

One final assumption…

The 80’Fortune superimposed on an 86’ ad quadratum (second)

Globe?

16-sided 86' plan

20-sided 100' plan

But before the fun starts,

an archaeological problem: would such a

structure fit the foundations?

A dig at the Globe site in 1989 uncovered some traces of a polygonal building, but the junction points are very rough.

Nevertheless the designers of the third Globe reconstruction jumped on them as validating a 20-sided 100’ diameter polygon.

But a 16-sided 86’ polygon has a bay size that sits on the foundations just as well as does the 20-sided 100’ plan.

Two guesses as to where one bay of the polygonal gallery might have sat on the

remains: the 16- and 20-sided plans are similar in size and orientation.

Hollar's POV 280.5deg

Bay junction post

Hollar's stair turret

2 metres scaled to this small-scale illustration from the ACT89 1:20 scale drawing

0 2m

N

16/86

Some considerable time later we had a CAD plan we could superimpose on

the remains

Two wall junctions suggest a bay of the polygon.

[N.B. Do not trust theatre

historians turned archaeologists: the scale and N orientation on the published

image are wrong, requiring our correction.]

CAD line overlays show where our polygon would run.

Hollar's POV 280.5deg

Bay junction post

Hollar's stair turret

2 metres scaled to this small-scale illustration from the ACT89 1:20 scale drawing

0 2m

N

16/86

Some features of the remains require comment

if we’re looking at them from Hollar’s point of view

To Southwark Cathedral:

we also know the bearing of Hollar’s

point of view (280.5 degrees), as

well as its distance and

relative elevation from this wall

junction.

The stair turret Hollar shows in the sketch would

have been south of the lean-to

structure.

A lean-to structure, probably an

entrance foyer, not a ‘stair

turret’.

We assumed an ad quadratum 16-sided polygon.

This meant the stage cover could be supported by four posts of the polygon, since the relevant posts (circled) are in a rectangular relation to each other due to the ad quadratum geometry of the design.

This stage cover would span the yard at its widest point, and cover the whole of the stage down to the mid-line of the polygon (as Hollar’s sketch shows).

TF Fig 8

How did we arrive at a CAD plan of

a 16-sided ad quadratum

structure that would sit on the

archaeology?

We started with the problem of the M-shaped rectangular stage cover anchored onto a polygon:

However its principal support is not at the widest point, but one bay back from the downstage line, where a beam (a---b) can be supported as well by two stage posts (c and d)—since we know the stages featured such posts.

These stage posts, on the 45 degree radials, would also (due to the mysteries of ad quadratum geometry) be in a rectangular relation to two other posts of the polygonal playhouse structure, e and f. This would provide a solid rectangular base for the stage cover.

We also found that if we positioned the stage posts (c and d) on these 45 degree radials, they gave us the right roof profile: the ‘nick’ in our M (the gully between the two roof ridges) matched Hollar’s! We’ll see below how the ‘nick’ in the M depends on the position of these posts.

TF Fig 10

a bc d

e f

gh j

Structural considerations: how to build it so it stays up…

It has no main beam or ‘bottom chord’ (it would be where the faint dotted

line is). Instead a collar tie

higher up stops the diagonal beams from splaying outwards.

These structural considerations result in a the final scale model that looks like this:

The diagonal raking beams

They are connected to the vertical posts of the polygon at the

downstage line.

The front of the M-shaped double-gabled stage cover

The serious structural work is happening one bay back from the front of the stage.

Here is the main horizontal beam, with the roof pitched on it as two overlapping triangles supported on their inwards ends by the two stage posts…

But the main support for the stage cover must be elsewhere…

A pair of overlapping triangles, each of them supported by a vertical post in the polygonal ring and by one of the two stage posts.

This is the best explanation for the choice of the complex M-shaped roof structure with two parallel ridges running down the stage, instead of a simple single-gabled structure running across the stage: it’s not an M, it’s two overlapping triangles.

…like this:

(Click a couple of times to turn some highlighting on and off.)

The positioning of the stage posts determines the overlap of the triangles and hence the size of the ‘nick’ in the M (the closer together they are, the bigger the ‘nick’ will be).

We discovered that when we positioned the stage posts in a structurally logical place, our ‘nick’ matched Hollar’s. It was then that we thought we were onto something…

In addition there are vertical ‘queen struts’ to support a collar tie (the horizontal beam higher up, which keeps the roof from splaying outwards).

‘Nicks’ and collar ties

(Click a couple of times to turn some highlighting on and off.)

Looking in under the collar tie

These structures one bay back from the front of the stage cover, in under the front collar tie and gable, can clearly be seen on the model from Hollar’s point of view.

Hollar too seems to suggest he could discern some structural complexity in the same part of his building: he sketches a number of lines there.

The model mapped onto Hollar’s sketch

When we photograph this model from the scaled distance and relative elevation of Hollar’s point of view, it looks very similar to the building in his sketch.

The image on the right superimposes a computer-generated outline of the model onto the sketch. Looking more closely…

The orange markers below map the outline of the model. The green indicate Hollar’s pencil lines which the model validates over his later ink lines. The blue markers bottom left indicate Hollar’s guesses for a baseline of the polygon.

>

>

> >

>

>>

>

<

>

>>

>

>

>

>>

>

>

> >> >

>> >>

>

>> >>

>>

Click anywhere, then again and again, to see how the model maps onto Hollar’s sketch.

A pretty good fit, but…

Hollar’s stage cover is further to the left than the model would have it.

On the right there is confusion in the sketch; the model matches Hollar’s pencil lines, not his later inking.

The model’s baseline is ‘underground’ compared to Hollar’s guesses at where the building’s baseline might be.

Let’s deal in turn with these discrepancies….

Discrepancy 1) is Hollar’s fault: he has sketched unevenly, making the left-hand gable smaller than the right. The left one should be bigger, as

it is closer to us, and this would move the whole structure further to the right and bring it into alignment with the model.

>>

>>

>

>

>

Click once to see the sketch, then again to see the model—and watch the markers

>

Discrepancy 2) On the right there are two pencil lines, one vertical and the other showing the curved ridge of the polygon roof, which Hollar has

not inked (and he has inked elsewhere, where there is no pencil line). Who knows why his later inking didn’t respect his pencil lines.

>

>

> >> >

>

>

Click once to see the sketch, then again to see the model—and watch the markers

Discrepancy 3) In the bottom left corner the model is lower than the lowest of the

three pencil lines, indicated by the blue markers, where Hollar was guessing at the baseline of the playhouse—a baseline that (according to

the sketch) was hidden behind bushes.

>>

> >>

>>

Click once to see the sketch, then again to see the model—and watch the markers

>

If our structure is the ‘skeleton’

underneath Hollar’s ‘skin’…

…then there are four important points of detail where it might cause us to reconsider our view of what the inside of the playhouse was like.

The inside is quite different from that of the third Globe reconstruction in London.

Further upstage another horizontal beam is supported by the next set of posts in the polygonal ring; it too is crowned by overlapping triangles.

When we position the dome-topped ‘lantern’ above this beam structure, it matches Hollar’s positioning.

You can see here how this ‘lantern’ would let in light to upstage — light which flares off the diagonal raking beams in this image.

Detail 1: getting

light onto the

stage

If the lantern (g) is positioned above that beam (the fine dotted horizontal line h—j), it is too far back to maximise light onto the rear part of a stage that had a straight back wall running along this chord (the red stage).

This suggests that the stage was deeper, with its back wall following the cants of the polygon, as was the case at the Rose playhouse (the green stage).

The need to get more light into this upstage area would explain why a lantern structure was built in the gully between the two roof ridges (you wouldn’t buy into the drainage problems unless you had to).TF Fig 10

a bc d

e f

gh j

The back wall of the stage may therefore have looked like this: angled, continuing the line of the polygon.

Detail 2: the shape of the stage

This angled wall forms a shallow triangular recess upstage centre.

Curtains could be hung on a rod across this recess to create a concealment space (it is about three feet deep at the centre).

Detail 3: a

recess to

serve as a

curtained

discovery

space

This curtained recess would have provided a ‘concealment space’ — a space where Polonius could hide unsuccessfully…

… or from where Volpone could observe: ‘Volpone peeps from behind a traverse’ (Ben Jonson, Volpone 5.3.8).

1

2

3

“b r e a d t h”“l e n g t

h”

Note how the stage posts, positioned one bay back from the downstage line, provide a substantial unencumbered downstage playing area.

The posts are 11’ from the front of the stage, 10’6” from the sides of the stage, and 22’ apart.

Detail 4: the

position of the

stage posts

11’

In contrast, the stage posts in London (further downstage and wider-set) constrict both the downstage and the extreme corners of the stage

Since this photograph was taken the square bases of the pillars have been ‘shaved’ and made octagonal to reduce their intrusiveness.

6’

Comparing the stage with that at the

London reconstruction

A FINAL SUMMARY….

We have constructed a theoretical object based on one interpretation of the evidence for the first

and second Globes.

If this object had been on the site of the second Globe when Hollar climbed up the tower of Southwark Cathedral in the 1630s, it would have looked very

much like the object he sketched.

This may just be a coincidence, of course — but if it’s not then the structural logic of our object

suggests four things about the inside of the second Globe that are quite different from the third

Globe.