Colonial Williamsburg Digital Library

KING'S ARMS TAVERN

1707--
FIRST OWNER OF LOT #23 WAS PROBABLY JAMES SHIELDS
1749--

The early history of lot #23 is scanty and confusing but, apparently its first owner was James Sheilds, although positive evidence of this has not been found. We do know, however, that Sheilds owned the neighboring lot #24 for he deeded it in May of 1707 to Colonel William Byrd II of Westover, and also lot #25 to the east because he conveyed this in the same year to the tavern keeper, John Marot, noting at the time that it was next "to the lot formerly sold by the said Sheilds to William Byrd, Esq." It is probable that Sheilds owned all three contiguous lots, #23, 24 and 25 for a certain James Crosby, a merchant of Glasgow, sold the 1749 lots in a block to Andrew and Archibald Buchanan and Company in 1749 and mentioned in a deed that they had formerly been the property of "the late Colonell William Bird of Charles City County." We know that Byrd never owned lots #25 or 26, so that it must have been lots #22, 23 and 24 which the deed covers. The document also furnishes the information that there were, on the three lots, a dwelling house and kitchen, a storehouse on the center lot and a warehouse and stable "upon the other lot." It is likely, then, that a storehouse stood on lot M, the middle lot. From this description it is not possible to determine on which of the two flanking lots, #22 or #24, the other buildings stood.

1757
PURDIE LOT, #24, SOLD TO DR. McKENZIE

On November 20, 1753, John Hyndman and Hugh Blackburn, attorneys for the Buchanans sold lot #24 to Dr. Kenneth McKenzie, a famous surgeon-apothecary who a few years before had established himself on the Carter-Saunders property near the Governor's Palace. He had turned this dwelling house over to Robert 9 Dinwiddie when the latter arrived as royal governor in Williamsburg in November, 1751, and found the Palace undergoing repairs which made it temporarily untenantable. McKenzie did not long enjoy his now home (Purdie's House' for he died sometime before June 1757, and his widow, Joanna had difficulty establishing her title to it.

1760
JOHN AND JAMES CARTER PURCHASE LOTS #23 & 24 RESPECTIVELY

Andrew and Archibald Buchanan sold lot #23 (King's Arms) to John Carter sometime before 1760 and in June of that year Joanna McKenzie sold her lot #24 to his brother, Dr. James Carter. Dr. James, we believe, built the Tayloe House sometime between 1752 and 1759 and we know that he built the brick double shop just west of the Raleigh Tavern in 1764. In this latter building his brothers, William and John later came to have an interest. This was an enterprising trio for Dr. William Carter became proprietor in 1761 of the Brick House Tavern a few doors to the west of the King's Arms while John owned the latter. John operated a general store on this property and though we do not know where this was it seems reasonable to think that it might have been in the small building on the western border of lot #23 which later became a barber shop.

1772
JANE VOBE TRANSFERS HER TAVERN BUSINESS TO LOT #23

By February, 1772 the famous Mrs. Jane Vobe had taken over lot #23 and transferred to it her tavern business from its erstwhile location on the east side of Waller Street behind the Capitol -- the area known as the Exchange. Before making the shift, Mrs. Vobe, in July, 1771, sold all her kitchen and other household furniture at auction and in September of that year, on the oft used pretext that she was leaving the colony, informed 10 her debtors that she would sue them if they failed to settle their accounts with her immediately. Obviously she wanted to place all her affairs in order before starting her new enterprise and also to provide her hostelry with brand new furniture and cooking equipment.

HER NOTICE IN GAZETTE IS FIRST KNOWN REFERENCE TO KING'S ARMS

So far as we know, Mrs. Vobe was the first to Operate a tavern on lot #23 and the advertisement which she placed in the Virginia Gazette of February 6, 1772 announcing the opening of business contained the earliest known mention of the King's Arms Tavern:

I BEG Leave to acquaint my former Customers and the Publick in General., that I have just opened Tavern opposite to the Raleigh, at the Sign of the King's Arms, being the House lately occupied by~: Mr. John Carter, and shall be much obliged to the Gentlemen who favour me with their Company.
JANE VOBE

I am in Want of a good COOK, and would be glad to hire or Purchase one.

MRS. JANE VOBE MAINTAINED A FASHIONABLE ESTABLISHMENT, THE SCENE OF VARIED EVENTS

In order that the reader of this report may have some notion of the character of Jane Vobe and of the establishment which she conducted we will quote a few passages concerning it from the research report on the King's Arms Tavern By Mary Stephenson:

Jane Vobe was among the better known tavern keepers in Williamsburg. In 1765, (when she was operating at a different location), a French traveller wrote that he "got a room at mrs. vaube's tavern, where all the best people resorted." ("French Traveller in the Colonies," American Historical Review, XXVI, 741-42.) Virginia gentlemen stopped at Mrs. Vobe's when in the city: among them, Colonel William Byrd 11 and George Washington. Her establishment was cited as a place where artists held displays, where meetings were held, where rewards were made for returned articles lost by tavern guests, as well as being cited with reference to locating neighboring property.

MORE OF THIS AND ABOUT HER FAMOUS PATRONS MAY BE READ IN RESEARCH REPORT

Evidently, then, Mrs. Vobe's King's Arms Tavern was a high class establishment. Those who wish to know more about what went on there and who patronized it may turn to page 6 of the Stephenson research report and inform themselves on these subjects. What we are chiefly interested in here is the architectural characteristics of the main and subordinate buildings on lot #23. In that connection any scraps of information bearing on the architecture are of Importance to us.

ACCOUNTS OF HARWOOD GIVE MUCH INFORMATION ABOUT KING'S ARMS AND ITS OUTBUILDINGS

During Jane Vobe's occupancy of the King's Arms, Tavern (1772- 1789) many repairs were made and much new building was done, as is evident from the accounts kept by Humphrey Harwood, Williamsburg carpenter and brick mason of work done for Mrs. Vobe. From these accounts, which cover a period of eight years, the number of rooms, passages, porches, stairways, etc. of the building can be more or less exactly determined as well as the identity of the outbuildings on the lot.

THE TAVERN CONTAINED AT LEAST 14 ROOMS, FOUR HALLS AND TWO STAIRWAYS

Whitewashing of the plastered walls of the interior of the King's Arms was a frequent occurrence, judging by the references to this work in Harwood's ledger. Harwood was called upon to do this work every year or so during the eight year period covered by the accounts because there are five entries recording the whitewashing of what must have been most, if not in each case, all of the interior rooms and hallways. The building contained, apparently, 14 rooms, for this number is mentioned twice while 12 three other references give lesser numbers -- 12, 13 and 13, respectively. There were evidently four hallways since there are four references to this number of "passages." Two "stare ways" are mentioned and we would probably be justified in assuming that this meant two stairways from the first to the second floor.

HARWOOD LISTS BUILDING OF CELLAR STEPS

In April, 1785, Harwood lists the building of "cellar steps" but whether this means outside or inside steps we have no means of knowing. Archaeological investigation has revealed that steps at one time existed within the house between the later, higher basement on the west side and the early, lower basement on the east side and that there was an areaway at the southwest a flight of steps leading to the west basement from the outside.

A BAR IS ALSO LISTED

A "barr" is mentioned and there is no reason to doubt that this meant the eighteenth century equivalent of our present day facility for serving drinks.

WHITEWASHING OF PORCHES MENTIONED

Two porches are listed in one instance as being whitewashed. In another case the "front porch" is undergoing the same treatment and, in still another, "ye porch". probably the main front porch, is being repaired and underpinned.

RECORD SPEAKS OF BUILDING PILLARS AND WALLS IN BASEMENT

Harwood lists, under the date of September 18, 1778, an item which he designates as "pillering cellar." This probably refers to the erection of the two 15" x 15" brick piers, remains of which were found in excavating the basement. These piers were built in the old east portion of the basement, apparently, to support the beams of the floor which was substituted for the vaulted ceiling after the latter had fallen down. Later, on January 27, 1780, he also lists an item entitled "building up Cellar walls." This 13 may refer to the 4" supplementary walls which were built up against the insides of the north and south walls of the old part of the basement, presumably to receive the beams of the new floor of which we have just spoken (see archaeological drawing, p.39).

REPAIR OF FIREPLACES, CHIMNEYS, STOVES AND WELL AND BUILDING OF OVEN CITED

There were at least four fireplaces in the house since this number is spoken of in 1779 as having been repaired. The "kitchen back" is mentioned twice as having been "mended." The kitchen back was, presumably the fireback or rear wall of the kitchen fireplace. There are also several references to "mending stove in kitchen." It is likely that the stove or stoves referred to in these items were one or another of the case iron types which we know to have existed with the eighteenth century in Virginia. In this connection it is well to mention that Harwood in 1779 speaks of building an oven. It is likely that this would have been a brick oven somewhere in the kitchen or in close proximity to it. An entry of 1784 covers the repair of the chimneys of the Kitchen and Laundry and there are also two references to repairing the well.

HARWOOD LISTS CHARGES FOR WORK ON STORE AND STOREHOUSE

As we have remarked, several of the outbuildings are mentioned in Harwood's ledger. Under the date of August 11, 1777, the Williamsburg carpenter and brick mason lists a charge for lime, lath and labor "for Storehouse." On October 8 of the same year there is a charge for plastering "Store." It is possible that the two items refer to the same building but this is not certain.

OUTBUILDINGS MENTIONED IN LEDGER RECONSTRUCTED ALONG WITH CERTAIN OTHERS

We know that between 1777-1785 the outbuildings mentioned above existed on the King's Arms lot, #23, because during those 14 years, Mrs. Vobe owned no other property in Williamsburg so that the work Harwood did for her must have been on King's Arms buildings. There were therefore, a kitchen, a laundry, a store and/or storehouse and a well on the lot during Mrs. Vobe's occupancy of it. If, as seems reasonable, the small building west of the Tavern which we know as the Barber Shop was at one time a store or storehouse, we may state that all of these buildings have been reconstructed, along with three others, the two Privies and the Stable which are not cited in Harwood's ledger. As for these latter structures, the Stable is listed in two insurance policies and the Privies, obviously, would have been on the lot. In addition to these buildings, the well mentioned by Harwood, which was still in existence on the plot has been reconstructed in its old location. This location was such as to cause the north wall of the Well House to coincide with a part of the south wall of the Tavern wing so that the Well House was reconstructed as an appendage of the latter.

THE KING'S ARMS A "GENTEEL TWO-STORY HOUSE" CONVERTED TO TAVERN USES

On November 10, 1785, Jane Vobe in an announcement in the Virginia Gazette and Weekly-Advertiser offered the Tavern for sale or for rent. She describes the building in the "ad" as a "genteel two story house, pleasantly situated on the main street, opposite the Rawleigh Tavern…" The quotation gives us a notion of the character of the building. It was obviously a rather large size dwelling converted to tavern uses. Something similar, indeed, was the case with many eighteenth century taverns and shops in Virginia. These began their existence usually as, houses and, 15 as time went on, they, or parts of them, were converted to some sort of business use.

MRS. VOBE DIES

Jane Vobe died in 1789 and the property passed on to her estate.

PART OF FRENCHMAN'S MAP WITH LOTS 22, 23 & 24

1782
FRENCHMAN'S MAP SHOWS KING'S ARMS AND PURDIE LOTS WITH MAIN HOUSES AND OUTBUILDINGS

The next date to take note of in connection with the King's Arms Tavern is 1782 when an unidentified cartographer probably attached to one of the French armies stationed on the peninsula during that period wiled away his time going around the to town and pacing off the sizes of the lots and the buildings on them. He set this information down on a map of Williamsburg which, fortunately, has been preserved and which has proven invaluable to the restorers as an aid in determining the approximate locations and sizes of the 16 buildings which existed on the various colonial lots of the city. We have reproduced here that portion of the map which shows lots #22, #23 and #24 in Block #9 on the south side of Duke of Gloucester Street. The lot which, to the best of our belief, represents the King's Arms property, has been marked with a red #23. It's companion at the right is the Purdie lot which has been marked with its colonial lot number, #24. The shadEd rectangle directly adjoining the King's Arms at the west, we believe, comprises two buildings, the King's Arms Barber Shop and the Charlton House. A faint line in the original map appears to divide this rectangle into a narrow east part (the Barber Shop) and a wider west portion (Charlton's). The squarish building west of the latter building has not been identified.

INSURANCE POLICY OF 1796 SHOWS TAVERN AND OUTBUILDINGS ON LOT #23

Jane Vobe's estate paid a land tax as late as 1795. An insurance policy issued in 1796 by the Mutual Assurance Society of Virginia indicates that the property had in the meantime been sold to Philip Moody for the policy is taken out in his name. It is referred to in the policy as the "Eagle Tavern." In the document Moody characterizes his property as "my Wooden Buildings on the main street at Williamsburg… ." As was customary, the policy includes a diagram in which the outline of each building is shown in the approximate position of the structure in respect to the others on the lot. The dimensions of each building and the material out of which it is made are indicated on the rectangle representing the building. Thus, we learn from the diagram that the four buildings shown, the Eagle Tavern, the Barber Shop, the Kitchen and the Stable are all of wood. The sizes 17 are as follows:

INSURANCE PLANS OF 1796 AND 1806 ARE, UNACCOUNTABLY, REVERSED

It should be remarked that the plan has been for some reason with which we are not acquainted, so that the Barber Shop stands east of the Tavern. The same error was made in the plan in the next policy of 1806, in which Botetourt Street is indicated as being east rather than west of the property. We are certain that this plan is reversed because we know from the description of the location of the lot given in the policy that it is the correct one on the south side of Duke of Gloucester Street.

POLICY GIVES VALUATION OF BUILDINGS AND BRIEF DESCRIPTION OF THEM; THE HOUSE AND BARBER SHOP

In his policy of 1806, Philip Moody insured his four buildings for the following amounts, which give us some idea of their relative pretentiousness, even though the value of the dollar has changed since that time:

THE KITCHEN AND STABLE

The Kitchen is "of wood 1 story high 30 by 16 feet." The Stable is "of wood 1 story high 52 by 20 feet." Attached to one end is a wood shed "52 by ?." The dimension of 52 [feet] would seem to indicate that the woodshed ran the entire length of the stable but it is nevertheless indicated as being on the 20 foot wide end.

SIZES OF BUILDINGS VARY FROM ONE POLICY TO ANOTHER

It will be noted that the sizes of the buildings as given in the two platS vary somewhat - generally only by a matter of a few feet. This was not unusual in these old insurance diagrams since the buildings were probably paced off rather than measured. The one large discrepancy is in the length of the kitchen, given in the plat of 1796 as 50 feet and in that of 1806 as 30 feet. If this is not simply an error it would seem to indicate that a portion of the building had been torn down between the two dates.

LITTLE KNOWN OF TAVERN AFTER 1809; IT PROBABLY FELL VICTIM TO FIRE

The last mention we have of Moody's ownership of lot #24 occurs in a document of 1809. To make matters short with the remainder of the architectural history of the King's Arms Tavern - although we can trace, except for certain hiatuses, the ownership of lot #23 from 1809 down to 1949 when the reconstruction of the buildings was undertaken, we have no further architectural references to it. No eighteenth century structure remained on the lot when the reconstruction of the building began, the only part of it which survived being the underground part - the foundations. We believe the building was destroyed by fire since the cellar area, when excavated, was found to be covered at one level by ashes a foot deep.

STREET FRONTS OF ALEXANDER PURDIE HOUSE (LEFT) AND KING'S ARMS TAVERN. THE COCNNECTING LINK IS NOT VISIBLE HERE.

GARDEN SIDE VIEW OF ALEXANDER PURDIE HOUSE SHOWING IN FOREGROUND, LEFT THE SMOKEHOUSE AND KITCHEN AND AT RIGHT THE DAIRY AND STOREHOUSE. THE WELL AND WOODSHED, ALSO ON RIGHT, ARE OUT OF SIGHT.

ALEXANDER PURDIE HOUSE
ARCHITECTURAL FACTS REVEALED BY OLD DOCUMENTS

1707
JAMES SHEILDS FIRST RECORDED OWNER OF LOT #24

As in the case of lot #23, the first owner of lot #24 of whom we have a record was James Sheilds who, as was noted on p. 6p conveyed the property to William Byrd II in 1707. The lot at that time had "one good dwelling house thereon built."

1707-1773
PURDIE BUYS LOT IN 1767; FACTS ABOUT HIS BUSINESS

The sequence of ownership of lot #24 from this point to 1760 when Joanna McKenzie sold it to James Carter has been given on pp. 6, 7. The next owner was Alexander Purdie, printer, editor of the Virginia Gazette and merchant who purchased the lot from Carter in 1767, Purdie, apparently, not only lived in the house but also ran a dry goods store in it for he advertised yard goods and wearing apparel for sale there in 1772 and 1773. We have good reason to believe that his printing establishment was located in Tarpley's Store a few buildings farther west, on the southeast corner of Duke of Gloucester and Botetourt Streets.

1777-1778
HARWOOD WORKS ON PURDIE'S KITCHEN

Humphrey Harwood, Williamsburg carpenter-mason repaired certain of the buildings on the lot for Purdie in 1777 and 1778 and lists the work done for him in his ledger. Certain of the items are of interest to us, such as the work done on the Kitchen, i.e., taking down and rebuilding the chimney; building an oven and laying the floor and lathing and plastering the Kitchen, including the stairs. He also repaired the well at this time.

HIS WORK ON LAUNDRY AND OFFICE

In October, 1777, Harwood whitewashed the Kitchen and Laundry. Just where the Laundry might have been we do not know, since no foundation evidence of such a building was discovered.

An entry for March., 1778, records the repair "Office Steps 22 & Chimney back." Again, we do not know where this may have been, of courses on the Tarplay lot in proximity to Purdie's printing office.

1779-1789
PURDIE DIES; LATER OWNERS & DESCRIPTIONS OF BUILDINGS ON LOT #24

Alexander Purdie died in 1779 and the executor sold his lot to Thomas Cartwright in 1780. Three years later the latter offered the property for sale in the Gazette. The advertisement describes the buildings on the lot as "a large commodious dwelling house, with four rooms on a floor, a kitchen, stable, and other convenient outhouses, situate on the main street, between Mrs. Vobe's and the capitol." A certain Sheldon Moss purchased the lot in 1783 and he was succeeded in the ownership of it by James Davis, a tailor, who acquired the property in 1786. Davis offered the lot for sale in 1787 and again in 1789. Since the later advertisement contains all the data concerning the property given in the first with some facts in addition, we will quote from it:

It is a commodious house, in good repair, has six rooms below stairs, and three above, with five fireplaces, and all convenient out houses, such as kitchen, laundry, dairy, smoke-house, carriage houses, fine dry cellars; an excellent well of water in the yard, and a large garden enclosed … .[Virginia Independent Chronicle and General Advertiser, November 25, 1789.]

"ADS" OF 1783 AND 1789 DO NOT AGREE AS TO NUMBER OF ROOMS IN THE PURDIE HOUSE

It will be noted from the above that a laundry is again mentioned, so that there must have been such a building on the plot although this has not been reconstructed. It is also of interest that the statements concerning the number of rooms in the house that the statements concerning the number of rooms in the house which Cartwright and Davis made in their "ads" are at variance. The interior of the house, of course, could have been altered in the six years which separate the two advertisements but it is also 23 possible that the two men counted the rooms in different manners.^*

1796
JAMES DAVIS INSURES THE PROPERTY

James Davis failed to sell the property for in May, 1796, he insured "his buildings on the main street now occupied by Cyrus Griffin."^** The policy covers three buildings and these are shown on a plat which describes them as follows:

1790-1809
HARWOOD LISTS REPAIRS TO SHED AND STABLE; POLICY OF 1809 DISAGREES IN DETAILS WITH THAT OF 1796
"EXPRESSED" PARTITIONS BRING FIRST FLOOR ROOM COUNT TO SIX

An account in Humphrey Harwood's ledger, dated April 20, 1790, covering work done for Davis, lists the underpinning of "the shead" and "the Stable." That the shed is not mentioned in the foregoing policy does not signify that it did not exist at the time the policy was written. These early insurance policies, we know, were not, 24 of necessity, either complete in their listing of the insured buildings or accurate in giving the dimensions of them. A later policy, #968 of 1809, for example, which covers the same property since the appraisers speak of "the buildings, heretofore declared for assurance by James Davis," disagrees with the policy of 1796 in a number of details. This shows on the plat two buildings, an "Outhouse" and a smokehouse, in addition to the three structures listed in the earlier policy. The house is described as follows: "a dwelling house one story high covered with wood and body of wood 64 feet front and 26 feet deep owned by John Coke and occupied by Samuel Sheild Jr." The kitchen is indicated as being "of wood one story high 34 feet by 16 feet" and the Stable "of wood one story 15 x 16 feet." Comparison of these sizes with the ones of the policy of 1796 reveals that the House is now longer and the Stable considerably smaller, only the Kitchen remaining the same size. This disparity in dimensions could, of course indicate that the buildings in question had been altered but it more probably signifies no more than that the dimensions of one or both instances were taken in a hasty and careless manner.

1809-1930
BUCKTROUT-MACON HOUSE & ITS REMOVAL; FATE OF PURDIE HOUSE NOT KNOWN

From the date of the second policy, 1809 to the time of the Civil War, nothing certain is known of this property. It is believed that Richard Bucktrout, son of the cabinetmaker, Benjamin Bucktrout, about the middle of the century built upon the lot the house which came to be known as the Bucktrout House and, subsequently, the Macon House. This house remained on the lot until the property was purchased by the Williamsburg Holding 25 Corporation which removed it in 1930 because it was late in period. Whether Bucktrout tore down the original Purdie House to build his own structure is a matter of conjecture.

GENERAL NOTES

KING'S ARMS AND PURDIE'S WERE SEPARATE STRUCTURES

It should be remembered that the King's Arms Tavern and the Alexander Purdie House were quite distinct structures, the present one story serving pantry link, as we have said, having been added in the reconstruction of the buildings to facilitate their joint use as a restaurant. After a brief comparison of the two, therefore we will treat the exteriors of the two houses (for the King's Arms was also originally a dwelling) separately.

COMPARATIVE DIMENSIONS OF THE TWO BUILDINGS

As for the relative sizes of the buildings, the Alexander Purdie House is somewhat larger than the King's Arms Tavern, when the south wing of the latter is left out of consideration. The measurements are: Length, Purdie's - 60'-0"; King's Arms 53'-9": Depth, Purdie's - 26'-2"; King's Arms - 24'-10". Very complete foundations for the main part of the Tavern existed and these, together with plan dimensions recorded in the insurance plats of 1796 and 1806, furnished with great exactness the external dimensions of the original part of this building. Only a few fragments of the eastern part of the old foundations of the Purdie House remained together with traces of the remainder of the foundation but, such as they were, they were in substantial agreement with the dimensions of the building as given in two early insurance policies (see pp. 23, 24).

BOTH HOUSES HAD WOOD FRAMEWORK AND BRICK FOUNDATIONS

The Purdie House and the main part of the Tavern were, therefore, pretty much of a size. The bodies of both were constructed 29 of wood, as we learn from information given on the insurance plats and the foundations and chimneys and the east end wall of the Tavern, as well, were made of the only material suitable for these purposes which was readily available in Williamsburg during colonial times, viz., brick. Both houses were assumed to have had walls sealed on the outside with weatherboarding and roofs covered with shingles, since these treatments of walls and roof were by far the commonest for wood houses in Tidewater Virginia.

EACH HAD COVERED FRONT PORCH

We believe that both buildings had covered front porches. We are certain that this was the case with the Tavern because Humphrey Hardwood in a listing of work done in 1778 for Mrs. Vobe specifically mentions whitewashing the front porch. In the case of Purdie's an advertisement of February 22, 1787 in the Virginia Gazette and Weekly Advertiser in which the house was offered for sale, mentions "good dry cellars" so that we know the first floor was raised above the ground and that, therefore, steps would have been required to reach it. We have no specific authority for the use of a porch roof, except that, in buildings of this size, some form of protection for a stoop was more customary than not in Williamsburg.

ROOF TYPES DIFFERENT; PURDIE'S HAS "A" ROOF AND KING'S ARMS A GAMBREL

The two buildings are, therefore, similar on the exterior i a number of respects. Their most striking difference lies in the fact that two different modes of roof were used to cover them, Purdie's having an A roof and the Tavern a gambrel. In making this last statement we are treading on very firm ground since the aforementioned insurance plat of 1806 describes the Tavern as having a "dutch roof", which was another way of saying "gambrel." The Purdie plat of 1809, on the other hand, describes that house as having one story which could have meant only that it was covered 30 by an A roof. That this house had living space under this A roof and, consequently, dormers is proven by the extracts from for sale "ads" given on p. 22 which speak, on the one hand, of a house with four rooms on a floor and, on the other, of one having three rooms above stairs.

SLOPE OF PURDIE ROOF REPRESENTS AVERAGE OF SLOPES OF SEVERAL OLD ROOFS

The roof of the Alexander Purdie House is composed of two equal slopes of about 47°. This slope was chosen as being characteristic of A roofed story and a half wood frame houses of the period in Williamsburg. A comparison of this slope with the "pitches" of the roofs of several other old story and a half wood frame houses in the town will reveal the justification for the use of this inclination.
(All slopes are approximate)

PURDIE ROOF "JERKINHEAD" TYPE WITH "CLIPPED" GABLE ENDS

The Purdie roof is of the "jerkinhead" type, having what are known in Virginia as "clipped" gable ends. This merely means that the upper part of each gable end is truncated by a plane passing through it at approximately the same angle as the main roof slope (i. e., 47°). The effect of this is to produce at either end a triangular area of roof sloping at the aforesaid angle. In this case the truncation or cut begins at a point 6'-0" (measured vertically) below the roof ridge. This starting point of the truncation was not fixed arbitrarily; it was customary in the eighteenth century to rest the short rafters 31 supporting this part of the roof on top of the end cross ties or collar beams and this in precisely what was done in this case.

The jerkinhead roof formed gable ends which were, roughly, half gable and half hip. It is readily apparent that if the end inclinations started at the cornice level we would have the common hipped roof.

PRECEDENT FOR THIS ROOF TYPE

There were many such clipped gable roofs in Tidewater Virginia in colonial times. The architects in this case, however, did not have to search far for their precedent, for the old Bland-Wetherburn Tavern in Block 9, a few houses west of the Purdie House, has this very type of roof.

GAMBREL ROOF OF TAVERN DIFFERENT IN CHARACTER; DISCUSSION OF THIS

The gambrel or "dutch" roof of the King's Arms Tavern is of a radically different order from the Purdie roof. The former is a "jointed" roof, the name "gambrel" being derived from the French "jambe", meaning leg. The joint in the Purdie roof occurs somewhat above the midpoint in the height of the roof. It is the meeting point of two planes of different slopes, the lower one being steep (75 ½°) and the upper one much flatter (31 ½°). The gambrel has the advantage over the A roof in that it makes the walls under the roof slopes almost vertical and, thus, provides a considerably greater of space in the second story of a house than that yielded, for the same plan area, by the A roof. The gambrel roofed dwelling is said to have been used, frequently, in lieu of a full two story house with A roof for the reason that the tax paid on a house in colonial times was in ratio to the number of stories under the roof. The gambrel roofed structure was classified as a story and a half house and fell into a lower tax bracket than a full story 32 house of the same floor area. Hence the popularity of the gambrel roof form which was widely used in Virginia in the seventeenth and eighteenth centuries.^* A number of old houses still exist, in fact, in which the second floor, as is evident from diagonal lines remaining in the brick ends, was expanded by the substitution of a gambrel for an A roof.

ROOF SLOPES OF KING'S ARMS ALSO BASED ON AVERAGE SLLOPES OF OLD EXAMPLES

Again, as in the case of the Alexander Purdie House, since no information relative to the roof slopes existed, the architects chose a lower and an upper inclination for the two parts of the roof which represented an average, more or less, of the slopes of a number of old gambrel roofs studied. The following table is a listing of several of the old gambrel roofs in Williamsburg with the approximate inclinations of their two surfaces:

It will be readily observed that the slopes of Tavern roof, 75 ½° and 31 ½°, respectively, are the averages for the five old roofs listed above.

SOUTH WING HAS GAMBREL ROOF WHICH CONTINUES ACROSS END

The wing which extends southward from the west side of the 33 south face of the Tavern is covered, like the main part of the buildings by a gambrel roof. This is a logical solution for the roof of this element since, to have joined an A roof with the double sloped gambrel of the main building at right angles would have been awkward. The wing is 20'-0" wide, 4'1-0" narrower than the main part of the Tavern. To gain space in the second floor room, which is a dining room, the lower slope of the gambrel was made somewhat steeper than the lower slope of the main roof; it is 81° as against the 75 ½° of the main roof. The upper slope, on the other hand, is 2° steeper than the 31 ½° of the main roof. The double slopes in contrast to the treatment at the ends of the main part of the building, continues around the end of the wing, making the roof, in effect, a Mansard.^*

PRECEDENT FOR MANSARD OF WING

The precedent for the hipped gambrel of the Tavern wing was the handsome Mansard type roof of the wing of Belle Farm, a colonial mansion in Gloucester County. The building has since been demolished but before this took place Colonial Williamsburg architects made measured drawings and photographs of it.

ROOFS OF APPENDAGES TO BE TREATED LATER IN TEXT

As for other, supplementary roofs not covered in the above discussion, such as dormer, porch and bulkhead roofs, etc. these will be treated under their respective headings in the course of the detailed discussion of the exteriors which follows.

ROOFS COVERED WITH CEMENT SHINGLES FOR FIRE PROTECTION

Both the main roofs and the roofs of all the all the appendages of 34 the Alexander Purdie House and the King's Arms Tavern are covered for reasons-of fire prevention with round butted asbestos cement shingles laid from 5 ½" to 6" to the weather, with a double starter course at the eaves. Round butted shingles are very frequently used in the eighteenth century. Whereas,it doubtless took more time to round the ends of wood shingles than to cut them straight, round butted shingles were more easily aligned and kept in alignment than were the square butted type. The above-mentioned wing of Belle Farm had this type of wood shingle when it was photographed in 1928.

CONCEALED FRAMING AND FOUNDATIONS MODERN IN CONSTRUCTION

Since it is the policy of Colonial Williamsburg to use modern materials and construction methods in the reconstruction of buildings and parts of buildings when the elements are involved are concealed from view, the wood framing of King's Arms Tavern and the Alexander Purdie House follows modern rather than eighteenth century practice. This holds true also in the construction of foundations, a subject to be discussed later (p. 36).

WALLS OF BOTH BUILDINGS COVERED WITH BEADED WEATHERBOARDS

The main walls of the two buildings with the exception, mentioned on p. 29, of the east end wall of the Tavern which is of brick, are covered with beaded weatherboards of Red Gulf cypress, with an overlap such that each board is exposed from 5 ½" to 6 ¼" to the weather. The boards are secured to the studding by galvanized iron nails with hand hammered heads. Copper nails with hand hammered heads are substituted for iron nails at the corners of the building and around door and window openings, where copper flashing occurs. The use of copper nails in the locations specified above was a precaution against the chemical reaction which takes place when copper and iron are brought into contact with each other.

TERRA COTTA NOGGING

Certain of the wood walls of the King's Arms Tavern and the Alexander Purdie House have between the studding and back of the weatherboarding of the exterior side and the plaster on the interior face an infill of 3" thick hollow terra cotta blocks. (See detail drawings above and on other side). These have, as their chief purpose, the retarding of the spread of fire, from one portion of the two buildings to another; from the inside of the buildings to the outside (protection of neighboring buildings) and from a neighboring building to one or the other of the two structures. The nogging has been placed only in those walls which stand most directly in the path of a fire originating within the two structures and those (the west wall of the main part of the Tavern and east wall of the Purdie House) which are most vulnerable in the event of a fire in the Barber Shop and the John Coke Office, respectively (see plan, p. nogging locations). (Over)

The use of solid brick as nogging was not uncommon in Virginia in the eighteenth century and was found in the exterior walls and one interior wall of the St. George Tucker House and in the north and west walls of the northwest first floor room of the Brush-Everard House as well as in the north wall of the old north wing of that house. These old walls with brick filling the spaces between the members of the wood framework resemble English half timber work except that in the latter the brick was generally stuccoed over while in the Virginia examples it is covered with weatherboarding.

We are uncertain as to the precise reasons for the use of brick nogging in weatherboarded walls in the eighteenth century. The brick may have been placed in the walls for its insulating value and possibly as a means of retarding the spread of fires. Whatever may have been the reason for its use, there is plenty of colonial precedent for employing it in the reconstructed buildings of Williamsburg.

BEADED WEATHERBOARDS TYPICAL WALL COVERING IN COLONIAL WILLIAMSBURG

Weatherboarding with the lower edge shaped to the form of a bead with a diameter of about ½" was the most common covering in Williamsburg in the eighteenth century. Old weatherboarding of this type existed on the Tayloe House, for example, down to the time this building was restored, though a great deal of it had to be removed because of its poor condition. Although the age of the Tayloe weatherboarding is uncertain since such wood materials are subject to periodic replacement, there is little question that in type they go back to the eighteenth century, inasmuch as building usages are apt to be perpetuated when such replacements occur.

APPENDAGES HAVE OTHER BOARDING

The wood appendages, such as porches, dormers and the attached well house are covered with types of boarding other than weatherboarding. This boarding will be discussed in every case at the time each of these elements is treated.

FOUNDATION BRICKWORK IS ENGLISH BOND; EAST WALL OF TAVERN AND CHIMNEYS ARE FLEMISH

The King's Arms Tavern and the Alexander Purdie House rest on masonry foundation walls which at the west end of the former building rise about 2'-5" above the ground level and at the east end of the latter, due both to the eastward slope of the ground and to the difference in the first floor levels of the two buildings, about 3'-4" above grade. The brick of these foundations are laid up in English bond which in eighteenth century Virginia was commonly used below the watertable in brick structures and for the foundations of wood buildings. On the east end wall of the Tavern, which is of brick masonry throughout, the bond changes above the watertable to Flemish. The engaged east chimney and, for that matter, the other two chimneys of the Tavern as well as the two of the Purdie House are also laid up in Flemish bond, except for certain of the chimney faces above the roof level which are too narrow to permit the carrying 36 out of the Flemish bond pattern. These narrow sides of the chimney are laid up in running or stretcher bond.^*

Lest the above statement that the foundations are laid up in Flemish bond be misinterpreted it should be made clear that the brickwork of the foundations of the two reconstructed buildings is merely a 4" facing of brick varying in height from 2'-6" to 4'-6" more or less, which rests upon a ledge provided for it in the reinforced concrete walls which constitute the actual foundations. The variation in the height of this brick facing is due to changes in the grade level at different points of the two buildings, the brickwork having been made at all points high enough to mask the concrete foundations. The latter walls (actually basement walls and foundations) with their footings vary in depth because of variation in the levels of the basement floor. At their deepest they descend about 11'-6" below the level of the finished first floor.

LOOKING SOUTHEAST TOWARD PURDIE KITCHEN

LOOKING EAST TOWARD JOHN COKE OFFICE

TWO VIEWS OF THE REINFORCED CONCRETE FOUNDATION WALLS AND COLUMNS OF THE KING'S ARMS TAVERN AND THE ALEXANDER PURDIE HOUSE. The trenches in the basement are for pipes. The rectangular openings at the upper edge of the walls are to receive basement windows. The squarish concrete slab at the right in the lower pictures and right center in the upper now covers a wine room which has a dropped ceiling.

BASEMENT AND FIRST FLOORS ARE CONCRETE SLABS; SUPERSTRUCTURE IS OF WOOD

The basement floors are 4" concrete slabs resting on cinder fill. The structural first floor is a 4" thick reinforced concrete slab upon which wood sleepers are laid to receive pine under flooring. This, in turn, carries the finish flooring which consists of old salvaged boarding. Above the level of the finished first floor the structure, i. e., studding, floor joists and roof rafters, is of normal contemporary wood construction.

BRICK USED IN RECONSTTRUCTION WORK MADE BY OLD METHODS; SIZES, IN CASE OF TAVERN ARE SIMILAR TO THOSE OF OLD BRICKWORK

The brick visible to the eye which was used in the reconstruction of the foundation walls and chimneys and other brick parts of the Tavern and Purdie's House are handmade brick made by Colonial Williamsburg, Incorporated in the colonial manner in its own kilns near Williamsburg. An is entirely reasonable since the two buildings were originally erected by different people at different times, these brick are not alike either in size or color. In the case of the Tavern, for example, the brick are not alike either in size or color. In the case of the Tavern, for example, the brick of the older, eastern part of the foundation measured approximately 9 ¼" x 4 ½" x 2 5/8" while that of the later, western part averaged 8 ½" x 4" x 2 ¾". The bond in each case was English and the mortar of the oyster shell variety current in Virginia during the eighteenth century. The handmade brick used as facing in the reconstruction of the Tavern walls vary in length from 8 7/8" to 9 1/8", in width from 4" to 4 1/8" and in thickness from 2 ½" to 2 5/8", six courses measured from edge of brick to edge of brick totaling about 17 ¼", so that it is evident that these were chosen as an approximate average of the dimensions of the two old brick types of the original foundations.^*

OLD FOUNDATIONS OF THE KING'S ARMS TAVERN

The plan drawing above shows the foundations on colonial lot #23 which have been identified as those of the King's Arms Tavern. Although existing records indicate that a wing extended southward from the rear of the building, no foundation evidence for this was discovered, which suggests that the foundations may have been superficial in character.

The rectangular house foundation consisted of brickwork of two periods, the older being that of the east portion, which must have constituted the original structure. The squarish west part of the foundation consists on three sides of later brickwork. The cellar here was not so deep as that of the older part of the building and steps connected the two. The remains also indicate that the more recent basement was entered from the outside by means of a flight of brick steps at the southwest corner.

The east part of the cellar was "roofed" with a barrel vault, the remains of which were found (see section drawings at left). The brick piers apparently served as support for a flat ceiling erected after the barrel vault had fallen. The 4" lining of brick inside of the north and south walls was composed of smaller brick not bonded to the brickwork of the main walls so that it was evidently added later, possibly to support floor joists.

SECTION OF FALLEN EAST WALL OF TAVERN. ITS HEIGHT, ADDED TO THAT OF THE END WALL FOUND IN PLACE, WOULD HAVE RAISED THE TOP OF IT 7'-0" ABOVE GRADE. THIS IS GOOD EVIDENCE THAT THE EAST WALL WAS OF BRICK.

VIEW OF SOUTH WALL OF EASTERN PART OF BASEMENT, THE BRICKWORK FORMING THE SPRING OF THE BARREL VAULT IS VISIBLE AS WELL AS THE TOP FO THE 4" SUPPLEMENTARY WALL WHICH WAS PORBABLY ERECTED, AFTER THE VAULT FELL, TO SUPPORT FLOOR BEAMS.

BASEMENT ENTRANCE STEPS AT SOUTHWEST CORNER, LOOKING WEST. THESE DETERMINED THE LOCATION OF THE PRESENT BASEMENT STEPS AND BULKHEAD.

SHALLOW WEST PART OF BASEMENT, LOOKING SOUTH. THE STEPS SHOWN ABOVE CAN BE SEEN AT THE JUNCTION OF THE WEST AND SOUTH WALLS. THE TRIANGULAR PATCH OF BRICKWORK IS PROBABLY THE BASE OF A CORNER CHIMNEY.

LOOKING NORTHWEST TOWARD OPENING IN WALL BETWEEN TWO BASEMENTS SHOWING PAVING OF A PERIOD LATER THAN THE FIRST. WHEN THE PAVING WAS REMOVED THE WALL SHOWN IN THE PICTURE BELOW WAS REVEALED.

LOOKING WEST AT WALL SEPARATING DEEP EAST PART OF BASEMENT FROM THE SHALLOW WEST PART THE TOOTHING PROBABLY SERVED AS ANCHORAGE FOR STEPS. THE BRICKWORK OF THE WALL IS CLEARLY SEEN TO BE ENGLISH BOND.

SIX VIEWS OF OLD FOUNDATIONS OF KING'S ARMS TAVERN

IN CASE OF PURDIE HOUSE NEW BRICK DEPARTS FROM SIZE OF OLD

In the case of the Alexander Purdie House the brick chosen for the reconstruction work varied from 8 1/8" to 8 ¼" in length, 3 5/8" to 3 ¾" in width, and 2 3/8" to 2 5/8" in thickness and six courses measured from brick edge to brick edge totaled about 17 ¼". This is a departure from the brick sizes found on the site. As a matter of fact, however, only two wall fragments, one about 10'-0" in length and the other approximately 3'-0" long were found and the average size of the brick in these was 9" x 4" x 2 ½". The only other actual brickwork discovered was the remains of three or four basement steps and here the brick averaged 8 ¾" x 4 ¼" x 2 ½" in size. It is likely that these sizes were not taken into account because of the fragmentary character of these foundation remains. The bond of the old foundation brickwork was English, as in the case of the Tavern foundations, and the mortar also of the oyster shell type. In the reconstruction work on both projects a mortar was used which simulates the color and texture of the old mortar.

KING'S ARMS BRICKWORK LIGHTER IN TONE THAN PURDIE BRICK

There is a considerable difference between the color of the Tavern brick and that of Purdie's. The former is generally lighter in tone than the latter. It varies in color from a salmon pink to 42 OLD FOUNDATIONS OF THE ALEXANDER PURDIE HOUSE
Relatively little brickwork appertaining to the old Purdie foundations remained (see plan above) although the position of the basement steps was indicated. Burnt brick fragments and wood ashes suggested the location of the east inside fireplace and chimney. Most of the external foundation walls of the house had disappeared and only a thin layer of oyster shell mortar and brick fragments remained to indicate the extent of these. The length and width of these remains, however, agreed substantially with the dimensions of the house (64'-0" x 26'-0") given on an insurance plat of 1909 and they were followed in locating the reconstructed building. 43 an ox blood red and includes some dark purples. All of these colors are somewhat modified or muted since the surfaces are "peppered" with particles of sand. There are some partially glazed headers. The Purdie brickwork contains a few salmon red brick but the predominant color is a purplish brown -- sometimes almost a blackish red. There is very little glazing on these brick.

BRICK GUTTERS & DRIPS TO BE TREATED HERE

To complete the general treatment of brickwork (certain items will be discussed in the course of the detailed treatment of the elevations and interiors) mention should be made of the brick gutters on the front (north) side of the two buildings and the brick drips or "spatters" on the rear (south) side of the main buildings and the wing. The plot plan (p. 3) should be consulted for the exact location of these features.

LOCATION OF GUTTERS AND THEIR DESCRIPTION

The gutters adjoin the building foundations, the Tavern having one only, between the front porch and the west end of the building and Purdie's having two lengths flanking its front porch. In each case they are made up of three rows of brick laid with their bedding surfaces level with the ground and an outside border of brick stood on edge. Of the flat-lying rows the two outside ones are roughly perpendicular to the face of the building and incline toward the center row in which the brick parallel the building. The gutters slope with the street slope, i.e., from west to east and each is provided with a drain at its east end.

BRICK DRIPS, THEIR LOCATION & DESCRIPTION

The drips are not designed to lead off the rain water which drops from the roof eaves but only to break its fall and prevent it from pitting the ground adjacent to the house. They consist at some points of a single row of brick lying with their bedding faces 44 flush with the surface of the ground and their long sides perpendicular to the building and in other locations of a double row similarly placed. The drips are not needed, of course, on the gable ends of the buildings but only on the sides toward which the roof planes incline. This means that they are found on the south side of the main building complex and would also be present on all three sides of the wing and Well House, since these two elements have roof slopes on all sides, were it not for the fact that brick paving runs to the foundation walls on all sides, except for a length of about 4'-0" on the east side of the south face of the wing. Here a planting strip strikes the building and makes it necessary to use a drip. In some cases the drips touch the foundation walls and in others they are a few inches removed from them.

COLOR AND SIZE OF BRICK OF GUTTERS AND DRIPS

Both the drips and the gutters are made of new handmade brick of a brownish red color. The length of these brick varies from 8 7/8" to 9"; width from 4" to 4 1/8" and the thickness from 2 3/8" to 2 ½".

PRECEDENT FOR USE OF GUTTERS AND DRIPS

Whatever gutters and drips may have existed about the two buildings had disappeared when the time for their reconstruction came, which, considering the fact that these features rest on the ground surface, is not remarkable. Remnants of old gutters and drips have nevertheless been found near old buildings in Williamsburg. An example of the former is the length of old gutter which was discovered adjacent to the north wall of the main part of the Wren Building, and of the latter, the brick drips found along the north and south sides of the Tayloe House.

KING'S ARMS TAVERN -- GAMBREL SECTION, SHOWING FRONT DORMER

KING'S ARMS TAVERN WING -- GAMBREL SECTION SHOWING DORMER

KING'S ARMS TAVERN, EXTERIOR
DETAILED ANALYSIS^*
NORTH ELEVATION

SHINGLES

Both slopes of the gambrel are covered with round-butted, asbestos cement shingles laid 5 ½" to 6" to the weather, with a double starter course at the eaves. The same type of shingle, laid in the same manner, has been used on the roof of the front porch and on the dormer roofs.

DORMERS & DORMER WINDOWS

There are five dormers on this elevation, centered over the four windows and the doorway of the first floor. Since there is no regularity in the spacing of the first floor openings, the dormer spacing is also irregular. The dormers are of the pedimented type, which is to say that the crown mold of the window enframement is carried along the upper edges of the triangular tympanum formed above the window by the sloping planes of the A roof. The face of this tympanum is formed by flush boarding.

SHEATHING AND MOLDING OF DORMER SIDES; UPPER CORNICE OF BUILDING

Dormers are actually "light boxes," designed to admit daylight to the roof sides of the second story of a building. The dormer sides are closed and, since they are cut off on a diagonal by the sloping surfaces of the roof of the building, they are triangular. In the case of the Tavern dormers, due to the steep slope of the lower "leg" of the gambrel, the inverted base of the triangular side is very short, being only about 18" long. These dormer sides are covered with random width flush boarding which parallels the slope of the adjacent face of the Tavern roof. A transition between the dormer sides and the overhanging roof covering (asbestos cement shingles) is effected by the use of a crown mold similar to that which caps the window enframement and runs up the sloping pediment sides. This crown mold consists of a cyma recta, a fillet and a cyma reversa. The molding when it strikes the building proper "turns the corner' and continues along in both directions between the lower edge of the upper gambrel slope and the upper edge of the lower, to form what is known as the "upper cornice."

MOLDINGS OF WINDOW ENFRAMEMENT

The moldings forming the stiles of the window frame are the same as those at the head except that the cyma recta which tops the latter is missing. These moldings, "reading" from the outside in (toward the sash) are: a cyma reversa (ogee) backband, the outside edge of which is beaded; a fascia and a bead. The total width of these jamb moldings is 6". The sill is of the square cut variety and runs the full width of the dormer, 3'-4". The front face of the sill is 2 ½" high.

WINDOW SASH AND GLASS

The window is of the double-hung type, five lights high and three wide. These lights are distributed between the upper and 48 lower sash so that there are six lights in the upper in the sash and nine in the lower. The glass pane size is 8" x 10 3¾". The glass used throughout is salvaged antique glass. It should be added that both sash operate with the aid of metal counterweights. The window frame and sash are joined by mortise and tenon joints secured by hardwood pegs.

COLONIAL PRECEDENT FOR VARIOUS DORMER FEATURES

The design of the various domer parts just described is derived from comparable old features observed in dormers of eighteenth century houses in Williamsburg and elsewhere in Virginia. Sufficient precedent for the pedimented front was furnished by the old dormers of this type on the Lightfoot and Powell-Hallam Houses and by the three south dormers of the Tayloe House. The sides of the latter dormers are covered, like those of the Tavern, with diagonally placed flush boarding, the angle of inclination of which is the same as that of the adjacent lower leg of the gambrel roof. The Tayloe dormers likewise have sash five lights high by three wide with six lights in the upper sash and nine lights in the lower. The molding sequence of the window architraves or frames (ogee, fascia, ogee, fascia and bead) is a very common one in old Williamsburg houses of the eighteenth century, where it is used, both inside and out, for door and window frames. A number of old door and window architraves of this type are found in the Tayloe House.

MAIN CORNICE OF MODILLION TYPE

The main cornice of the Tavern is of the more ornate type of Williamsburg and Virginia cornice which, in addition to the usual moldings, was decorated with modillion blocks. This cornice was used not alone upon the north elevation but also on the three wing elevations.

ENUMERATION OF MOLDINGS COMPOSING CORNICE

The cornice is a relatively large one having a height and width, that is, projection from the wall, of slightly more than 15". Its component elements, listing them from the top down, are a crown molding consisting of an ogee, fillet and cyma reversa; a fascia projecting 8" from a second recessed fascia and a base mold the main elements of which are an arc of a circle, a fillet and a cyma reversa.

The modillion blocks and the soffit to which these are attached by nailing from above were omitted in the above enumeration. The modillions are rectangular blocks placed 9 3/8" on centers, the end ones on the north elevation being applied against the cornice stops or end boards. At the corners of the wing, from which a section of the cornice runs across the south end as well as along the sides, modillion blocks were placed at an angle of 45°, to effect a transition between the bands running at right angles to each other. A 5/8" high cyma reversa mold runs along the top edges of the blocks covering up the joints between the blocks and the soffit board and this continues along the top of the recessed fascia board, again hiding the joint between the latter and the soffit. Thus the blocks are joined by a continuous ornamental band.

DIMENSIONS OF BLOCKS

The blocks themselves, including the molding which caps them, are 3 7/8" high with a length and width respectively of 6 ¼" and 2 3/8" , not counting the projection of the top molding.

OLD EXAMPLES OF CORNICES WITH MODILLION BLOCKS

This cornice is like the old cornice on the north and south elevations of the Tayloe House in all particulars except in the size and spacing of the blocks. Other old cornices which might be cited for precedent are those of the Powell-Hallam House (this has been largely renewed); of the Mansard-roofed wing of Belle Farm 50 in Gloucester County and of Wales near Petersburg.

ROOF "KICK"; DESCRIPTION OF THIS AND OLD EXAMPLES OF IT; ITS FUNCTION

Attention should be called to a somewhat unusual feature which, though it is part of the roof treatment, is so closely associated with the cornice that it will be examined there. On all elevations of the Tavern where the cornice is found (i.e., north elevation and the wing elevations) the roof, by means of blocking starting at about 3'-0" above the bottom edge, is made to curve outward and upward so that, at the level of the top of the cornice, the sheathing which bears the shingles stands out about 4" from the undeviating diagonal of the roof rafters. This feature, known as a "kick" is found on both the north and south elevations of the Tayloe House, except that in this case it does not carry across the entire facade but, instead, is subtly merged again with the straight slope as it proceeds from either side toward the middle. An old example of a kick which carries across an entire facade is the one found on the roof of the Powell-Hallam House. In this case the kick does not, as in the Tavern roof, take the form of a suave, gradual curve but is, rather, formed by the meeting of two straight roof planes of different angles of inclination. It should be said, in conclusion, that the purpose of the kick was, undoubtedly, an aesthetic one -- to soften the otherwise unrelieved straight slope of a roof.

GUTTERS

See pp. 43 - 44 for treatment of these.

WALL COVERING

See pp. 34, 35 for treatment of this.

FRONT PORCH; GENERAL DESCRIPTION OF THIS

A comparison of the front (north) porches of the King's Arms Tavern and of the Alexander Purdie House will reveal that the two are very unlike in design. Since the Purdie porch will be described 51 in detail later it will be sufficient to note here tat it is shed-roofed and possesses characteristics to which one can refer as Virginian. The King's Arms porch, on the other hand, is definitely classic (Roman) in derivation. It is a pedimented A-roofed member, the roof of which is supported on two-round-sectioned columns in front and two pilasters on the wall side. The porch platform is elevated about 2'-9" above the grade level and is approached from either side by means of four steps. A wrought iron handrailing with balusters starting on the lowest step ascends on either side to terminate against the column.

DIMENSIONS OF PORCH

As for some of the major dimensions of the porch: the platform is about 9'-0" in length and 7'-3" wide. The column height is 6'-7" and the top of the cap about 4'-7 ½" below the ridge of the porch roof. The slope of the roof is about 28°. Its lower edge lines up with the bottom of the main cornice of the house, so that the roof interrupts the cornice.

DETAILED DESCRIPTION; FOUNDATION AND STEPS

The porch floor and the flanking steps are supported on three concrete foundation walls - on the north, east and west sides. The structural floor and steps, themselves, are of reinforced concrete. The north foundation wall is given the appearance of brickwork by a 5" facing of hand made brick, which continues about a foot below grade. The steps are solid stone blocks which, with the facing brickwork of the supporting wall, form the part of the stairs which is visible to the eye. The blocks are of "rustic gray" Indiana limestone, a likely substitute for the Portland stone which might well have been imported from England in the eighteenth century for use here. The steps have been carved on their two exposed sides to 52 PHOTOSTATS OF TWO PLATES IN PALLADIO LONDINENSIS OR, THE LONDON ART OF BUILDING BY WILLIAM SALMON, THIRD EDITION, 1748. THIS HANDBOOK WAS KNOWN TO HAVE BEEN USED IN VIRGINIA IN THE EIGHTEENTH CENTURY. NORTH PORCH OF TAVERN. THE DRAWING, RIGHT ABOVE, WAS PRECEDENT FOR THE DESIGN OF THIS PORCH. THE DOOR WAS BASED ON THE DRAWING, LEFT ABOVE AND EXAMPLES SUCH AS TWO DOORS OF WALES IN DINWIDDIE COUNTY. 53 form a nosing which, running across the end as well as the long side, returns upon itself, giving an effect of symmetry to the tread end.

STEP NOSING AND ITS PRECEDENT

The nosing profile consists of a half round at the top followed by a fillet and a cove, the three moldings together forming an ornamental member about 3 ½" in height. This sequence of stone step moldings is an oft-recurring one in Virginia architecture so that the precedent for it is not difficult to find. The two sets of old steps, reputedly salvaged from the first theatre site, and which now stand before the John Blair House have nosing profiles similar to those used on the Tavern steps, as do the old steps before the President's House of the College of William and Mary. In the latter example the moldings are of about the same "weight" and proportion as those of the Tavern steps whereas the half round or cushion of the Blair House steps is somewhat heavier. It should be noted also, that the moldings of the President's House steps continue across the free end and return on themselves, as in the case of the Tavern steps.

PORCH PLATFORM WITH STONE BORDER AND BRICK CENTER

A border of the same limestone used for the steps has been carried about the three free sides of the porch platform. The steps are the same height (7") as the steps and are 1'-1" wide on the street side and 1'-5" on the two sides toward the steps. The stones are set in beds of non-staining white cement mortar. Where two stones are brought together end to end, either on the platform or in the case of the steps, corresponding notches are cut in these ends to receive wrought iron cramps. The cramps are set in molten lead.

The platform surface within the stone border is made up of 8" x 8" brick tiles, 1 ¾" thick. These are set in cement mortar placed an the concrete fill which in turn rests upon the reinforced concrete slab forming the structural support for the finished platform. An old example in Williamsburg of a porch such as that of the Tavern with a stone and brick tile floor, is the porch near the western end of the south front of the Coke-Garrett House.

PORCH SUPERSTRUCTURE IS MODIFIED ROMAN DORIC, SUBSTITUTING DENTIL BAND FOR MUTULES

We have in the columns, entablature and pediment of the porch a modified or mixed Roman Doric order, one in the case of which a certain element, the dentils of the cornice, do not belong, typically, to this order, but rather to the Ionic and Corinthian styles. The component parts of the Palladian Roman Doric order are shown in our photostat (p. 55) of a plate from the third edition of Palladio Londinensis^* by William Salmon, published in London in 1748. A second plate from the same book (our p. 52) shows this order in use in the design of an entrance enframement. It will be noted that in the latter plate the columns are fluted. In both cases, in conformity with classic usage, mutules (see first plate) are used above the triglyphs of the frieze and on the soffit of the sloping cornice members of the pediment, rather than dentils.

PRECEDENT FOR THIS MIXTURE OF STYLES

It will be discovered, if a number of old examples are consulted, that in Virginia and elsewhere in the eighteenth century for the use of the various elements of the several classic orders (RIGHT) ROMAN DORIC ORDER AS PRESENTED BY WILLIAM SALMON IN PALLADIO LONDINENSIS, A BUILDER'S HANDBOOK WIDELY USED IN THE 18TH CENTURY. THIS WAS THE ACCEPTED FORM OF THE ORDER. THE DESIGN OF THE NORTH PORCH OF THE TAVERN, WITH DENTILS REPLACING THE MUTULES CUSTOMARILY USED WITH THE TRIGLYPHS, REPRESENTS A MIXED FORM OF THIS ORDER. THIS SUBSTITUTION OF AN IONIC DETAIL FOR A DORIC ONE WAS, NEVERTHELESS, BASED UPON PRECEDENT NO LESS AUTHORITATIVE THAN THAT PROVIDED BY THE WEST PORCH OF GUNSTON HALL IN FAIRFAX COUNTY. THE DRAWING ABOVE WAS BASED UPON THE DESIGN OF THE OLD PORCH IN QUESTION. 56 were pretty generally respected. There were, however departures from them and it is from one of these that the present mixture of styles was derived. We present, on p. 55, a reproduction of a detail drawing made of a portion of the north porch of Gunston Hall built by George Mason in Fairfax County in 1758. It will be noted that in this example of a Roman Doric entablature, as in our own, dentils have been substituted for the mutules used normally with the triglyphs and other features of the Doric style.

BOTTOM THIRD OF COLUMNS & PILASTERS IS UNFLUTED; REASONS FOR THIS TREATMENT

Another detail worthy of comment is the treatment of the lower third of the columns and pilasters. This has been left unfluted to a point about 3'-0" above the porch platform. The treatment here was probably dictated by the circumstance that the iron and wood handrailings were to be fastened to the columns and a plain surface would render the connection both easier and more satisfactory aesthetically. Whatever was done in the case of the columns would naturally be reflected in the pilasters, though no railing was to strike them.

COLUMNS OF BARTON MEYERS PORCH WERE MODELS FOR THOSE USED ON TAVERN

Roman Doric columns such as those described above in which the lower third, approximately, of the shaft is smooth and the remainder fluted^* were found by Singleton P. Moorehead on the east porch of the Barton Meyers House, a late eighteenth century dwelling in Norfolk. 57 The two pilasters against the face of the brickwork of the building are treated in the same manner. This porch is otherwise quite different from the Tavern one, having a barrel-vaulted hood and a pediment in which the horizontal cornice member is broken and the tympanum missing.

The general principle involved, i. e., adding variety or richness of effect to the column or pilaster by treating the bottom of the shaft differently from the upper part is, on the other hand, illustrated in a considerable number of colonial Virginian examples and others. Most frequently, in cases of this sort, when the column is fluted from cap to base the flutes of the lowest third, approximately, have, carved in them, elongated convex features known as "reeds", which may be plain or decorated. (See our reproduction of William Salmon's Plate XXIII, p. 52)

STRUCTURAL DETAILS OF THE COLUMNS

The columns of the Tavern porch are hollow, being set together of 8 wood sections 1 ½" thick. The 3" high bases which rest on the porch platform are made of cast iron. Although cast iron was not used for column bases in the eighteenth century, it was necessary to employ it here in the interest of permanence, wood bases being subject to rotting. The top molding of the column capitol is covered with metal for the same reason.

IRON PORCH RAILING; DESCRIPTION OF THIS

The wrought iron railing which ascends the steps at either side of the porch consists of a handrail, circular balusters and an octagonal newel. The handrail is made up of two parts, a top element which is a half ellipse in section, with a 1 of 1 ½" and a short one of a trifle over 3/8". This rests upon a rectangular bar ¾" x 7/8" in section, to which it is screwed. This bar is supported by the balusters and the newel which are 58 tapped^* into it. The top bar is bent upward at the column and screwed to it. The elliptical hand rail is flattened out over the newel to a shape which in plan is a nearly complete circle, with a diameter of about 2 ½". Upon this rests a brass finial, 7 ½" high, which is turned to a typical vase shape. The finial is held to the newel beneath by means of a threaded dowel which extends through the rail into the newel. The newel is octagonal from the top to a point within 8" from the bottom where it becomes square. The octagonal part of the newel tapers from a width, measured from flat side to flat side, of ¾" at the top to 1 1/8" at the bottom. The transition from the octagonal cross section to the square is effected by means of four concave curves.

OLD PORCH RAILINGS WHICH GAVE PRECEDENT FOR THIS RAIL; THOSE OF PRESIDENT'S HOUSE AND CIVIL WAR PHOTOGRAPH

The north porch of the President's House of the College of William and Mary has an old stairrailing with a handrail hammered to the shape of a truncated ellipse and is, therefore, of the same type as the member of our handrail. It also has octagonal newels with square bases similar to those of the Tavern handrail. The balusters are rectangular rather than circular in section, however. A wrought iron stairrail with a handrail half elliptical in section is also clearly visible in a much enlarged Civil War photograph (#B-56, Colonial Williamsburg Office of Research and Record) of the wood porch and stone steps of an unidentified colonial mansion. Here, again, the balusters appear to be square or rectangular in section rather than circular.

STAIRRAIL IN HARWOOD-HAMMOND HOUSE IN ANNAPOLIS

A stairrailing which approaches our railing more closely than either of the above-mentioned examples, apparently is one in the Harwood-Hammond House in Annapolis, Maryland. The handrail in this case is closely similar to ours both in the dimensions and the character of the parts. It has both the half-elliptical top rail and the square supporting bar beneath this. The balusters of this railing, furthermore, are circular in section with the same ¾" diameter as the balusters of the Tavern railing.

CHARLESTON EXAMPLES

There are furthermore, numerous porch railings of this same general design in the colonial sections of Charleston, though the balusters seem to be square-sectioned in the main. Many of these railings terminate at the bottom in a baluster surmounted by a brass finial of one design or another, related in ornamental purpose and character to our finial.

WOOD RAIL BETWEEN POSTS OF PLATFORM

The horizontal rail suspended between the two posts of the platform is, as we have said, of wood. It has two horizontal members, a richly molded handrail and a plainer bottom rail. These support between them a series of balusters, rectangular in section. The height of the rail in about 2'-4 ¼" and the top of the hand rail in 2'-10" above the surface of the platform.

HANDRAIL PROFILE IS EXCEPTIONAL IN TYPE; ITS PRECEDENT

The handrail, symmetrical about a center axis, has a profile which, starting at the top with a cyma curve, passes successively to a half round, a cyma reversa and a bead, with the usual fillets, etc. between these moldings. It belongs to a genus of colonial handrailings which were used frequently in eighteenth century Williamsburg and Virginia. A very characteristic profile in this handrailing series is one which is similar to that of the Tavern porch in everything except the part lying between the half round 60 THE HANDRAILING PROFILE ON THE LEFT IS THAT OF THE RAILING OF THE NORTH PORCH OF TH EKING'S ARMS TAVERN. THE ONE ON THE RIGHT WAS COMMONLY USED IN VIRGINIA IN THE EIGHTEENTH CENTURY. THE OLD HANDRAIL OF THE MOODY HOUSE STAIR HAS THE LATTER PROFILE. THE HANDRAIL OF THE ORIGINAL STAIR OF THE BARRAUD HOUSE WAS ALSO SIMILAR TO IT. and the bead at the bottom of the rail. This part is made up of a cove at the top which merges into a fascia instead of the S-curve of our handrail. We find this latter cyma recta curve used frequently in architraves, cornices, stairrailings and other trim features but instances of its use in Virginia in a wood handrail of this type are difficult to find. The King's Arms handrail was derived, an a matter of fact, not from a colonial example but from a handrailing in Swan House, Chichester, England. This is reproduced on p. 24 of The Architectural Reprint which took it from Belcher and Macartney's The Later Renaissance Architecture in England. Other examples are to be found in English Interior Woodwork of the 15th to 18th Centuries by Henry Tanner, Jr. and in English books on interior details of the seventeenth and eighteenth centuries.

BOTTOM RAIL OF VARIANT OF A FORM WIDELY USED IN VIRGINIA

The lower rail of the porch handrailing is another variant of that frequently used type of eighteenth century railing which, in 61 section, in an elongated rectangle, beaded on either side top and bottom. These rails generally have a horizontal lower surface and a flat or rounded upper one. An old stair with such rails exists in the Nicolson Shop. The departure from the normal detail in our railing consists in the fact that the top part has been made wedge-shaped in order, probably, to shed rain water better, this being an outside railing. A detail of this sort was found on the bottom member of the railing of a porch of Bel Air in Charles City County and another on a railing of St. John's Church, King William County. The overall height of our rail is, by the way, about 3 ½" and its width 1 13/16".

RECTANGULAR BALUSTERS A FAMILIAR VARIETY IN COLONIAL ERA

The balusters, spaced about 3 ½" on centers, are rectangular in section, with a long dimension of 1 3/8" and a short one of a little under 7/8". This baluster type was very common in the eighteenth century as it is today. An old example of it is the series of them used in the porch railing of Bel Air, mentioned above.

USE OF TWO RAILING MATERIALS ON PORCH IS UNORTHODOX DESIGN

It may seem strange, at first glance, that both wood and iron railings should have been used on the front porch of the King's Arms Tavern. It would have been more in conformity with usual design practice to have used a single railing material throughout. Iron, indeed, might have been used between the columns, as well as on the steps, since many old railings of this sort can be found. It would have been less feasible, on the other hand, to have used a wood railing on the steps for the supports would have rested on stone and been subject to rotting.

IN SOME CASES THE TWO RAILS ARE OF DIFFERENT PERIODS; PRECEDENT FOR USE OF TWO MATERIALS IN TAVERN RAILINGS

Combinations of two railings of different materials have been found, however, and we may assume that in certain cases, at least, they represent two periods of building. Many flights of outside 62 steps of a height which we, today, would consider dangerous without guard rails, had no railings at all in eighteenth century Virginia. Examples of this are the four sets of stone steps at Tuckahoe in Goochland County, near Richmond which are high enough to be hazardous and which, nonetheless, have no railings to render them less so. In many cases like the above, probably, later residents provided railings which were absent originally. Had this been done to the south entrance steps at Tuckahoe we would have an example of mixing railing materials on a porch since the platform has a wood railing. The actual porch with railings both of wood and iron which furnished the precedent for this usage in the north porch of the Tavern was a porch shown in the Civil War photograph, #B-56, mentioned above. It was believed by the architects that this house with its porch stemmed from the eighteenth century.

PORCH CEILING IS RECESSED AND PLASTERED

The porch ceiling is recessed behind the wood entablature of the order and is raised about 11" above the top of the column caps. It is flat and plastered. A crown mold composed of a cyma recta, a fillet and a cyma reversa covers the joint between the plaster and the wood fascia forming the back of the entablature. On the door side this crown mold is stopped by the stiles of the door frame.

The plastering of porch ceilings or soffits was not unusual in the eighteenth century. The old porch at the west end of the south front of the Coke-Garrett House has a plastered ceiling, although this is curved.

PORCH LIGHT

Lighting for the entrance is provided by a reproduced colonial type lantern, square in plan, with metal frame and glass sides and containing a simulated candle wired for electricity. This light is suspended from the center of the porch ceiling. It is designated as 63 being from Colonial Williamsburg Warehouse stock which is marked "Chowning's Tavern."

PERIOD REPRESENTED BY PORCH IS LATE EIGHTEENTH CENTURY

Before leaving the subject of the King's Arms porch, a word should be said about the period it is intended to represent. It is well to recall here that the first owner of lot #24 of whom we have a record was James Sheilds who sold the property to William Byrd II in 1707 (p. 21). It is possible and, indeed, likely that Sheilds had built a house upon the lot since, otherwise, by the provisions of the Act of 1705^* he would have forfeited his right to it. So, the east portion of the original house may have stemmed from early in the eighteenth century. The foundations indicate that an extension was made toward the west at a later date (pp. 39, 40). We have reason to assume, since she made many alterations to her property, that Mrs. Vobe added the south wing and that she probably also built or rebuilt the front porch during her occupancy of the property (1772-1787). The design period of the porch was chosen on this assumption. It is, for Virginia, late eighteenth century in character since it may, in general, be stated that the exterior use of classic motives as they are employed here, stems from that period.

ENTRANCE DOOR; ITS DIMENSIONS

The main entrance door is located on the center axis of the porch and is thus sheltered by the porch roof. It is 7'-0" high, 3'-6" wide and 2" thick and it is panelled both sides. It has a four light fixed transom above it of the same width as the door and 1'-2 7/8" high. The glass panes are 9" x 11" in size.

DOOR IS "CROSS PANEL" TYPE: ITS FEATURES

This door is what is known an a "cross panel" door since it has 64 at the bottom, in lieu of the usual paired rectangular or square panels four triangular panels formed, with the help of the surrounding rails and stiles, by the crossing of two diagonal rails running from corner to corner of the panel area. This composite panel motive is 2'-3" high and 2'-10" wide. The upper part of the door has the normal type of paired rectangular panels, the lower ones being 2'-1" high by 1' 3 1/8" wide and the upper, 10 ½" high and 1 3/8" wide.

OLD EXAMPLES OF THIS AND SIMILAR DOORS

A cross panelled door of the character of this one is illustrated on p. 52 by a photostat of a drawing in the 1748 edition of William Salmon's Palladio Londinensis. Two old doors whose design may have been based on this plate are found on the south front of Wales in Dinwiddie County. This somewhat exceptional treatment of the bottom part of the panelled door sometimes took other forms. Salmon in Plate XXVI of his book shows another variant in which the rails form a "diamond" with curved sides, the "points" of this diamond falling on the axes of the panel motive. Five panels result from this design, a "diamond" with concave sides in the center and four quarter-circular ones in the corners. An old Virginian example of a panelled door utilizing this diamond pattern in the old south entrance door of Tuckahoe in Goochland County. This door has a cross panel motive at the bottom, a curved diamond motive above this and two square panels at the top.

DESCRIPTION OF THE PANEL MOLDINGS

The paneling of the entrance door is depressed 1/8" below the surface level of the stiles and rails which make up the door framework. The tongue of the panel which is inserted into the groove or channel cut in the stiles and rails is set back on either side about 3/16" from the panel surface. The frame edges on the panel side 65 have been shaped into quarter rounds which appear, in section, to grip the wedge-shaped panel tongue. The quarter rounds are also set back 1/8" from the surface of the framework of the door.

KING'S ARMS TAVERN, NORTH DOOR. SECTION THROUGH JUNCTION OF STILE AND PANEL.

The profile which results in one which was commonly used in Colonial Virginia. The old wall paneling of the Brush-Everard, St. George Tucker and Pitt-Dixon^* Houses, to mention only a few, has a profile which is very similar to that of our door. This paneling, of course, is single-sided since the back is turned against the wall. Old double-paneled doors utilizing the moldings described above are not rare, however. Several of these can be seen in the Brush-Everard House.

JOINING OF STILES AND RAILS

The stiles and rails of the door are set together with mortise and tenon joints and held in place by hardwood pegs driven through the joints. The tenons, formed on the rail ends, run completely through the mortises (slots) cut in the stiles.

PRECEDENT FOR USE OF TANSOM ABOVE DOOR

The precedent for the use, as we find it here, of a transom above an entrance door is not difficult to find since there are a number of old examples of it in Williamsburg. If one, however, includes in the listing only those old transoms which are one light 66 high, the number is reduced to about a half dozen. If we narrow the selection down still further to those one light high transoms which are four lights wide the number becomes still smaller. If we fix the condition that the door architrave or frame, as in the case of the King's Arms doorway, continue upward without interruption past the doorhead to enframe the transom, we find that we have only two old entrance doors left, those in the James Semple and the Timson Houses. Other examples, of course, are to be found outside of Williamsburg. The front door of Belle Farm, which once stood in Gloucester County^* had a continuous frame enclosing a door which had a four light wide, one light high transom above it.

BOLECTION PROFILES, AT LEFT, OF NORTH DOOR OF TAVERN AND, AT RIGHT, OF WINDOW FRAMES IN PEYTON RANDOLPH HOUSE.

DOOR FRAME IS BOLECTION MOLD SIMILAR TO OLD ONE IN PEYTON RANDOLPH HOUSE

The exterior door architrave or frame is formed by a molding of the type known as "bolection" which is slightly over 6" high and 2" wide at its thickest point. The headpiece of the frame extends high enough to overlap the plastered ceiling of the porch. Since it is difficult in words to convey a picture of this molding, it is illustrated here, together with an old example of a bolection which is found in use as interior trim on the windows of the library 67 the Peyton Randolph House. This latter molding may legitimately be viewed as precedent for the Tavern door frame since the profile, although smaller in scale, is very much like that of the door architrave, differing from it chiefly in the fact that its topmost bead is not present in the door molding. It is also of importance that the Peyton Randolph molding was used, as is the Tavern example, as enframement of an opening. The fact that in the one case the opening is that of a window and in the other of a door is not material, since the same types of moldings were frequently used in the eighteenth century in both of these locations.^*

DOOR SILL IS UNMOLDED, BLOCK TYPE

The door sill is of an unmolded, square-cut type. This means that it lacks the beads and other moldings possessed by many eighteenth century sills and is cut off at the ends at right angles to its front face. It projects outward about 1 ¾" and beyond the opening at the sides a little more than 6" to act as a stop for the stiles of the door. It is 3" thick at its front face. It is based on the precedent of old block sills such as that of the north (front) door of the Lightfoot House.

HARDWARE OF NORTH DOOR

The hardware of the entrance door has been reproduced for Colonial Williamsburg, Incorporated after colonial models. It consists of one pair of wrought iron HL hinges, 10" wide and 12" long and similar to CW F-4; a 12" high H hinge of the same design and one brass rim lock, 5 ½" x 10" with a brass knob and ring handle, similar in type to two old brass rim locks found on the exterior doors of the Tayloe House.

HINGES AND LEATHER WASHERS

The hinges are of the same decorative character as an old HL hinge found in the Casey House (now demolished) which has floriated ends, though the "flower" profile is somewhat simpler here than in the Casey example. The hinges are held in place on the inside face of the door by nails with hand hammered heads driven through the door and clinched on the outside in the colonial manner. Leather washers have been inserted between the nail heads and the hinge surface. Washers of this sort have been found very frequently on the interiors of old Williamsburg houses. A few cases of Williamsburg houses the doors of which had them (generally painted over) are the Benjamin Waller, the Brush-Everard and the Tayloe Houses.^*

POSSIBLE EXPLANATIONS OF PURPOSE OF WASHERS
FOUR WINDOWS ON FIRST FLOOR; ARRANGEMENT OF THESE AND DOOR IS NOT SYMMETRICAL

There are four identical 28 light first floor windows (four lights wide and seven high) on the north front of the building. An was noted on p. 46, there is no regularity in the spacing of 69 these windows, the three windows west of the porch having different intervals between them and the single window east of it being separated from the latter by a distance considerably greater than that between the porch and the first of the windows west of it.

This irregularity in the window spacing is traceable, in part, to the manner in which this house was built. The old foundations indicated that the main part of this eighteenth century house, which eventually became a tavern, was built at two different periods (see archaeological plan, p. 39). This two-period building history has been expressed in the present floor layouts in that the east walls of the northwest dining rooms on the two floors have been placed over the position of the original north-south foundation wall which separated the old east part of the house from the newer west part. A joint has also been left in the brickwork of the north basement wall at the approximate point where the north wall of the extension would have butted against the north-south dividing wall. Additions of this sort were made frequently to houses in Williamsburg in the eighteenth century. These generally led, as in this case, to an irregularity of window spacing in the enlarged facade since the first consideration in the location of the windows in the new part would be their serviceability to the interior and not their external appearance.

RECONSTRUCTED EASTERN PART ALSO WITHOUT SYMMETRY BECAUSE OF ARRANGEMENT OF INTERIOR

The foregoing would not, however, explain the fact that in the reconstruction of the early east part of the structure the door was located west of the central axis of this part and the flanking windows at different distances from it. This appears, on the face of it, to be an arbitrary departure from the usual colonial practice of maintaining bilateral symmetry about a central axis. The chief reason for this departure, it seems, springs from plan arrangement 70 which resulted from the present day use of the building, a subject which will be discussed more fully later.

Whatever may have caused it and even though it is not typical of eighteenth century building, there is in Williamsburg plenty of precedent for a non-axial, asymmetrical arrangement of the three openings of the north front of the original part of the King's Arms Tavern. There are several old central hall houses which, so far as we know, were never enlarged which depart to a greater or less extent from the principle of bilateral symmetry. Among these are the Moody, Bracken and Ewing Houses and Captain Orr's Dwelling. The street front of the John Blair House, furthermore, a building which underwent enlargement and which, with its two front entrance doors, has the appearance of two houses joined together, is definitely asymmetrical as a whole and also within the two parts, that is, the windows adjacent to the doors are arranged in an asymmetrical manner around these. If we suggest that when the John Blair House was lengthened the positions of the openings in the original part of the building were altered making the south facade of this part asymmetrical, we must grant that the same possibility exists that the facade of the original part of the Tavern was altered when that building was enlarged, making it also asymmetrical. In any event, as we have seen, we have sufficient precedent in Williamsburg for the non-symmetrical arrangement of the openings of the north facade of the King's Arms Tavern.

DESCRIPTION OF WINDOWS; NUMBER AND DISTRIBUTION OF GLASS PANES; MUNTIN PROFILE

The two sash of each opening of the north elevation combine to form a window 6'-11 7/8" high and 3'-1 ½" wide. Both sash move with the aid of metal weights. The 28 lights of the windows are distributed as follows between the upper and lower sash: 12 in the upper and 16 in the lower. The light size is 8" x 10 ¾" and the 71 glass is rippled and uneven, having been salvaged from old buildings. The frame members and the muntins have been given on the inside the quarter round profile so characteristic of window ornamentation in colonial Williamsburg (see adjoining section).

PRECEDENT FOR THESE WINDOWS

In the matter of the number of glass lights in the windows, there is precedent in Williamsburg for almost any combination of lights from two high and one wide to eight high and four wide. Old windows of the latter sort are found on the first floor of the Brafferton and the President's House of the College. The second floor windows of both of these buildings have the same number of lights as those of the King's Arms and those are distributed in the same manner between the upper and lower sash. The Bland-Wetherburn Tavern also has 28-light windows, four lights wide, but in this case the lower sash is three lights high and the upper sash four high.

PRECEDENT FOR MUNTIN PROFILE

The muntin profile, as we have said, was common in Virginia in the eighteenth century. Among a number of old examples in Williamsburg of this profile type, the muntin profiles of the sash of Bruton Church and the President's House of the College may be mentioned.

WINDOW FRAMES SIMPLE IN TYPE; PRECEDENT FOR THESE

The exterior window frames are of a type less complex than many 72 consisting of two members, one of which is superimposed upon the other. The external member is rectangular in section, 2 3/8" x 1 ¼". This rests upon a wood piece 1 1/8" thick, of which a length of about 1 5/8" is exposed. The outside band which, at the top, touches the bottom of the bed mold of the Tavern cornice, receives the weatherboarding and acts as a stop for the shutters. The inside member, the free, outside edge of which is beaded, serves as a guide for the upper sash and as one side of the chase in which the sash weights move. The overall height of this two-part trim member is 4". A number of old first and second window frames of the Brush-Everard House are of the same character as the King's Arms frames and can be considered the precedent for them.

SILLS ARE OF MOLDED VARIETY

The window sills are of the molded type, approximately 4" high at the face of the building. They consist of an approximately semicircular curve with a diameter of 1 ½" at the top, followed in succession by a fillet, cove and bead, the latter lining up with the bead of the adjacent weatherboard whose lower edge has been made level with the bottom edge of the sill. All of these moldings return against themselves at the sill ends.

PRECEDENT FOR SILL

This sill type has precedents in Williamsburg in the first floor window sills of the Tayloe House and in two old first floor window sills on the north elevation of the Taliaferro-Cole House. The first floor windows of Wales in Dinwiddie County also have two or three variants of this sill.

WINDOW PARTS SET TOGETHER

The parts of the window members discussed above -- sash, frames and sills -- are held together by mortise and tenon joints. These are rendered secure by hardwood pegs driven through them. Pegged mortise and tenon joints were a conventional joinery device of the eighteenth century.

SHUTTERS OF BUILDING PROPER ARE PANELLED

The window shutters of the north elevation and for that matter, also of the west elevation, the only other facade of the main part of the house which has them, are panelled. The first floor shutters are three-panelled, the top and bottom panels being elongated and the middle one square. In the case of the smaller second floor windows of the west elevation the shutters have two equal rectangular panels only.

DESCRIPTION OF PANEL MOLDINGS

The shutters have the panelling showing on both sides although they are molded only on that side which is turned outward when the shutters are open. The flat back faces of the panels are recessed about 5/16". The ornamental profile on the molded side, as was customary, consists of moldings both on the panel edge itself and on the surrounding framework. The molding of the frame is a slightly recessed quarter round. The panel edges are shaped into the usual sloping, wedge-shaped tongues which fit into rebates or grooves cut in the stiles and rails. The panels "float" in these grooves; i. e., they are not fastened and are free to move as the panel expands or contracts with changes of temperature. The panel tongues, themselves, are ornamented by a recessed quarter round effecting the transition between the surface of the panel and the panel tongue. When the shutters are closed the adjacent stiles meet in a shiplapped joint and their overlapping parts on either face of the shutter are cut out to form a 3/8" bead. The parts of the shutter frame are mortised and tenoned together in the colonial manner and secured in place by means of hardwood pegs.

The panel molding just described is a typical early Georgian profile. Old examples of it have been found at Christ Church in Lancaster County (1732) where it is used on the inside of the entrance 74 door and also on the pew panels, and at the Towle's House (now destroyed) in the same county where it was used on a door panel.

Our illustration is of a part of the mansion, Wales, in Dinwiddie County. It shows a pair of old three-panelled shutters with the panels arranged like those of the shutters of first floor of the Tavern. It also shows, incidentally, a window sill of the same molded type as those of the Tavern and a modillion cornice of the character of the main cornice of that building. The dwelling on Maine Farm near Williamsburg also had three-panelled shutters similar to the Tavern shutters. (This house has long since fallen into ruin.

OLD THREE-PANELLED SHUTTERS OF WALES NEAR PETERSBURG. THE SHUTTER HOLDBACKS ARE VISIBLE AS WELL AS THE BAR STRAP USED TO FASTEN THE CLOSED SHUTTERS FROM INSIDE. AN EYE IN THE END OF THE BAR HOOKS OVER BUTTON ON OPPOSITE SHUTTER.

SHUTTER HARDWARE

The shutter hardware consists throughout of wrought iron reproductions made for Colonial Williamsburg after colonial models. It includes the following items: Pintel hinges, CWI Folder Plate 1-A, Type 2 with pintels; surface bolts, 8" long (CW F-18); shutter holdbacks similar to CWI Folder Plate 22, Type 4 and cabin hooks and 75 staples similar to CW F-21, 4" long.

Something should be said concerning the significance of the use of panelled shutters on the main portion of the King's Arms building and the employment, on the other hand, of louvered or slatted shutters on the south wing. This change in shutter type, it may be said at the outset, was occasioned by the architects' objective of expressing as far as possible in the detailing of the south wing the fact that the original version of this wing was built later than the main part of the building. This has the following basis in the history of the development of shutter types and their uses: Shutters serve several purposes today -- they can be closed in order partially or fully to shut out the light and as a protection against housebreaking. In the less secure early colonial days shutters served this last function in the first instance and the other secondarily. Since the Tavern, to the best of our belief, was built early enough in the eighteenth century to warrant the taking of precautionary measures against unlawful entry of the building, the reconstructed older part was provided with closed shutters. There are many instances of the use of such closed shutters on the first floors of early building. These shutters are very frequently boarded for added security on the side which is out when the shutters are closed, and panelled on the reverse side. Frequently the shutters of the first floor, where protection is principally needed, are of the closed type and those of the second louvered. As time went on the need for the protection afforded by the closed shutter was no longer felt so that louvered blinds or shutters which permit the passage of air through the shutter when the latter is closed, were substituted for it. This is the state of affairs which is intended to be represented by the louvered shutters of the Tavern wing.

SOURCES OF EXAMPLES OF HOUSE WITH PANELLED AND LOUVERED SHUTTERS

The book, Early Architecture of Delaware by George Fletcher Bennett is rich in illustrations of original examples both of houses which have panelled shutters on the first and second floors, and of those which have panelled shutters on the first floor and louvered ones on the second. The book also contains examples of buildings in Delaware with shutters boarded on one side and panelled on the other end, finally, of houses having the simplest type of all, board and batten shutters.

BASEMENT WINDOWS LOCATED IN ACCORDANCE WITH COLONIAL PRACTICE

There are but two basement windows on the north elevation and these are centered beneath the two windows nearest the western end of the building. Neither the archaeology nor other evidence gave any indication as to the numbers and positions of the basement openings so that these were placed beneath the first floor windows wherever this was feasible, as they would have been in the eighteenth century. No basement windows were placed in the eastern two-thirds of the front facade because the original basement here was "roofed" with a barrel vault, which would have made the use of such windows on the north and south sides impossible (see archaeological drawings, p. 39). In most cases the basement windows are useful in bringing some natural light to the kitchen work spaces in the basement but in three instances, as in the case of the westmost window of the north elevation, the windows are blocked off and rendered useless by sealed-up refrigerator rooms existing behind them.

OPENINGS HAVE GRILLES ON OUTSIDE AND SASH ON INSIDE; PRECEDENT FOR THIS

The basement windows are designed in such a way as to have the appearance on the outside of typical eighteenth century grilled openings while, at the same time, they are provided on the inside with glazed bottom-hinged wood sash, four lights wide and one light high. These have the same type of muntin as the first floor windows. 77 Although this combination of bars and glazed sash is exceptional in eighteenth century buildings in Virginia, sash, used in conjunction with basement grilles, did exist. Examples of these have been found at Scotchtown, the Hanover County house occupied by Patrick Henry between 1771-1777, and at Rosegill, the Wormeley mansion in Urbanna in Middlesex County. Old grille frames with original hand wrought bars were found in several basement windows of the Wythe House. These had sash rebates on their interior faces and, though the original sash were missing, remnants of the leather hinges on which they swung inward from the top were still in place.

DESCRIPTION OF BASEMENT WINDOW FRAMES AND BARS

The open, exterior grilles consist of a heavy wood frame with four horizontal wood bars let into it. The frame is 3'-8 ½" long by 1'-11" high. The head and jamb members are rectangular pieces approximately 3 ¾" x 4 ¼" in size. These are set in about 1 ¼" from the face of the brick basement wall and have an ogee backband applied to them on the outside covering the joint formed by their meeting with the brickwork. The frame is divided veridically in the center by a mullion about 3 ¼" x 3 ½" in size. The salient corners of the head and jamb members have been shaped to the form of a bead, ½" in diameter. The jamb members rest on a sill whose front face, about 2 7/8" high, has been brought out flush with the surface of the brickwork. The members of the frame are mortised together and secured with hardwood pegs driven through the joints. The four bars constituting the horizontal grille are 1 1/8" square in section and are spaced 3 ¼" center to center. They are, of course, divided into two equal lengths at the middle of the opening by the mullion. They have been rotated through a 90° arc so that they present an edge to the observer rather than one of their flat surfaces.

WILLIAMSBURG HOUSES WITH OLD BASEMENT GRILLES; COMPARISON OF THESE WITH GRILLES OF TAVERN

A number of old basement grilles and parts of grilles have been found in Williamsburg and these represent the precedent for the reconstructed grilles of the King's Arms Tavern. The following houses, among others, had them: the Orrell, the Bracken, the Lightfoot, the Brush-Everard, the Tayloe and the President's House of the College. These old grilles are basically like those used on the Tavern, although they differ in some respects from them and from each other.

The frames are all beaded but they are of varying weights of material and some are divided by mullions and some are not. The number of bars vary from three (Bracken, Brush-Everard and Lightfoot Houses) to five in the grilles of the President's House. In some instances the bars are all of wood and in others one or two of them are of wrought iron for greater security, as in the Brush-Everard grilles and those of the Tayloe House. In the case of the President's House a single old grille was found in a good state of preservation beneath the later south porch, where it had been protected from the weather. This grille was exceptional in that it was set in a segmental-arched opening and in consequence of this a filler piece or tympanum of wood was needed between the top of the headpiece of the grille frame and the bottom of the brick arch. None of the houses mentioned had basement grille frames ornamented with backbands but plenty of examples of these have been discovered. One instance of this treatment was found, for example, in the basement grilles of the place called Farmington in Charles City County.

CORNER BOARDS AT NORTH CORNERS OF BUILDING; THE PURPOSE OF THESE

There are cornerboards of the two-faced type at both the northeast and northwest corners of the building. These are vertically placed boards rising, as does the weatherboarding, from the foundation to the bottom of the cornice. Their purpose is to "dress up" the 79 building corners and to effect a transition between the plane of the north face of the building and those of the west and east faces.

CORNER BOARD AT NORTHWEST CORNER

The corner board (or boards) at the northwest corner is 3 ½" wide and 1 ¼" thick on each face. The two pieces meet at the corner in a mitre which is hidden by a half inch bead cut from the end of the west board. The edges of the boards adjacent to the weatherboarding for which they act as stops, are unbeaded.

CONDITION AT NORTHEAST CORNER MADE WIDER BOARDS NECESSARY

The structural condition at the northeast corner of the building is different from that at the northwest corner since at the former corner we have the meeting of the wood frame wall of the north front with the brick wall of the east end. As is structurally logical the frame wall butts against the 9" brick wall, so that the end of the latter wall would appear on the north face of the building if it were not masked. This masking of the brickwork is accomplished by a corner board wide enough (11 ½") to cover the end of the brick wall and to overlap this 1 ¼" on either side. This north corner board is paired with one at right angles to it, which is 5 ½" long. This board overlaps the brickwork of the east end wall about 4 ¼". These two boards are similar to the pair at the northwest corner except in length and in the fact that their outer edges are beaded.

PRECEDENT FOR NORMAL WIDTH AND EXTRA WIDE CORNER BOARDS

Old beaded corner boards, along with some new replacements of old ones, are found on a number of wood frame dwellings in Williamsburg, among which are the Lightfoot, the Peyton Randolph and the Tayloe Houses. In the case of the Travis House, an old building which has been moved to a temporary site on Francis Street, a condition similar to that of the King's Arms Tavern exists, since the house is of wood but has a single brick end (the end which in the present location of the building faces east). At the junction of the present north face 80 with the adjacent west end we have a two-sided corner board the pieces of which are of normal width -- about 3 ½". At the opposite end where the wood frame strikes the brick end wall, we find the brick end capped by a corner board nearly a foot in width. The other member of the corner board pair, which is superimposed upon the brickwork, is again of normal width.

"VESTIGIAL" CORNER BOARDS, A SIGN THAT A HOUSE WAS LENGTHENED

It is interesting and of some importance to note that the corner board situation of the north elevation of the Travis House has not been fully covered by the above discussion. Between the door and the easterly of the two middle windows and also between this pair and the pair near the west end, two old corner boards or, better, "vestigial" corner boards since they are no longer on a house corner, exist which interrupt the weatherboarding at those points. These boards were each once located at what were at two successive periods corners of the house so that their presence signifies that the house was twice lengthened. When, in each instance, the house was extended these boards were simply left in place.

KING'S ARMS LACKS THESE; REASONS FOR THEIR OMISSION

Though the north front of the King's Arms was extended westward no vertical board was inserted in the weatherboarding at the point where the extension took place in the eighteenth century, though a joint was left in the foundation brickwork to mark the line where the second period foundation brickwork formerly butted against the old wall of the first building period. The omission of the corner board is not an oversight in the reconstruction of the Tavern for, in the eighteenth century, in the course of making an addition to a building it would in many cases have been necessary or have been thought desirable to renew the weatherboarding, particularly on the street face of the structure, and in the course of this a cornerboard 81 which had become useless might well have been removed. A restored house which is a good example of this is the John Blair House which was lengthened and which, nevertheless, has no vestigial corner boards on the street elevation.

EAST ELEVATION

GENERAL

It has been noted on p. 35 that the east wall of the Tavern is of brick throughout and the bond of this and of the chimney, which from the top of the upper haunch downward is united with it, has been discussed, as well as the size and color of the brick. It should be stated at the outset that there are no window or door openings in this wall. The south half of this exterior, up to the level of the finished second floor, has, by the presence of the (unauthentic) serving pantry element which butts against the east facade at this point, been eliminated, along with the features, such as the lower left half of the chimney, bottom of cornice end board, the corner board, etc., which would otherwise have been located here. Again, elements not present on this facade will be omitted from the outline followed in the detailed treatment of this elevation.

BRICK WALL TAKES FORM FROM THAT OF BUILDING FRAMEWORK; THE LATTER BUTTS AGAINST THE WALL

The form of the wall (see elevation, p. 27) is, naturally, that which is given to it by the cross-sectional shape of the building framework. Though the superimposed trim members which follow the contour of the wall might lead one at first glance to the conclusion that the brick wall is contained within the framework of the building end, it stands, in actuality, outside of this. It reaches, on the north side, to the front of the structure where its end is masked, as we have said, by an unusually wide corner board (pp. 78, 79) and it rises at all points of the roof to the height of the tops of the roof rafters, so that only the roof covering overlaps it. Evidently, therefore, the east end wall is not enclosed by the structural framework of the building but, rather, stands outside of it and acts as a sort of abutment to it.

FALLEN SLAB INDICATES THAT EAST WALL WAS OF BRICK

We have good reason to believe that the east wall was originally of brick. In excavating within the foundation walls of the older, east part of the basement, a horizontal slab of brickwork of considerable area was uncovered near the east foundation wall (see archaeological plan, p. 39 and photo, N 3488 p. 40). Because of its thickness and the oyster shell lime mortar which still held the brick together the slab was identified positively as a section of fallen wall rather than an area of paving. The height of this slab of brickwork added to that of the still upright foundation wall near which it lay and from which it must have become broken off, was such that the top of it, when the wall was in its original position, would have been about 7'-0" above grade. This seems to be sufficiently conclusive evidence that the end wall was of brick.

OTHER WAYS OF EXPLAINING SLAB; THESE DO NOT SEEM TO APPLY HERE

It may be asserted, on the other hand, that the fallen slab can be explained in other ways. There are, indeed, two or three other wall conditions into which such a brick panel might conceivably have fitted but, on consideration of these it does not appear that the slab in question was a part of any of them. It might be suggested, for example, that though the first story of the east end was of brick,the gable above it might have been constructed of wood. This is possible although there were few eighteenth century Virginia houses (Windsor Shade, formerly Waterville, in King William County is one) which did not have the same material, either brick or wood, on both stories. It might, on the other hand, be argued that this was one of those fairly frequent cases in which a fireplace on the inside of an end wall was expressed as a rectangular panel of brickwork on an exterior which was otherwise of wood. No evidence of interior chimney was found at this point however, and, furthermore, the fallen slab 84 appears to have been too wide to have been such a panel. Finally, it might be contended that this was one of those cases in which an exterior chimney, standing in contact with the end wall, was rendered more fire safe by the insertion in the wood wall of panels of brickwork above the chimney haunches. The Bracken House in Williamsburg has these on both the east and west end elevations. The same objection can be raised here as in the previous case -- our panel is too broad to have served in this capacity since such panels are generally the width of the chimney only and sometimes, as in the Bracken examples, less than this. The most reasonable conclusion to reach, therefore, on the basis of the above discussion, is the one the architects arrived at, viz., that the fallen slab was at one time part of an entire wall of brick.

END CHIMNEY; REASONS FOR BELIEVING ON ESTOOD AGAINST EAST WALL

Another question arises immediately in this connection. The foundation drawing, p. 39, gives no indication that a chimney once stood, either inside or outside, against the east end wall. The archaeologist, Herbert S. Ragland, who investigated the property, points out in the report which he wrote in 1933 about the foundations discovered on the King's Arms Tavern site that fireplaces and, hence, a chimney would have been necessary to heat the rooms of this early part of the house. Since no indication of a chimney was discovered within the foundation, there must have been one on the outside. He concludes that the chimney must have stood against the east wall and that it was probably built on shallow foundations which have since disappeared. He mentions having seen a landscape drawing of 1929 made by Arthur A. Shurtleff (Shurcliff) which shows a "brick wall" of a shape and in a location which confirms the supposition of this chimney.

Singleton P. Moorehead suggests that such a chimney might have 85 been erected after the original house had come into existence and have been corbelled^* out from the foundation wall. Examples of corner chimneys supported by corbelling rather than foundations discovered in Williamsburg. This is a possible exp it had been the case we would expect to find racking corbelled brick had been bonded in) in the outside f foundation wall, but the archaeologist did not report condition. It seems that m definitive word can be whether or not the present chimney stands in the pos one. It is, nevertheless, reasonable to believe, with the architects, that it does.

GENERAL DESCRIPTION OF CHIMNEY

The chimney is a three-staged stack centered on the axis of the east facade (see elevations, pp. 26, 27). It is bilaterally symmetrical or would be if the south half of the lowest stage had not been eliminated because of the interposition of the serving pantry element between the Tavern and Purdie's House. It rises with two pairs of lateral setbacks to terminate in its shaft but unlike many old chimneys of its type (see Bracken and Moody chimneys) the east face remains throughout the entire height of the stack, with the exception of parts of the cap, in a single vertical plane. The chimney serves a fireplace on the first floor and, in the eighteenth century, would have served another on the second but the latter was omitted in the reconstruction of the building since public toilets rather than a bedroom were placed in the east end of the building.

LOWEST STAGE OF CHIMNEY

The lowest part of the chimney, which would be a little over 11'-0" wide if it were complete, is broader than the section immediately above it, since, in colonial times, at least, the fireplace which it would have accommodated on the first floor would have been larger than that on the second. It would also have been larger as a matter of stability since the superimposed weight which it supports is considerable.

TRANSITION TO MIDDLE SECTION BY MEANS OF SLOPING PLANES

The chimney width is diminished above the level of the fireplace opening by means of two haunches inclined at about 61° from the horizontal. (In actuality the serving pantry has necessitated the elimination of the south haunch but since this is an unauthentic departure from the eighteenth century condition we will treat the chimney as if both haunches existed.) The haunches terminate in the second, narrowed stage of the stack at the level of the finished second floor. The width of this middle part of the chimney is about 7'-4". Two haunches of the same slope as those below effect a transition between the sides of the middle section of the chimney and those of the shaft above, the junction of the latter with the haunches taking place about a foot above the first floor ceiling.

SHAFT IS T-SHAPED; REASON FOR THIS FORM

The chimney shaft, which is T-shaped, rises about 6'-6" above the ridge line of the roof. The present shaft houses three flues, the uses of two of which are unauthentic. The eighteenth century counterpart of this chimney would have contained two flues, a large rectangular one in the "head" of the "T" to serve the large fireplace on the first floor and a smaller squarish one in the shaft of the "T" for the smaller second floor fireplace. The existence of the elongated flue accounts for the T shape of the chimney in those cases in which only two fireplaces were served by the stack.

DESCRIPTION OF PROFILE OF ORNAMENTAL CAP OF CHIMNEY

The profile of the chimney caps, describing this from the bottom upward consists, first, of three brick courses corbelled outward, each overhanging the one below it about 1", followed by a band two courses high which likewise projects about 1". A sloping plaster surface atop this joins the outer edge of the double course with the bottom of the crowning band which consists of a single row of brickwork, the faces of which fall in the vertical planes of the sides of the shaft. A cement wash protects the upper surface of this from the infiltration of moisture.

HAUNCH SURFACES COVERED WITH TILE; OTHER DETAILS AT HAUNCHES

The sloping planes of the haunches are covered with square brick tiles, 8 ¾" on a side and 1 ¾" thick. Above each slope a single brick course is projected out about 1" as both an ornamental feature and as a covering for the joint between the brickwork of the chimney sides and the tile-surfaced slope. The haunch at the bottom also projects 1" over the chimney sides below to carry the water falling from the slope away from the surface of the brickwork below.

CHIMNEY STANDS FREE OF WALL ABOVE LEVEL OF UPPER HAUNCHES

The chimney is engaged with the brick end wall as far up as the top of the upper haunches. From this point the shaft rises free of the building. The T-shaped section of the shaft does not fully cover the top surface of the part of the chimney which supports it and a U-shaped surface is left exposed. This is given a slope flatter than that of the haunches and protected, like the latter, by a covering of brick tile.

RUDIMENTARY WATERTABLE NEAR BOTTOM OF CHIMNEY AND END WALL

At the level of the bottom of the weatherboarding -- about 2'-10" above grade at the east end -- the chimney brickwork; like that of the east wall itself, moves out about ½", thus indicating the top of the foundation. There is no watertable in the usual sense of this term.

VENT HOLES IN EAST WALL OF CHIMNEY

In the brick wall directly behind the base of the "T" of the chimney shaft, a few inches below the roof peak is a vertical vent opening, 17 ¾" by 5" and for a similar purpose two headers have been omitted from the face of the second stage of the chimney. None of these openings would have existed in the eighteenth century. The putlock (or putlog) holes found in colonial brick walls, which are similar in appearance to the vent holes in the second stage, served another purpose, viz., to receive the timbers of scaffolding. Putlock holes did not run through the wall, as do our vent holes.

KING'S ARMS CHIMNEY HAS PRECEDENT IN STACKS OF WILLIAMSBURG; EAST CHIMNEY OF BLAIR HOUSE

Practically all of the features of the Tavern chimney have their precedent in old Williamsburg chimneys so that it is scarcely necessary to look outside of the town for examples to support them. Outside end chimneys serving fireplaces on two floors were, in the eighteenth century, more often than not three-staged structures, consisting, as in the Tavern chimney, of a lower part housing the first floor fireplace; a narrower second stage serving the smaller second floor fireplace and a third level devoted to carrying off the smoke (the smokestack). We find all of these elements in the east end chimney of the John Blair House, in the two similar though not identical outside end chimneys of the Bracken House and in the two chimneys of the Moody House, though the detailing of all of these departs in certain respects from that of the Tavern chimney. It should be observed, in the case of the John Blair chimney, that the lowest stage is, in reality, only a base since it is below the level of the first floor window sills and accommodates no fireplace. The first and second floor fireplaces are here served by an elongated second stage. This element diminishes in three directions (in depth as well as laterally) to attain the base of the T-shaped shaft.

DETAILS OF INTEREST IN CHIMNEYS OF BRACKEN HOUSE

The east end chimney of the Bracken House is larger than the west one, its first stage being broader and its second being higher than the same elements of the west chimney. It also exceeds the west one slightly in overall height. The second stage in these chimneys also diminishes in depth as well as laterally with this difference from the condition which obtains in the Blair chimney that the step back in depth occurs above the lateral set backs. A noteworthy feature of these chimneys is the brick wall panels mentioned on p. 84.

THE MOODY CHIMNEYS, THE SMALLEST IN SIZE OF ALL THOSE MENTIONED

The two exterior end chimneys of the Moody House are of the same general type as the examples just mentioned and as the Tavern chimney, but they are much slenderer than any of them and particularly than the Tavern stack. The bottom stage of the larger of the two is only about 6'-0" wide as against 8'-0" for the larger Bracken stack and 11'-0" for the Tavern chimney, considering the width of this as it would be if it were completed. Only one of the Moody chimneys, the west one, accommodates a fireplace on the second floor and, in consequence it has greater depth than the east one. These chimneys also have brick wall panels, but behind and above the lower haunches only.

THE CHIMNEYS TREATED ABOVE ARE T-SHAPED WITH CAPS LIKE THOSE OF TAVERN CHIMNEY

The east chimney of the John Blair House and the chimneys of the Moody and Bracken houses all have T-shaped chimneys. The caps in all these cases are of the same basic form as that of the Tavern stack, except that there are only two corbelled courses beneath the doubled course which forms the most salient feature of the member. Williamsburg chimneys which do, however, have the three projecting courses beneath the two-coursed band are the chimney pairs of the Clerk's Office and the President's House of the College, (th latter are old chimneys rebuilt). These are not, of course, end chimneys but, rather inside ones which rise through the roof.

THEY ALSO HAVE PROJECTING BRICK COURSES ABOVE AND BELOW HAUNCHES

The haunches of the Blair, Moody and Bracken chimneys have, except for one case, protective "drip" courses both above and below the haunch slopes, similar to those of the King's Arms chimney. The exception is found in the Moody chimneys which lack these projections above the upper haunches, probably because the sloping planes here are relatively so curtailed.

HAUNCH SURFACES COVERED WITH BRICK; OLD EXAMPLES ON WHICH TILE WAS USED

All of the chimneys described above, unlike the Tavern chimney, have brick on the haunch slopes, laid with their long sides paralleling the direction of the slope. We have to go outside of Williamsburg to find old chimneys on the haunch slopes of which square brick tile have been used. The three-staged end chimney of the Curtis House near Lee Hall in the vicinity of Williamsburg has square tile on the lower haunches and brick on the upper ones. A reasonable explanation for the use of brick on the upper slopes is that these surfaces, due to the fact that the stack is T-shaped, are irregular and the elongated shape of the brick made it easier to fit these in than it would have been to use square tile. Another example is the two-staged chimney of the west office of Greenway in Charles City County which serves only one fireplace. Two rows of brick tile have been used on the haunch slopes of this chimney, with a single row of brick above them. Here again the problem of fitting the square tile to the area of sloping surface assigned to them may have led to the substitution of brick for tile in the upper row.

TRIM PARTS ELIMINATED BY PANTRY ADDITION; CORNICE END BOARDS

As has already been remarked in the general notes, p. 82, the south corner board of the east elevation and the lower part of the south cornice end board have, perforce, been omitted from this elevation since the west end of the serving pantry link butts against the part of the east facade on which they would have been located 91 We have, therefore, on this front, one complete cornice end board (the north one) and the upper half of the other (the south end board). The remaining half of the south end board is scarcely representative since its characteristic feature, the contour cut approximately to the cornice profile, is not present.

PURPOSE OF END BOARDS; OTHER WAYS OF TREATING CORNICE ENDS

The function of cornice end boards is evident from the title given them. These are flat boards placed at either end of a cornice to terminate it. Their use presupposes that the moldings continue to the corners of the building so that their "raw" ends require to be covered in some way. In certain cases some or all of the cornice moldings return upon themselves and consequently require an end board to cover only the moldings which continue through or no end board at all.^* In post-eighteenth-century gable ended buildings the problem was solved by carrying the moldings around the corner and across the gable end a short distance and then returning them against the building. The problem, of course, did not exist at all in hip-roofed buildings in which the cornice ran at the eaves line along all faces of the building equally.

CORNICE END BOARDS FALL INTO TWO GROUPS

Cornice end boards varied widely in design in the eighteenth century. They may all, for convenience, be placed in one or the other of two groups, viz., that comprising boards the outer edges of which are cut more or less to the profile of the cornice and that which includes the boards which are designed, in whole or in part, freely and without reference to the cornice profile and consequently do not follow it.

TAVERN END BOARDS FOLLOW APPROXIMATELY CONTOUR OF CORNICE

Although not every detail of the cornice moldings is reflected in them, the cornice end boards of the King's Arms (this holds for the one at the southwest corner as well as the two at the north corners of the building) may be considered as belonging in the category of those end boards whose profile matches, more or less, the profile of the cornice (see pp. 48-50). The end board follows exactly the contour of the crown mold and the fascia, except, in the case of the latter, in so far as it extends ½" below the lowest part of the fascia. From this point the end board runs horizontally toward the building, screening the modillion blocks from the view of an observer looking in a line perpendicular to the end of the building. The horizontal edge drops about ½" below the bottom of the blocks and has cut out of it near the front a near semicircle with a diameter of 2" which acts as a drip. The bed mold of the cornice is echoed in a cyma reversa curve cut from the lower part of the end board, but this shape lacks the fillet which in the cornice divides the bed mold into two parts.

RELATION OF CORNICE END BOARDS TO RAKE AND CORNER BOARDS

The cornice end boards can not well be treated apart from the corner boards (p. 78) and the rake or barge boards (p. 94) since they are continuous with these members. They taper upward so that at the point where they join the rake boards they are only 5 ½" wide. They also diminish in width toward the bottom, following the setbacks of the cornice, and join the corner boards about 1 ½" below the bottom of the cornice bed mold. Their outer edge projects about 5/8" beyond the outer edge of the corner board and the beaded inside edge of the corner board is continued upward in the vertical inside edge of the end board. Since the corner board on the east face of the Tavern is 5 ½" wide while those on the west face are only 3 ½" and since 93 the relationship of the parts of the end boards to the edges of the corner board remains constant, the cornice end board on the east end is 2" wider than that on the west face. This is, however, the only respect in which the two boards differ, their exterior profiling and height being exactly the same. The cornice end board at the southwest corner is the counterpart in reverse of the one at the northwest corner.

DIMENSIONS OF CORNICE END BOARDS

Measuring from the top of the crown mold to the top of the corner board the height of the end boards is approximately 17". The widths of the boards measuring from the outer tip of the crown mold to the inside vertical edge of the end boards are 18" (west face) and 20" (east face) respectively. From the dimensions just noted it is evident that the cornice end boards could not readily be cut from a single piece of wood. They are each, as a matter of fact, made of three pieces of wood whose edges run diagonally upward and are fitted together by tongue and groove joints. The boards are 1 1/16" thick.

PRECEDENT FOR TAVERN END BOARDS, NO SPECIFIC COLONIAL EXAMPLE WAS FOLLOWED

This writer has been unable to find in an old building the exact counterpart of the cornice end board of the King's Arms Tavern and, for that matter it is unlikely that it was copied in toto from any one old example. It would surely be the privilege of the restorers to do with this feature what the carpenter-builders of the eighteenth century did with it, i. e., design it freely in any form which struck their fancy since these boards took such a variety of shapes in colonial times as to make an attempt to record and classify them all well-nigh impossible. One example may be mentioned here, however, which is like the Tavern end board in principle, viz., the end boards of Toddsbury in Gloucester County, which follow more or less closely the contours of the modillion cornice. These also have the semicircular cutout which is a somewhat striking feature of our cornice.

END BOARD OF UPPER CORNICE IS SIMPLE MEMBER WHICH FOLLOWS CORNICE PROFILE

The upper cornice, so-called, as we have noted (p. 47), is a crown molding (cyma recta, fillet and cyma reversa) used to effect a transition between the upper gambrel slope and the lower. Just as in the case of the main cornice, this molding requires to be covered at the ends of the building with end boards. Since the upper cornice is relatively so simple, the end board is also simple. It is a flat piece of wood with four sides, no two of which are parallel. It is wedged in, so-to-speak, between the underside of the overlapping shingles of the upper roof slope and the backband running along the outer edge of the rakeboard. Its outside edge is cut to the profile of the crown mold. Its bottom edge is horizontal and overlaps the beaded board which is inserted between the crown mold and the shingles of the lower roof slope. Its greatest height and width are about 5 ¼".

PRECEDENT FOR USE OF THIS FEATURE; TRAVIS ROOF DOES NOT REQUIRE IT

A pair of these boards is found on each of the gable ends of the Tavern. They were used in the eighteenth century on most gambrel roofs. Old gambrel roofed buildings in Williamsburg which have them are the Tayloe, Lightfoot, Powell-Hallem and Orrell Houses though these are, in part, replacements of old ones found on the houses. The Travis House gambrel, on the other hand, does not require them because the upper cornice in this case returns upon itself.

RAKE OR BARGE BOARDS; THEIR FUNCTION

The rake or barge boards serve a purpose related to that of the end boards and, as we have said, they are continuous with them. They consist of narrow wood strips which follow the roof slopes on the gable ends. They overlap the weatherboarding of the west end and the brickwork of the east end of the Tavern and the shingles which project over the edges of the roof slopes, in turn, overhang them.

DESCRIPTION OF THESE

Due to the "jointed" character of the gambrel roof the rake boards change direction at the level of the upper cornice. Both "legs" of the rake taper upwards, the lower one from a width of 5 ½" at the point where it joins the main cornice end board to 4 ½" at the junction with the upper leg and the latter from 4 ½" at this lowest point to 4" at the peak of the gable where it meets a similar rake board which runs up the opposite slope. The boards are 1 1/16" thick and have their salient inner edges cut to the form of a bead, ½" in diameter.

The upper or outer edge of the rake boards is ornamented with a backband, a molding consisting of a fillet and an ogee (cyma reversa). The backband changes direction at the joint between the two legs of the rake. It follows the lower leg down to the main cornice end board and continues along the upper edge of this to the point where the outer, profiled edge of the end board strikes the shingles and here it is returned against the face of the end board. In similar fashion the length of backband running along the upper rake is continued out beyond the joint formed by its meeting with the lower length of backband. It runs out along the upper edge of the upper cornice end board to a point about ¾" from the outer, profiled edge of the latter, where it is returned against the face of the end board.

RAKE BOARDS A COMMON FEATURE OF GABLE ENDED HOUSES; OLD EXAMPLES IN WILLIAMSBURG

Rake boards were a feature of eighteenth century gable ended houses whether A roofed or gambrel roofed. All of the old examples of gambrel roofed dwellings in Williamsburg -- the Orrell, the Powell-Hallam, the Tayloe, the Lightfoot and the Travis Houses -- have rake boards although not every piece is original, some of them having been replaced since they were in a deteriorated condition. In most of these cases the boards have both the bead and the backband. The Travis House 96 rake has the bead but lacks the superimposed molding, however.

Rake boards were almost invariably tapered in colonial times but old records fail to give us the reason for this. It is the opinion of the architects, however, that the taper was used for aesthetic purposes -- that it created a false perspective and accentuated the slopes of the roof.

SOUTH ELEVATION

GENERAL

The south elevation of the King's Arms Tavern (see drawing, p. 26) presents two complicating factors, viz., the serving pantry link which unites the Tavern with the Purdie House and the wing which is attached to the western part of the south front of the building. It seems feasible to complete the treatment of the exterior of the pantry here to obviate the necessity of dealing with it later in connection with the south elevation of the Purdie House. The Tavern wing, on the other hand, will be discussed separately, elevation for elevation, following the treatment of the four facades of the main part of the structure.

ROOF SLOPES; ROOF COVERING

The gambrel roof slopes are the same here as on the north elevation (see p. 32). For a discussion of the shingles see p. 46. The shed roof of the pantry extension will be discussed under the latter subject.

DORMERS AND DORMER WINDOWS SIMILAR TO THOSE OF NORTH FRONT

There are three dormers on the south front which are, in all respects, of the same size and type as the five front dormers. For a discussion of the latter and their sash, see pp. 46-48. The middle and west dormers are not centered on their counterparts on the north front but the east one is, its center, like that of front dormer, being about 7'-5" from the east edge of the roof. The two bays between the dormers are almost though not quite uniform, the distance between the center of the east dormer and that of the middle one being about 10'-2" and between the latter and the center of the west one being approximately 10'-5".

MAIN CORNICE; ONLY A FRAGMENT OF THIS APPEARS ON SOUTH FACADE

The shed roof of the modern pantry extension eliminates the major portion of the main cornice which this facade doubtless had in the 98 eighteenth century. Only a length of it about a foot wide remains, between the west side of the wing and the west end of the south facade of the main building. This fragment and the wing cornice are identical in character with the front cornice (see pp. 48-50).

REAR PORCH PROBABLY EXISTED; ITS NATURE NOT KNOWN

We have reason to believe that the Tavern in its second building stage covered the area indicated by the foundations discovered on the site (see archaeological plan, p. 39). The south wing was thereafter added to this, probably by extending the west end wall and the cross wall which divides the basement into two parts. A rear porch may have been built about this time, also, for we know that the Tavern had a second porch^* and the logical place for this would have been at the rear. Though we grant that a rear porch probably existed, we have no way of knowing exactly what the nature of it was.

SOUTH PART OF SERVING PANTRY & DISH PANTRY INCORPORATED IN ENCLOSED PORCH & ITS EXTENSION

In building the admittedly unauthentic serving pantry (see plan, p. and elevation, p. 26) the south part of this element was incorporated within a leanto porch addition such as the original building might well have had. In order to enclose the portion of the pantry known as the dish pantry, by means of which the Tavern is linked with the Alexander Purdie House, the porch element was extended some 19'-0" east of the east end of the Tavern proper and attached to the south facade of the Purdie House. This links of course, did not exist in the eighteenth century and it is a frank concession to the requirements of the present use to which the two buildings are being put.

ENCLOSURE OF PORCH TO PROVIDE MORE SPACE ON INTERIOR HAS COLONIAL PRECEDENT

The enclosure of the porch and its extension was necessary, of course, to make this area available for use on the interior of the building. Only the first bay adjacent to the east side of the wing 99 remains open to form a covered approach to the rear entrance door. However unauthentic the space enclosed may be the enclosing of the porch is a legitimate colonial treatment. Porches were frequently enclosed, in the eighteenth century, when the need for additional space within the house made itself apparent. An old instance of this is found in the James Galt House in Williamsburg, in which what was once a porch at the southeast corner was walled in either in the late eighteenth century or the early nineteenth. The example, though somewhat late, is, without question, representative of the colonial practice. This will be discussed following the description of the pantry leanto of the Tavern.

PORCH MADE DISTINGUISHABLE FROM EXTENSION BY DIFFERENCE IN DETAILING

The enclosed porch element extends about 11'-0" southward beyond what would have been the rear face of the structure if this part of the building had been carried out in an authentic manner. The extent of what purports to represent the porch has been made distinguishable from the unauthentic extension in several ways. One of these is the use of columns and half columns, spaced about 6'-0" on centers, on the porch part but not on the extension. Secondly, the enclosing material used here is random width tongue and groove flush boarding, recessed about 2 ½" back of the column faces, while weatherboarding similar to that found elsewhere on the two buildings, is employed as the external covering of the extension. In addition to these devices, the floor of the open entrance porch is extended as a projecting edge to the point where the King's Arms Tavern would normally have ended (the east front of the building) and there, along with the ogee backband and 7" high beaded skirting beneath, it is returned against the foundation brickwork.

PORCH HAS FIVE "FULL" POSTS AND ONE HALF POST

The porch has what are ostensibly five full columns or posts, 100 the spaces between which are filled with flush boarding, which is also carried across the east side of the open entrance porch. In actuality, of course, these "posts" are merely pilasters applied to a wood wall of modern construction. The face of the post which stands at the simulated east end of the porch element is partly covered by one of the two end boards which receive the weatherboarding of the extension. In addition to these would-be posts there is a half post applied to the east face of the Tavern wing. The half post forms the west terminus of the row.

DESCRIPTION OF POST

The posts are 8'-0" high and, if they were complete, would be square in section at the top and bottom, with 4 ¾" sides. They remain square to a height of 3'-2" above the floor and, at the top, to a point 4 ½" below the bottom of the porch cornice. The cap has a crowning cyma reversa topped by a fillet, the molding being 1 ½" high. Between the square cap and the base the edges of each post are chamfered so that this middle part becomes eight sided, the four sides which are turned on the diagonal being somewhat less wide than the others. The transition from the square capital to the octagonal shaft is effected by means of "lamb's tongues," curved surfaces which in profile, take the form of a cyma recta. The lamb's tongue starts at the edge formed by two sides of the cap and, following the cyma curve, PART OF POST HEAD AND BASE SHOWING CURVES WHICH JOIN CHAMFER WITH EDGE FORMED BETWEEN TWO SIDES OF POST 101 spreads to embrace at the bottom the diagonal upper edge of the chamfered surface. At the bottom of the chamfer the similar problem of joining the chamfer and the arris between two sides of the square base is solved more simply with the aid of a spheroidal curve.

CHAMFERS DIMINISH TOWARD BOTTOM LENDING REFINEMENT TO POSTS

It should be noted that while the cap and base have the same cross section, the chamfers, starting at the top with a width, measured on the diagonal, of 1 7/8" taper toward the bottom to become, at the point where the lower transitional curve begins, less than half the top width. In consequence of this the faces of the posts which are parallel with or perpendicular to the facade of the building diminish in the opposite direction. Thus, with the chamfers increasing in width toward the top and the other faces of the post decreasing, the former approach in width to with ½" of the width of the latter. This detail has an aesthetic rather than a practical basis, giving the post a refinement suggestive of the entasis of classic columns.

CORNICE OF LEANTO; DETAILS OF THIS

Above the columns is a cornice consisting of a fascia board, beaded at the bottom, and a crown mold of that frequently-used type which is composed of a cyma recta, a fillet and a cyma reversa. The roof shingles project beyond the crown mold and this, at the bottom, rests against the fascia. The cornice runs, without interruption, the entire length of the porch and the porch extension. At the east corner of the latter the fascia is stopped by the corner board and the crown mold by the rake board which runs along the inclined upper edge of the end elevation of the extension.

LEANTO ROOF; RELATION TO TWO MAIN ROOFS

The roof is of the single sloped shed type, having an inclination of about 17 ½°. It strikes the lower plane of the King's Arms gambrel at the level of the finished second floor, which lies a few 102 inches above the top of the main cornice of the building. It continues across the gap between the east edge of the Tavern roof and the west edge of Purdie's roof, striking the latter at the eaves level which is, in the Purdie House, slightly below the level of the finished, second floor. The shed roof then runs about 13'-0" to the end of the extension.

ROOF OF PART OF LEANTO BETWEEN THE BUILDINGS

The part of the extension which "flows" through the gap between the two buildings and runs as far as the middle of the Tavern chimney is covered, between the bottom of the south gambrel of the Tavern and the south side of the chimney, by a roof sloping very slightly in each direction from a center line. The remainder of the roof, from the south side of the chimney to the middle of the latter, slopes at an angle of about 30°.

WEATHERBOARDED WALLS OF PORCH EXTENSION

The roof of the pantry leanto is covered with the same type of round-butted shingles used on the main roofs of the Tavern and,Purdie's House, except for the nearly flat part between the two buildings, which has a covering of lead coated copper. The treatment of the 3'-0" width of north wall between the base of the Tavern chimney and the west face of Purdie's is the same as on the south elevation of the extension, i. e., it is covered with beaded weatherboarding and has a cornice similar to that on the south face, between the eastmost porch column and the east corner of the leanto and the eastmost porch column and the east corner of the leanto and the east end of that element, has a beaded lower edge and its exposure, 5 ½" to 6 3/8", corresponds with that of the weatherboarding of the Tavern and the Purdie House. The lower edge of the bottom weatherboard lines up with the face of the simulated floor of the Tavern porch and overlaps the foundation brickwork of the link element.

TREATMENT OF EAST END OF EXTENSION

The weatherboards are stopped at either end of the south face of the porch extension by single faced corner boards (3 ½" x 1 ¼"), the broader side of which face south. The gable end of the leanto is covered with weatherboarding throughout its entire height except, of course, where the window is located. It is received on the south end by the corner board, while it butts against the weatherboarding of Purdie's on the north. The rake board, which is beaded on the lower edge, tapers from a width of 5" at its outer extremity to 4" at the point of junction with the crown mold of the Purdie cornice. At its lower end it receives the crown mold of the cornice of the porch extension and this end is cut to the profile of the crown mold.

WALL COVERING OF ENTRANCE PORCH

The north wall of the rear entrance porch is covered with weatherboarding similar to that of the east side of the Tavern wing, a part of which forms the west side of the porch enclosure. The east wall has flush boarding similar to that used on the south elevation of the enclosed part of the porch.

BRICK PIERS SUPPORT LEANTO

The south side of the simulated porch is supported on five brick piers about 1'-6 ½" in width and the contiguous link element by similar piers somewhat higher than those of the porch since, as we have said, the bottom of the weatherboards and, hence, the sill, are here at the level of the surface of the porch floor. The piers of the porch part of the leanto are composed of brick similar to that of the Tavern foundations while Purdie foundation brick were used for the piers of the link element. For a description of these brick types see pp. 38 and 41.

WOOD SASH AND GRILLES CLOSE OPENINGS BETWEEN PIERS OF LEANTO

The spaces (about 4'-2" wide) between the piers of the porch element are closed by wood basement sash resting on a concrete foundation wall. These sash are five lights wide and the three light center 104 portion of them is hinged at the bottom to swing in. The sash are hidden from view from the outside by removable wood grilles set between the piers and extending from the skirting board to about 1" above the ground. The grilles are composed of horizontal and vertical slats, 2" x 5/8". the verticals running through and the horizontals consisting of short pieces wedged between them. Beneath the steps a grille, but no window, closes off the open space under the entrance porch.

WINDOWS AND GRILLES OF EXTENSION ARE LARGER; BRICK DRIP

In similar fashion the three openings between the brick piers on the south side of the link element and the one on the east side are provided with four-light wood basement windows, the lights of which are larger in both dimensions than those of the porch. The bottom-hinged part here embraces only the two center lights. These windows are also hidden from view by latticed grilles which are larger than the others since the openings between the brick piers here are both higher and wider than those beneath the porch proper. The wood windows used beneath the porch and its extension serve a practical purpose and have no eighteenth century basis. They are admissible here since they are hidden from view. A brick drip runs level with the surface of the ground from the base of the porch steps to the east end of the pantry extension, the bricks being perpendicular to the face of the building and adjacent to it. (See discussion of ground gutters and drips, pp. 43, 44.)

ENTRANCE PORCH STEPS; DESCRIPTION OF THESE

The entrance porch (5'-0" deep by 6'-0" wide) is approached from the ground by two riserless square-cut board steps supported between stringers 9" high and 2 ¼" thick, the two top edges of which are ornamented with a ½" bead. The treads are held in place by tenons which fit into mortises in the stringers. The outer edges of the treads advance somewhat beyond the sloping top surfaces of the stringer 105 so that, to eliminate sharp projecting corners these have been chamfered off, the chamfers being 1 3/8" wide and 10 deep. The steps, measured between the outside faces of the stringers are 6'-3" wide.

PORCH FLOOR

The floor of the open porch is continuous with the simulated floor lip which runs eastward to the would-be corner of the enclosed porch. At the sill of the recessed door the floor is 6 ¼" below the level of the finished floor within the building and it has been given a slight slope toward the outside.

SOUTH (LEFT) AND EAST PART ELEVATIONS OF JAMES GALT HOUSE SHOWING FORMER PORCH WITH ENCLOSURE OF FLUSH BOARDING BETWEEN THE CHAMFERED POSTS. TWO ADDITIONAL, OPEN PORCHES ARE ALSO VISIBLE WHICH HAVE A CORNICE TREATMENT SIMILAR TO THAT OF THE KING'S ARMS TAVERN LEANTO

ENCLOSED PORCH OF GALT HOUSE WAS MODEL FOR TAVERN LEANTO

The precedent for most of the details of the porch leanto which were discussed on the foregoing pages is found in Williamsburg in a similar enclosed porch at the southeast corner of the James Galt House. The exact period of this is not known and though it may stem from the 106 first quarter of the nineteenth century, the architects considered it legitimate to follow it as precedent in view of the fact that it doubtless represents the persistence of usages going back to the previous century.

GALT PORCH ORIGINALLY OPEN, WAS FILLED IN TO GAIN SPACE

The enclosed porch of the Galt House occupies an area 10'-0" x 12'-0" in size and houses, at present, an entry hall and a pantry. It was originally an open porch supported by four full columns and two half ones. To gain space within the house this porch was at some unknown period enclosed, random-width flush, unbeaded boarding being inserted to a depth of 1 ½" between the posts and nailed to studding placed between the columns. The studding, on the inside, carries the plaster which forms the finished surfaces of the walls of the interior spaces thus created.

SIZES OF PARTS OF GALT POSTS COMPARED WITH THOSE OF TAVERN COLUMNS

The column spacing of the Galt "porch" is about 3'-3" on the south face and 5'-3" on the east front which, though longer than the south face of this element (12'-0" as against 10'-0") has one less post. The columns are in character and in their cross-sectional size (4 ¾" x 4 ¾") exactly like those of the Tavern pantry leanto, though they are shorter than the Tavern posts (6'-7 ½" as against 8'-0"). The heights of the three component parts of the Galt posts, because of their lesser length, are necessarily somewhat less than those of the same parts of the Tavern posts, the caps, being 3 5/8" high as against 4 ½" for those of the Tavern; the bases 2'-9" as against 3'-2" and the chamfered length being 3'-6" as against 4'-3 ½". It will be noted that the relative proportions of the parts in both columns are similar. There is, however, a difference in the treatment of the chamfers of the Galt posts and those of the Tavern posts. When, as the Tavern chamfers diminish from top to bottom, those of the Galt 107 posts remain uniformly wide between the transitional curves at top and bottom. Other old Virginia porches, however, have posts of this type with chamfers which diminish toward the bottom. The posts of a porch of Little England in Gloucester County, although, compared with those of the Tavern these are crude in workmanship, have chamfers with the above characteristics.

DETAILS OF CHAMFERS AND TRANSITIONAL CURVES SAME IN GALT POSTS AS IN THOSE OF TAVERN

The full columns in the case of the Galt porch are, of course, actual full columns whose apparent depth has been reduced by the infill of boarding. These columns and the half columns have, like those of the Tavern, been chamfered off in such wise that the diagonal sides are somewhat shorter than the ones parallel with or perpendicular to the face of the building. The transitions from the square cap and base to the chamfers have been effected by means of lamb's tongues and spheroidal curves as in the case of the Tavern posts.

OPENINGS IN BOARDED WALLS OF GALT PORCH ARE PRECEDENT FOR SIMILAR OPENINGS IN TAVERN PORCH

The enclosed Galt porch has, in the boarded space between the two center columns of the south-wall, an eight-light window and, on the east side, a four-panelled door in the south bay and a four-light casement window in the other. Both door and windows, are replacements of old ones. These openings are legitimate precedent for the placing of openings in the boarded walls of the Tavern porch. (The latter openings will be discussed below.)

SKIRTING OF GALT PORCH THE BASIS FOR THAT OF TAVERN

The porch has an 8" skirting board, beaded at the bottom, beneath the overhanging edge of the porch floor. It is similar to the same feature of the Tavern leanto except that the cyma reversa at the top of the skirting is missing in the Galt example.

GALT PORCH CORNICE MORE COMPLEX THAN OURS; SIMPLER CORNICES ON SAME BUILDING MAY BE TAKEN AS PRECEDENT

The cornice of the Galt porch is continuous with the cornice of the south side of the house. It is more elaborate than the cornice of the Tavern leanto having, in addition to the crown mold and fascia 108 of the latter, an overhanging soffit which ends in a base mold on the building side. This porch cornice, then, since it is more complex than ours, will not serve as precedent. It will be noted in the drawings on p. 105 that there are, however, two other leanto porches on the east side of the Galt House. In these cases the cornice consists, like that of the Tavern leanto, of a crown mold and fascia only, so that these may properly be viewed as precedents for our cornice.

GALT PORCH FURNISHED MAJORITY OF DETAILS FOR TAVERN LEANTO

As we have seen the enclosed porch of the James Galt House provided the basis for the major part of the details of the leanto porch of the Tavern. We must still find the precedent supports of the porch and the lattice work which cloaks the openings between them and for the porch steps. Thereafter we will discuss the windows and doors of this element.

PIERS OFTEN USED AS PORCH SUPPORTS IN EIGHTEENTH CENTURY

Porches of any size were more likely than not to be supported on brick piers in eighteenth century Virginia, leaving the space between them entirely open or provided with some type of perforated screen so that air could circulate under the porches. Many porches are still built this way in Virginia. Old examples of this method of construction have been found and photographed by Colonial Williamsburg architects. An old house with a leanto porch running its entire length was photographed by Singleton P. Moorehead in Smithfield in 1930. This porch was supported on brick piers, the spaces between which were left open or, at least, were found open when the house was examined. Little England in Gloucester County has a much smaller porch supported on the outside by two brick piers. In this case also the openings between the piers remain free of any sort of grillework. It should be mentioned in this connection that whole buildings were in colonial times often 109 supported on piers, as small structures are today, leaving the space beneath open to the outside or partially closed by screening. An example of this construction in Williamsburg is found in the Tayloe Office. The spaces between the brick piers in this case have been filled in with boarding to make the building, which is used for living purposes, more immune to temperature changes.

PRECEDENT FOR USE OF LATTICEWORK BETWEEN PIERS

Since the space beneath the Tavern porch was needed for use as part of the basement kitchen, the steel sash mentioned above were inserted between the piers. These were unauthentic so that it was deemed wise to mask them in a way which would not exclude light and air. Wood latticework was the logical answer to this since this was frequently used in the eighteenth century, where a screen to obstruct the view but which would not prevent the circulation of air was needed. Among other places these were used in the upper part of the walls of well enclosures. An old example of this use of latticework is seen in the well head on the Captain Orr property, the parts of which are either original or are exact replacements of original features. Other similar latticework examples are visible in the photographs taken by Colonial Williamsburg architects of an old well head in King William County and another in the picture of a similar structure in Edenton, North Carolina. The latticework in all of these cases consists of slate diagonally placed rather than horizontally as in the Tavern examples, but the principle involved is the same.

EXAMPLES OF GRILLEWORK BETWEEN PIERS OF PORCHES IN NORTH CAROLINA

A number of different treatments of the open spaces between piers supporting porches and, for that matter, houses themselves, may be seen in photographs reproduced in The Early Architecture of North Carolina by Johnston and Waterman. Among these are several types of wood latticework, including vertical, diagonal and others and also 110 perforated brick screens. In some cases, incidentally, the piers are connected by brick arches. It should be stated in closing this subject that the open barred grilles used so frequently in Virginia in the eighteenth century in lieu of basement windows served the same general purpose as the latticework which we have discussed, i.e... admission of air and interruption of view.

BASIS FOR PORCH STEPS

Riserless steps with treads supported by simple stringers cut from heavy lumber are found at Little England in Gloucester County. The stringers here have rounded top surfaces without beads. Many other examples of steps of this same general character have been observed in various parts of Virginia.

OPENINGS IN WALLS OF LEANTO

The Tavern leanto has six openings above the foundation level, viz., the door, in the north wall of the entrance porch, two windows inserted in the boarding of the first and third bays (counting from west to east) of the south front of the enclosed "porch"; two windows similar to those in the south face of the weatherboarded extension and one in the east end of the latter.

ENTRANCE DOOR IS PANELLED AND HAS GLAZED "PEEP" HOLES

The door, 7'-0" high and 3'-0" wide, has four equal rectangular panels about 2'-1 ½" high and 1'-1" wide. Above these are two panels 10" high by 1'-1" wide which contain two circular glazed "peep" holes. The panel moldings are, on the outside similar to those of the front (north) entrance door but on the inside the moldings are omitted so that we have only square-edged, recessed rectangular panels. The stiles and rails of this door, like those of the front door (p. 65) are set together with pegged mortise and tenon joints.

DOOR FRAME AND SILL

The outside door frame (5 ½" high) is much simpler than that 111 of the front door, having a typical colonial profile consisting at the top of a cyma reversa backband followed in succession by a fascia, a smaller cyma reversa, a smaller fascia and a ½" bead. The sill is of the square cut, unmolded type, the ends of which line up with the outer edges of the frame.

DOOR HARDWARE

The door hardware consists of one pair of wrought iron HL hinges, 10" high, which are reproductions of a colonial model and similar to CW F-3 and an iron rim lock (Reading No. C-625) with brass knobs, a rose and a swivel spindle.

PRECEDENT FOR THIS DOOR IS OLD ONE AT GREENWAY

This door, with its pair of squarish panels above two pairs of similar rectangular panels, is one variant of a type within which there were several different combinations of the basic panel shapes. This particular combination is found, among other places, in the east entrance door of Greenway in Charles City County. The Greenway example like the Tavern door, is molded on the outside only, the inside face having only sunken panels. The quarter round mold of the stiles and rails of the Greenway door and the sloping tongue of the panels are similar to these details in our door.

DISCUSSION OF GLAZED DISCS OF TAVERN DOOR

Though not common, doors with glazing were occasionally used in the eighteenth century. An example of this is the old nine-light glazed door in the restored Nicolson Shop in Williamsburg. Glazing of the sort used here -- glass discs inserted in the panel faces -- was not however, to the best of our knowledge, used in colonial times. This represents a concession to modern utility, the glazing introducing some natural light into the passage and reducing the chance of accidents in this much-used doorway by affording a person about to pass through it a view of someone who may be approaching from the other side.

WINDOWS OF LEANTO; NUMBER OF LIGHTS; SASH AND FRAMES

The windows of the leanto porch and those of the weatherboarded extension are alike, having twelve lights (four high and three wide) and being divided into two movable sash of six lights each. They are 4'-1 ¼" high and 2'-4 ½" wide and have glass panes 10 ¾" high and 8" wide. The detailing of the sash of these windows is similar to that of the windows of the north facade (pp. 70-72). The exterior frames of the windows are of the simple block type described on p. 72. The two windows in the sheathed wall are recessed about 2 ½" more deeply than those of the weatherboarded part of the leanto, due to a difference in the construction of these two parts of the leanto wall.

WINDOW SILLS; PRECEDENT FOR THESE

The sills, in contrast to those on the north front, are of the simple, unmolded block type the ends of which line up with the outside edges of the window frames. In the case of both the flush-boarded and the weatherboarded walls the forward projection of the sills is minimal, being 3/8" in the case of the former and 9/16" in the latter. The sills lack the customary bead on the bottom outside edge, probably because the slight overhang did not lend itself to its use. Old examples of block sills, with and without beads along their lower edges, are found in the Brush-Everard House (see diagrams of these on p. 55 of the architectural report on this house).

THREE OF FIVE LEANTO WINDOWS HAVE SHUTTERS

Only three of the five windows of the pantry leanto have shutters, viz., the three windows of the weatherboarded extension. There is no logic in the omission of shutters from the two windows of the enclosed porch since the need for them is as great here as in the neighboring part of the leanto. Practical necessity, apparently, dictated the omission of the shutters in the two flush-boarded bays since their width is such that had shutters been used they would have struck, when opened, against the caps of the chamfered posts.

DESCRIPTION OF SHUTTERS

For the reasons given on pp. 73-75 louvered shutters were used on the leanto. These shutters which, of course, are in pairs, are divided into two nearly equal parts by a center rail. The framework, which is mortised, tenoned and pegged together, is relatively sturdy, with 2 ¾" wide stiles, 3" top and center rails and a 3 ¾" bottom rail. The slats are 5/16" thick and, on the side which is out when the shutters are closed, are cut flush with the framework. On the side which is seen when the shutters are open the slats project about ¼" and have half-round ends. The slats are slanted at an angle slightly less than 45° and the direction of the slant is such that the slats expel rainwater to the outside when the shutters stand before the windows.

PRECEDENT FOR LOUVERED SHUTTERS

louvered shutters, divided in the middle, are numerous in Virginia. It is not always easy, however, to determine which were made in the eighteenth century and which represent replacements of shutters made then. The following old houses have louvered shutters which have the appearance of being eighteenth century examples: Kittewan, Charles City County; the Wigwam, Amelia County and Belle Farm, Gloucester County.

SHUTTER HARDWARE

The shutter hardware here is the same as that of the shutters of the north elevation (see p. 74).

CHIMNEY WHICH SERVES FIREPLACES OF NORTHWEST DINING ROOMS

We will discuss at this point the chimney which serves the fireplaces of the northwest dining rooms on the first and second floors. This is most closely related to the south front, probably, since, in the eighteenth century, before the southwest wing was built, the chimney would have appeared prominently in this elevation. As things now stand, because of the existence of the wing and the enclosed pantry leanto the chimney is set fairly deeply within the area of 114 the house and rises out of the roof near the point where the top of the lower gambrel slope of the wing meets that of the main building no part of it showing below the level of the bottom of the upper gambrel slope. (See south and east elevations, pp. 26 and 27, respectively).

CAP DESIGN SIMILART TO THAT OF EAST CHIMNEY

This chimney contains two square flues only and is, consequently, rectangular (2'-4" x 3'-10 ½") in section above the roof line. The maximum height of the shaft above the roof is 11'-9". The cap is of the same design as that of the east chimney (p. 87).

WEST ELEVATION

GENERAL; COMPOSITION OF FACADE

The west elevation of the main element of the King's Arms Tavern (see drawing, p.27) is, if we except the fact that a wall of the basement bulkhead strikes against its southwest corner, completely regular. It has two windows each on the first and second floors and a pair of basement openings, all of which are arranged symmetrically about the vertical center axis of the gable end. Furthermore, in each of the two vertical tiers of three openings the windows are grouped symmetrically about a common center line.

SECOND FLOOR END WINDOWS LINE UP WITH DORMER WINDOWS IN VIRGINIA HOUSES OF EIGHTEENTH CENTURY

It seems appropriate at this point to explain a practice so customary in eighteenth century Virginia architecture that it almost amounts to a principle, viz., that the heads, and frequently, also, the sills of the second story end windows in gambrel and A roofed houses are made to line up with these parts of the dormer windows on the long sides of the house. There are, of course, exceptions to this rule. The prevalence of this practice, in any case, is a great aid in the location of the gable windows in relation to those of the dormers and by this token, also, the location of the first floor windows of the ends is facilitated. It should be added that in both the King's Arms Tavern and the Purdie House the heads and the sills of the end windows of the second story line up with these parts of their respective dormer windows.

Most of the features of the west elevation have their counterparts on the three elevations already discussed so that in many cases we need only to refer back to matters already covered.

CORNICE END BOARDS AND RAKE BOARDS

The west elevation has cornice end boards and rake boards which are similar to those on the east facade. (pp. 91-96).

FIRST FLOOR WINDOWS AND SHUTTERS

The two 28-light windows of the first floor are similar in all of their elements (sash, frame, sills) to the four first floor windows of the north facade (pp. 68-72). The three-panel shutters with their hardware also have exact counterparts on the north facade (pp. 73-74).

SECOND FLOOR WINDOWS

The two second floor windows, the opening size of which is 2'4 1/[cut off] x 5'-0 ½", have 15 lights and are three lights wide and five high. They are composed of two movable sash, the top one of which is two lights high and the bottom one three lights high. The detailing is similar to that of the first floor windows. Old 15-light windows in which the lights are divided in the above manner between two sash were not uncommon in the eighteenth century. The two second floor windows of the west facade of the Tayloe House are an example of 15-light windows with this distribution of panes. These are old windows which have been repaired.

SILLS OF SECOND FLOOR WINDOWS

The sills of the second floor windows are of the block type which signifies that there are no moldings on the upper outside edge as there are on the first floor sills, the sloping upper surface being joined with the vertical front face by a simple curvature. The sills are approximately 2 ¾" thick and project about 5/8" beyond the weatherboarding, the lower outside edge being cut to form a ½" bead. A groove about 5/8" wide and deep receives the top of the weatherboard directly beneath the sill. The weatherboards laterally adjacent to the sill are so placed that their beaded lower edge is at the same level as the bead running along the lower edge of the window frame. The Brush Everard House has several old sills of the same type as the Tavern sills described above.

SECOND FLOOR SHUTTERS

The shutters of the second floor windows are of the same panelled 117 type as the ones of the first floor but they are smaller (each leaf is approximately 1'-3 ½" x 5'-3"). The leaves each have two panels only and these are rectangles of the same size. The panel moldings and construction are similar to those of the first floor shutters. Two-panelled shutters are a common enough type. A pair of old ones is visible in a photograph made by Colonial Williamsburg architects of a story and a half house near the village of Crocker in the Williamsburg vicinity.

SHUTTER HARDWARE

The hardware of the shutters of this facade is the same as that used on the shutters of the north elevation (p. 74).

BASEMENT GRILLES

The two basement openings are provided with open grilles and also, behind these, wood sash. These features are identical with the pair of grilles and sash of the north elevation (pp. 76-78).

VENTILATOR IN GABLE; CONSTRUCTION OF THIS

Near the peak of the gable of this facade is located an attic ventilator approximately 2'-0" high and 3'-6" wide. This was introduced to make possible the circulation of outside air throughout the attic space. The vent has been rendered well nigh invisible when viewed from the ground by the manner of its construction. The air enters through slots created between vertically adjacent weatherboards by inserting 1" x 2" blocks at intervals of 16" between the lower part of the inside face of the upper weatherboard and the upper part of the outside face of the lower weatherboard. To accomplish this without making the weatherboards in the vent area spring out beyond the plane or planes of the surrounding weatherboards the parts of the studs fall within this area were cut back about ½" so that the tops of the weatherboards which are nailed to them could be tipped inwards. This caused a warping of these weatherboards out of line neighbors, that is, they slope inward more than the latter. The exclude 118 GABLE END (ABOVE) OF OLD ICE HOUSE AT MARMION, KING GEORGE COUNTY, AND (BELOW) SECTION THROUGH THE GABLE END. THE TWO VIEWS SHOW THE MANNER OF PROVIDING VENTILATION IN THE STRUCTURE BY BLOCKING OUT THE TOP SIX LENGTHS OF WEATHERBOARDING TO FORM LOUVRES THROUGH THE SLOTS OF WHICH THE AIR CAN PASS. insects from the interior of the house bronze screen wire was nailed to the studs back of the tipped weatherboards. This ventilator has its precedent in the similar treatment of the upper weatherboards in the gable ends of the old colonial ice house at Marmion (see illustration).

BULKHEAD OF BASEMENT STAIR STRIKES SOUTHWEST CORNER OF MAIN BUILDING

As we noted above, the bulkhead housing the outside stair to the basement strikes the southwest corner of the west facade of the building (see elevations, pp. 26,27), overlapping the latter about 9" and advancing about 3'-6" in front of it. Although it stands for most of its width of 7'- 4 ½" in front of the north end of the west facade of the Tavern wing, we will discuss the bulkhead in conjunction with the west elevation of the building proper and then proceed with the treatment of the wing.

BASEMENT STEPS BUILT IN LOCATION OF OLD ONES DISCOVERED ON SITE

The basement steps and their enclosure are located over the site of the foundations of an eighteenth century basement stair. These may be seen in the archaeological plan drawing, p. 39, and in three of the photographs on folding p. 40. The reason why the bulkhead overlaps the main building and why it stands 3'-6" in front of it is evident from an examination of the old foundations in the plan. Although the steps were fragmentary it is apparent that four or five risers led to an areaway, which, as the plan indicates, was paved with brick and the entrance to the basement was, evidently, in the western end of the south wall of the building. This stair, it seems, became necessary when the west enlargement was made to the building. Though the present reconstructed bulkhead is in the old location, the cellar entrance is in the west wall of the wing and the steps have many more risers and are completely modernized. The architects attempted to maintain authenticity only in the external features of the bulkhead.

SIDE WALLS WERE BUILT OF BRICK ON BASIS OF FOUNDATION THICKNESS; FRONT WALL MADE OF WOOD

We have no way of knowing positively whether the side walls of the bulkhead were of wood or brick. Eighteenth century bulkheads were built of both of these materials. The thickness of the foundation walls (about 18") led the architects to believe, however, that the walls were of brick since so small a structure, had it been built of 120 wood, would have required no more than an 8" foundation. A case in point is that of the 8" bulkhead foundations which were discovered adjacent to the east wall of the Tayloe House. In that instance the architects decided that the foundations indicated that the bulkhead walls had been of wood and, accordingly, built them of this material. On the same principle they constructed the lateral walls of the Tavern bulkhead of brick because the 18" foundation walls suggested that they had been of masonry. The vertical face of the bulkhead containing the doors, with the exception of the west foundation, which is of brick, was made of wood, the most feasible material available in the eighteenth century for a wall occupied to a great extent by doors.

DESCRIPTION OF BULKHEAD; ROOF AND MASONRY WALLS

All of the exterior details of the bulkhead were designed with a careful regard for authenticity. The structure is covered by an A roof of low slope (about 20°) which creates a gable on the front or western face. The roofing material is round-butted asbestos cement shingles of the same size and type used on the main roof of the Tavern. This roof is supported by the two brick walls mentioned above. These walls are 9" thick and, as we have said, the north one is 3'-6" in length. The south wall is 4'-9 ½" long and this discrepancy in the length of the two sides of the bulkhead is explained by the fact that the south wall runs back to the west wall of the wing which is set back of the west face of the main building. The two brick side walls and the low foundation wall which runs across the west end of the bulkhead and joins these are supported by concrete foundations. The latter remain invisible since they are covered by a 4" facing of brickwork which runs about 1'-6" below grade. The brick walls are laid up in English bond with a mortar made to resemble old oyster shell mortar in color and texture.

BULKHEAD; WOOD FRONT (WEST WALL)

The wood facing of the west front consists of a double valve door the top edge of which is rounded and a frame holding this. The headpiece of the frame, made of 1 1/8" stock, has two sloping top edges which follow the roof slopes and a lower edge which is cut to the curvature of the doors. This curvature is the arc of a circle having a radius of 4'-1". Superimposed upon the arch board are two rake boards which taper from a width of 4 ¼" at the outer end to 3 ½" at their meeting point. The lower edge of these has a ½" bead and the outer end terminates in a flattish ogee-curve. The curve extends out far enough so that the squared-off end of the piece covers a similar feature of the 5" high crown mold which runs along the eaves on the side elevation.

DESCRIPTION OF DOUBLE VALVE DOOR OF BULKHEAD

The two vertical sides of the frame are made of 11 5/8" wide boards both of whose vertical edges are beaded. At the points of junction with the ends of the top piece the bead on the inner edge of each vertical board is discontinued and carried around the curved lower edge of the headboard. The vertical sides of the frame are secured to a pair of studs set against the inner faces of the brickwork of the side walls and they are also anchored to the ends of the brickwork which they completely cover. The anchorage is by means of 1 ¾" screws which are screwed into wood plugs set in brickwork. The screw heads are recessed about 3/8" and covered by round wood plugs.

DESCRIPTION OF DOUBLE VALVE DOOR

Each valve of the door is 2'-9" wide and 5'-1 ½" high at the crown of the curve. The two leaves meet in a shiplapped joint, the "lip" of one overlapping member of the joint being cut to form a ½" bead. The vertical boards of which the doors are made also meet in shiplapped joints. The doors run to within 2" of the ground, overlapping the brick foundation which rises above the ground a little 122 over 2" and acts as a sill for the doorway. Each leaf on the door has on the inside face two horizontal wood battens with beveled edges. These are held to the vertical boarding of the doors by nails with hand hammered heads which are driven in from the outside and clinched on the face of the battens.

DOOR HARDWARE

Each valve swings on two wrought iron strap hinges about 2'-4 ½" in length. These are shown on p. 11 of the Colonial Williamsburg hardware folder where they are designated as Type 3. The folder states that Type 3 strap hinges are reproductions of an old colonial hinge found on the door of the old smokehouse on the Moody lot. The other items of hardware on this door are as follows: two wrought iron reproductions of surface bolts similar to Colonial Williamsburg hardware folder, Type 2, Plate 8; a wrought iron reproduction thumb latch, CW F-14, with latch, F-17, as well as a rim night latch 356RB with a wrought iron cover plate.

GREENWAY HAS BULKHEAD SIMILAR TO ENCLOSURE OF TAVERN; DESCRIPTION OF THIS

An old bulkhead of Greenway, Charles City County is very similar in principle to the Tavern bulkhead even though it differs from it in certain ways. The Greenway example forms an "ingle recess" or lean on the west side of the house between the south side of the chimney and the southwest corner of the building. Its shed roof is wood shingled. Its depth is considerably less than that of the Tavern bulkhead but its width is comparable to that of the latter. The flanking walls are of brick (one, of course, is the chimney wall) and the brickwork is revealed on the west front of the bulkhead. The wood enframement of the opening and the single valve wood door are inserted between the brickwork of the south side of the chimney and the brick south wall of the bulkhead. The lower edge of the top frame member follows a curve similar to that of the Tavern "arch". The door head necessarily 123 has the same curvature. The door is built up of random width flush boards which are, however, unbeaded. The door hardware, two strap hinges and a hasp, appear nondescript and are probably later replacements.

BULKHEAD AT TUCKAHOE

Another upright basement bulkhead which has certain of the characteristics of the Tavern bulkhead is an example at Tuckahoe in Goochland County. This also combines brick side walls with a wood front and it has an A roof. The entrance opening and the pair of doors are, however, square headed and the gable is treated as a pediment.

NEW ORLEANS LAMP AT SOUTHWEST CORNER

Before taking leave of the west facade we should not omit to mention that, attached by iron straps to the southwest corner, with its base about 8'-0" off the ground, is an original New Orleans type lantern about 3'-0" high and wired for electricity (see photo of wing, p. 129). The lantern, from the Colonial Williamsburg warehouse stock, is a wrought iron and glass fixture which once did service as a street lamp in the Crescent City. With several others, it was purchased years ago in New Orleans. Although it may not be an actual eighteenth century lamp, in the character of its design, it is entirely representative of that period. It serves at present to light the passage between the west facade of the Tavern, on the one side, and the King's Arms Barber Shop and Kitchen, on the other.

SOUTH WING

MATERIAL ON LOCATION AND SIZE OF WING TO BE REVIEWED BEFORE TREATMENT OF FACADE DETAILS

We will review the evidence bearing upon the size and location of the wing before proceeding to a treatment of the details of its three facades. It is recommended at this point that the material on pp. 16-18 dealing with the two insurance policies taken out on the King's Arms Tavern property by Philip Moody in 1796 and 1805 respectively, be reexamined. Both of the plats accompanying the policies show a southward extension of the main building and in each case this is attached to the east rather than the west side of the south elevation. We are confident that this reversal of the position of the wing is an error and that it should have been shown on the west side (see p. 17).

DIMENSIONS OF TAVERN AND WING GIVEN ON MOODY PLATS

In the policy of 1796 no dimensions are given for the wing but in the later one of 1806 the following handwritten notation appears on the outline plan of the wing: "a wood wing 57 34 feet." At the same time, the dimensions of the main element of the building are given as "57 by 24 feet." If we were to take these dimensions at their face value, the wing would have been the same length as the main part of the building and 10 feet wider, so that the tail really would have wagged the dog.

PLAN SIZE OF MAIN BUILDING AS REVEALED BY FOUNDATIONS IS IN CLOSE AGREEMENT WITH THAT GIVEN IN PLATS & WITH THAT OF TAVERN AS REBUILT

Let us consider for a moment the size of the old foundations discovered on the site (see archaeological plan, p. 39). These are, as is usual with such old walls, irregular in their contours so that only approximate measurements can be given. The width of these foundations, measured from outside of wall to outside of wall, varies from slightly over 24'-0" to 24'-6" or thereabouts. The length, taken along the longitudinal center line, is about 54'-3". if however, we add to this 3'-6", the distance the bulkhead foundation extends west 125 of the west face of the building, the total length of the old foundations of the King's Arms Tavern is 57'-9". Both the width and the length, taken in the manner just described, are surprisingly close to the 57'-0" x 24'-0" given by both of the Moody insurance plats as the dimensions of the Tavern. The present length of the building is 53'-9" without the bulkhead and 57'-3" with it. Its reconstructed width, exclusive of the pantry leanto, is 24'-0". It is evident, therefore, that the dimensions of the old foundations, the insurance plat dimensions and the actual size of the reconstructed building are very closely in agreement.

NO FOUNDATIONS OF WING WERE DISCOVERED; A COLONIAL WELL FIXED THE LIMIT OF ITS SOUTHWARD EXTENSION; THE WELL HOUSE

The matter which here concerns us chiefly is the determination of the basis for the present size of the wing in the light of documentary evidence which seems to indicate that it was once wider tan it now is. It should be stated in this connection that no foundations whatever which might have appertained to this wing were discovered. Since we know that the wing existed the foundations, which may have been shallow ones, must have been removed sometime after the building burned (p. 18)^[*] It was of considerable importance to the architects, however, that the foundations of a colonial well were discovered about 26'-0" south of the face of the south foundation wall of the building and about 3'-6" east of the line of the face of the west foundation wall. The existence of this well set a limit to the southward extension of the wing since [it] is unlikely that the well would have been within the wing itself. Under this circumstance the possible southward extension of the wing proved to be 24'-8". This brought the well within 1'-4" of the face of its south wall so that it seemed reasonable to enclose this in a Well House 126 which would abut against the south face of the wing. As it eventuated, the Well House extended 8'-5" south of the south face of the wing.

TOTAL LENGTH OF WING AND ATTACHED WELL HOUSE, ADDED TO WIDTH OF MAIN PART OF BUILDING, GIVES DIMENSION FOUND IN PLAT

The total extension of the wing and the Well House south of the south wall of the building proper thus became 33'-1" (24'-8" plus 8'-5"). If we add this dimension to the width, 24'-0", of the main part of the building we get a total length of 57'-1". The insurance plat of 1806 speaks of a wing 57'-0" x 24'-0". It appears, therefore, that the maker of the plat considered the length of the wing to be the sum of the width of the main part of the building and the distance the wing extended to the south of this. It is not unusual in these insurance policies to find the measurements of a part of a building taken in this unorthodox way.

WIDTH OF WING AS REBUILT DIFFERS FROM DIMENSION GIVEN ON PLAT

A discrepancy, however, immediately becomes apparent when we compare the width of the wing as given in the plat (34'-0") with its width as it was reconstructed, which is 20'-0". If we accept the other dimensions of the Tavern given on the plat as approximately true ones, we should, for consistency's sake, also accept this width of 34'-0". Consistency would, again, demand that we reckon the present width of the wing as the total of its east-west dimension and the extension of the basement bulkhead (4'-9 ½") west of its west face. The total width then becomes 24'-9 ½", or 9' 2½" less than the 34'-0" given on the plat. This discrepancy, evidently, is too great to be the result of inexact measurement on the part of the platmaker and we are uncertain as to the reason for it.

WEST FACE OF WING SET IN FROM WEST END OF MAIN BUILDING FOR REASONS OF AESTHETICS

It will be observed that the architects, in building the wing, set its west face 1'-3 ½" east of the west end of the building. Aesthetic considerations, presumably, dictated this setting back of the wing for, had this not been done, the fronts of the west dormers 127 of the wing, part of the lower gambrel slope and the entire cornice would have projected beyond the west gable end of the main part of the building. This would have been a clumsy solution and it is quite unlikely that an eighteenth century builder would have considered joining the two parts of the structure in this way.^*

ARCHITECTS DERIVED WING WIDTH BY PROJECTING SOUTHWARD THE LINE OF THE OLD TRANSVERSE WALL

Since the wing is set 1'-3 ½" in from the west end of the building, the wing extends 21'-3 ½" east of this end (1'-3 ½" plus width of wing, 20'-0"). If, on the archaeological plan, we measure eastward from the outside face of the west foundation wall a distance of 21'-3 ½" we come approximately to the middle of the transverse wall which divides the new part of the basement from the old. It is evident from this that the architects used this wall in determining the width of the wing, that is, they apparently carried the east wall of the latter southward at a point such that the west interior face of the east wall would, approximately, line up with the position of the west face of the transverse wall. In the absence of evidence pointing to a wing of greater width than 20'-0", this would be a sound method of establishing the wing width since it would be the part of orderly planning to extend this existing partition southward as the east wall of the wing.

AN ALTERNATIVE METHOD OF DIMENSIONING THE WING; THE FIGURES THUS DERIVED AGREE WITH THOSE OF PLAT

There is one further possible interpretation of the meaning of the dimensions, 57'-0" and 34'-0", which are given for the wing in the insurance plat. Just as, in obtaining a length for the wing of 57'-0" we added to the actual wing length the width of the main building, so, 128 in deriving the width we might assume this to be the entire length, 57'-9", of the building proper. This would be very close, of course, to the larger of the two wing dimensions given on the plat. If, in addition to this, we assume the shorter dimension of the wing to be its length of 24'-8", as given on p. 126, plus the southward extension of its attached Well House (8'-5") we derive the figure 33'-1". This again, is sufficiently close to the 34'-0" given as the smaller of the two dimensions of the wing to be acceptable. In this manner the two plat dimensions and the present wing sizes can be brought into agreement. This last hypothesis may seem far-fetched and yet it departs no further from conventional methods of taking measurements than the case in which we add the width of the main part of the building to the length of the wing to derive one dimension of the latter.

UNCERTAINTY STILL EXISTS AS TO SIZE OF WING IN 18TH CENTURY; THE ARCHITECTS CHOSE THE MOST LIKELY DIMENSIONS

From the above detailed discussion of the various possible methods of determining the dimensions of the wing, it is apparent that we are still uncertain as to what these dimensions in actuality were. After weighing the considerations we have reviewed above, the architects determined upon the present net dimensions, 20'-0" x 24'-8", as being as close to the true dimensions as, in the incomplete state of our information concerning them, it was possible to come.

TAVERN WING FROM SOUTHWEST, SHOWING ATTACHED WELL HOUSE WITH RUSTICATED SHEATHING

DETAILED ANALYSIS

ROOF, SHINGLES, WEATHERBOARDS

The mansard roof of the wing has been discussed at some length (pp. 32, 33) so that we will not concern ourselves further with it here. The shingle roof covering is the same as that of the main part of the building. The weatherboarding of all three sides of the wing is identical in character and treatment with that of the main structure (see pp. 34, 35). The exceptional external wall covering of the well house will be dealt with in the course of the discussion of the south elevation of the wing.

FOUNDATION BRICKWORK

The foundation brickwork of the wing proper rises to the same level as that of the main part of the building and it is laid up in English bond like the latter. The brick and mortar used are similar to those of the main structure (see pp. 36, 38, 41, 43 for general discussion of the foundation brickwork of the Tavern).

EAST ELEVATION

GENERAL

The shed roof of the south leanto strikes against this facade (see elevation drawings, pp. 26, 27) and continues its contact with it for a distance of some 11'-0" from the line of what would be, were it not for the presence of the leanto, the south face of the south wall of the main structure, cutting off a part of the main cornice of the wing. The outer 5' -0" of this distance is occupied, on the south elevation, by the open entrance porch and the balance by a vestibule (Passage #3 on first floor plan, p. 197) which eliminates the weatherboarding which would have been placed on this wall had it remained a part of the east exterior wall of the wing. A half post or pilaster 131 of the porch leanto is also applied to the east elevation of the wing (p. 100).

DORMERS AND DORMER WINDOWS SIMILAR TO THOSE OF MAIN BUILDING

There are on this elevation two identical dormers which penetrate the lower slope of the gambrel, the cornices of which line up and are continuous with the upper cornice between the two slopes of the roof. These dormers are of the pedimented variety, like those of the north and south elevations of the main structure. They are, in fact, similar in every respect to these other dormers except in the depth of the sides and the roofs which the somewhat steeper slopes of the wing gambrel have made less than in the dormers of the main roof. Except in the last respects, therefore, the analysis of the north dormers on pp. 46-48 will hold equally well for these.

CORNICES ALSO SIMILAR TO THOSE OF MAIN PART OF BUILDING

The main cornice is of the modillion type and it is similar to that of the north elevation of the building. The remarks made on pp. 48-50 concerning the north cornice and the wing cornice cover the latter in sufficient detail and the reader may refer back to these. The upper cornice is also like that of the main gambrel roof. For a discussion of the latter, see p. 47.

BRICK DRIPS

As in the case of the main part of the building there are no metal gutters and leaders on the wing, the water being allowed to run down the roof slopes and drip to the pavement. The latter runs up to the brick foundation on all sides of the wing and Well House, with the single exception of the eastern four feet or so of the south wall of the wing, where a band of planting touches the building. Here a brick drip two brick lengths wide placed in the ground adjacent to the building receives the rainwater which runs off the roof. For a discussion of the brick drips and ground gutters of the King's Arms 132 Tavern which lists the precedent for these, see pp. 43, 44.

WINDOW OF FIRST FLOOR HAS 21 LIGHTS

There is a single 21-light window (seven lights high and three wide) on the east face of the wing and this is locateD on the vertical axis of the south dormer of this elevation. The window is 6'-11 7/8" high and 2'-5 ¼" wide and it is divided into two movable sash, an upper one of nine lights and a lower one of twelve. The construction and details of the sash and window frame and the size of the glass panes are similar to those of the first floor windows of the north elevation, for a discussion of which see pp. 70-72. The sill is of the block type, similar to those of the pantry leanto (pp. 109, 110).

PRECEDENT FOR THIS WINDOW

The elongation of this window is somewhat exceptional, windows seven lights high and four wide being more usual than those seven high and three wide. Examples of the latter can be found in eighteenth century Virginia buildings, however, and an instance is the windows of an outbuilding in Stratford, Westmoreland County. In this case, how ever, it is the lower sash which has 9 lights and the upper 12.

SHUTTERS

The above window is equipped with a pair of louvered shutters, each leaf of which is 6'-11 7/8" high by 1'-3 ¼" wide. Except in their height and number of slats these shutters are exactly similar in detailing, construction and hardware to the shutters of the three leanto windows of the south elevation of the main house. For a discussion of the latter and their precedent, see p. 113).

CORNER BOARDS

This elevation has at the south end one of a pair of 1 ¼" wide cornerboards, the other being around the corner on the south elevation. These are in every way similar in design to the corner boards at the northwest and southwest corners of the main building (p. 79).

RAKE AND CORNER END BOARDS

There are on the wing no rake boards or cornice end boards since 133 these are features found only on gable-ended structures, whereas the roof of the wing is a mansard.

FOUNDATION

For material on the foundation brickwork see p. 130 and pp. 36, 38, 41 and 42.

SOUTH ELEVATION

DORMERS AND DORMER WINDOWS

There are two dormer windows on this elevation, symmetrically located in respect to the central, vertical axis of the elevation. These are exactly similar to the pair of dormer windows on the east elevation and to those on the north and south elevations of the main building. For a discussion of the north dormers, see pp. 46-48.

CORNICES

The main and upper cornices are identical in character with the cornices of the east elevation of the wing and of the north elevation of the building proper (for the latter, see pp. 47-50).

WINDOW; SHUTTERS

There is a single 21-light window on this elevation provided with slatted shutters. This window is in the eastern half of the facade and is centered on the central vertical axis of the dormer above. The window and the shutters are similar to those on the east elevation of the wing, q. v., p. 132.

FOUNDATION WALL

For material on the foundation brickwork see p. 130 and pp. 36, 38, 41 and 42. This foundation wall has in it a ventilation duct, 1'-2 ¾" high by 5" wide, placed on the axis of the first floor window above. The opening is covered on the outside by a framed bronze screen. This feature has its precedent in old ventilation openings of the same character which are located below the watertable in the east and west walls of the Court House of 1770 but there is no colonial precedent for the use of wire screening with them.

CHIMNEY; DESCRIPTION OF THIS

One of the three chimneys of the Tavern rises through the upper south slope of the mansard roof of the wing about 2'-3" (measured horizontally) from the junction of the upper slope with the lower. This stack serves two fireplaces, located on the south sides of the dining rooms of the first and second floors, respectively, of the wing. The chimney houses the two flues for these fireplace and a ventilation 135 duct. The shaft rises to a height of about 7'-0" above the ridge of the wing roof, which brings its crest to the level of that of the other two stacks of the Tavern. The cross-sectional dimensions of the shaft are 3'-1 ½" x 2'-2".

BRICK BOND; CHIMNEY CAP

The chimney, like the other two, is laid up in Flemish bond as far as the widths of the sides permit this pattern to be carried out. The cap is the corbelled type similar to the other chimney caps of the building. For a discussion of these and the precedent for them, see pp. 87 and 89. For a general discussion of the Tavern brickwork, see pp. 35, 36 and 38.

WELL HOUSE; ITS POSITION WEST OF WING CENTER LINE DETERMINED BY LOCATION OF OLD WELL

The location of the old well over which the Well House was built and its role in the determination of the limits of the southward extension of the wing have been discussed on pp. 125 and 126. The circular form of the old well may be seen on the archaeological plan, p. 39, The placement of the well house west of the center line of the wing is traceable to the circumstance that the old well itself was considerably west of the hypothetic center line of the wing, assuming the line to have run approximately midway between the west end of the main building and the transverse wall of the old foundations (p.127). The west side of the Well House is set in about 8 ½" from the west face of the cornerboard of the south-west corner of the wing, so that no part of the appendage, with the possible exception of the outermost edge of the lowest row of shingles, projects beyond the west face of the wing. On the east side this roof edge approaches very close to the center Line of the wing.

DIMENSIONS OF WELL HOUSE; ITS ROOF

The Well House is a wood, hip-roofed structure 8'-6" wide (east-west dimension), 8'-5" long (north-south dimension) and about 12'-6" 136 high, measured from the ground on the south side to the roof ridge which rises to the level of the bottom of the wing cornice. Of the total building height of 12'-6" the brick foundation wall accounts for about 6 ½" (2 ½ brick courses show above grade on the south side), a rusticated wood wall for 7'-6 ¾" and the roof, which slopes at an angle of 36 ½°, for the remainder. This roof, which is covered with round-butted as asbestos cement shingles similar to those used elsewhere on the Tavern, has, like the roofs of both the wing and the main part of the building, a "kick" (described on p. 50) but, because of the much lower slope of this roof, the upward curvature of the shingles (the kick) has been made less pronounced than in the case of the other, steeper roofs.

CORNICE; DESCRIPTION OF THIS

The cornice, which unites the underside of the roof edge with the faces of the Well House, is 9 ½" high and 8 ½" wide, the width being the amount of its projection from the face of the wall studding. The upper two-thirds of this is composed of three very commonly used elements, a crown mold consisting of a cyma recta, a fillet and a cyma reversa; a fascia 3 ¼" high, terminating in a drip, and a soffit board, the exposed width of which in about 2 3/8". For the bed mold of the cornice the designers chose an unusual molding. This molding, the longest dimension of which in 4 ¼", is basically a cove but it ends at the top in a ½" bead and at the bottom in a similar bead and a fillet. This cove connects CORNICE BED MOLD OF TAVERN WELL HOUSE 137 the soffit board with the exterior wall boarding.

OLD CORNICE SIMILAR IN TYPE TO THAT OF WELL HOUSE; PRECEDENT FOR BED MOLD

Several old examples of modillionless cornices of the same general character as this one have been found in Williamsburg. The cornices of the Travis House and Captain Orr's Dwelling are instances, as was the cornice of the now demolished small structure known as Casey's Gift. The Travis and Orr cornices have, and Casey' Gift had a much more usual type of bed molding, however, composed of a quarter round, fillet and cyma reversa. An example of a cove, beaded top and bottom, may be seen on the cornice over the main entrance doorway of the Mary Washington House in Fredericksburg. The west dormers of the Brush-Everard House have as bed molds forming the bases of their pediments and continuing along the sides beneath the eaves a cove molding beaded at the bottom but with a fillet instead of a bead at the top.

RUSTICATED SIDING AN UNUSUAL FEATURE; DESCRIPTION OF THIS

The feature of most unusual interest in this small building or building part is the rusticated siding which constitutes its exterior wall covering. This consists of boards about 6 ½" high and 1" thick whose upper and lower edges and ends have been bevelled so that when two boards are brought together V-shaped grooves are formed between them on both the long sides and the ends. The grooves have a depth of about ½" and they are 1 ½" across at the top. Since the design idea here is to simulate the effect of dressed, rusticated stonework, similar vertical grooves are cut at regular intervals in the boards so that full panels or blocks and half blocks result. The blocks are staggered so that the vertical joints of one row are opposite the centers of the blocks of the rows above and below. The half blocks around the openings and at the corners result from this staggering of the joints in alternate rows. The staggering lends verisimilitude to the appearance of this imitation stonework. The length of the 138 full blocks is 1'-3 1/8", except at the corners where it is 1'3 3/16". The full and half blocks are less than their normal dimensions where, on the east and west sides, they strike the south end of the wing and they are likewise somewhat shorter where they run up against the sides of the frames of the "window" and door openings, because at these points the bevel has been omitted. The height of the blocks is almost but not quite 6 ½" and there are 14 courses in the height of the walls.

BOARDS MEET IN BEVELLED JOINTS; NATURE OF THESE

The edges of the boards are joined by what are known as bevelled joints (see illustration). This signifies that one edge of each board is cut to the shape of a half diamond while the other has a bevel which runs all the way across it. In the bevelled joint this continuous sloped edge overlaps the inner face of the diamond about ½" so that when the boards contract in dry whether the joint remains closed. The rusticated boards are held in place by blind nailing with modern wire cut nails, driven through the part of the full bevel which is hidden by the overlap of the half diamond edge.

BEVELLED JOINTS OF RUSTICATED BOARDING OF KING'S ARMS TAVERN

MOUNT VERNON COVERED WITH RUSTICATED WOOD SHEATHING

The most famous example in Virginia of the use of boards with bevelled joints to simulate the effect of stonework is Mount Vernon, George Washington's country seat in Fairfax County near Alexandria. Here all of the exterior walls, as well as the pediment of the west [cut off] have been covered with bevelled boards which, to increase the similar 139 EAST OR FRONT ELEVATION OF THE MERCHANT HOUSE IN DUMFRIES, PRINCE WILLIAM COUNTY AS THIS APPEARED TO A DRAFTSMAN OF THE HISTORIC AMERICAN BUILDINGS SURVEY IN 1935. THIS DRAWING IS PRESENTED HERE SINCE IT SHOWS THE USE ON THIS BUILDING OF RUSTICATED WOOD SIDING SUCH AS WAS REPRODUCED ON THE KING'S ARMS WELL HOUSE. SOME OF THE WOOD QUOINS HAVE BEEN RESTROED IN THE DRAWING ON THE BASIS OF UNQUESTIONABLE EVIDENCE IN THE BUILDING. IT IS LIKELY THAT QUOINS ALSO ONCE EXISTED AT THE NORTHEAST CORNER BUT THE EVIDENCE OF THIS HAD BEEN DESTROYED BY THE REPLACEMENT OF THE RUSTICATION AT THIS POINT. AS IS APPARENT FROM THE PHOTOGRAPH ON p. 129, NO QUOINS WERE USED AT THE CORNERS OF THE WELL HOUSE. 140 to stonework, have been painted and sanded. It is worthy of mention that the courses here are not completely uniform in height, although the discrepancy in their heights is not marked.

ANOTHER OLD EXAMPLE OF RUSTICATION IN WOOD FOUND ON HOUSE IN DUMFRIES

Another example of the use of rusticated stone masonry simulated in wood was found by Milton L. Grigg in 1929 on a house known as the Merchant House in Dumfries, Prince William County. The rusticated work was here confined to the two long sides of the building. HABS drawings made of this house in 1935 show uniform course heights. Judging by photographs made by Grigg the house had never been painted so that sanding was out of the question. It should be stated that the rusticated sheathing of the Well House is painted but not sanded.

LOUVERED DOOR; DESCRIPTION OF THIS

The Well as was stated above, has three openings, a door on the south side and a louvered opening on each of the other two sides. The door is a louvered wood member 3'-0" wide by 6'-1 5/8" high. It is made in the form of a panelled door, with a 1 3/8" thick framework of stiles and rails holding six "panels" between them. The members of the framework are set together with mortise and tenon joints, secured by hard wood pegs driven through them. The "panels" are not the usual "floating" bevel-edged pieces of wood, but consist of openings with wood louvers inserted in them. In the two lower pairs of openings these are set horizontally while in the upper pair they stand upright, to permit a person about to open the door to peer through them. The two bottom louvered openings are rectangular, 2'-4 1/8" high and 1'-0¼" wide and the two upper pairs are squares, 1'-0 ¼" on a side. The louvers are 5/16" thick wood slats turned on an angle of 45°. The outer ends of these project about ¼" beyond the frame and are rounded while the inner ends are cut flush with the framework of 141 the door. They are spaced about 1 1/8" apart, measuring the distance at right angles to their faces. The edges of the openings on the outside face of the door have been given a 5/16" bead, cut from the thickness of the wood. Bronze wire insect screening has been applied over the louvered openings on the inside face of the door. This is held in place by being inserted in small grooves cut in the door framework and is kept taut by means of applied 5/8" beads which form a raised frame around the louvered openings on the inside face of the door.

HARDWARE OF LOUVERED DOOR OF WELL HOUSE

The louvered door is equipped with the following hardware: one pair of reproductions of colonial wrought iron pintel strap hinges, similar to that designated at Type 3 on Plate 24 of the Colonial Williamsburg Hardware Folder; one reproduction of a colonial wrought iron pull similar to the one designated as Type 3 on Plate 14 of the Hardware Folder and one modern, stock screen door closer.

PRECEDENT FOR LOUVERED SCREEN DOOR

louvered doors of the general character of this one existed in the eighteenth century although insect screening, so far as we know, was not used with them as it is here.^* Examples of eighteenth century louvered doors can still be seen in Alexandria, Virginia and New Castle, Delaware.

PANELLED ENTRANCE DOOR OF WELL HOUSE

Since the door just discussed is, in actuality, an open screen door there is, for temperature control and security purposes a second panel door of nearly the same size as the louvered door, its dimensions being 6'-1 1/8" x 3'-0" x 1 3/8". The paneling is raised on the outside only the inside faces of the panels being slightly recessed and without moldings.

There are three old four-panelled doors of this description on the first and second floors of the hallway of the Tayloe House and others of this type could readily be found. This door is equipped with modern hardware.

DOOR FRAME AND SILL; WELL HOUSE INTERIOR IS SERVICE ENTRY

The exterior door frame is 5 ½" wide. It consists of a cyma reversa backband followed by a fascia, a cove, another fascia and a ½" bead. The sill is of the molded type with a profile resembling in a general way those of the first floor window sills of the Tayloe House but, in the interest of durability, it is made of cast iron instead of wood. The profile consists of a half round, a fillet, a cove and a fascia. The sill is about ¾" high and it lines up at either end with the outside edges of the door frame. A metal sill was used here because the Well House interior functions as a service entrance so that the sill gets very hard wear. A good part of this room is occupied by a sidewalk type service elevator which communicates with a receiving room in the basement. The Well House interior, therefore, like so many other interior spaces of the Tavern, no longer serves its original purpose but rather a modern, utilitarian one and, consequently, no attempt has been made to give it authenticity.

LOUVERED OPENINGS; THEIR LOCATION AND SIZE

The two louvered openings are located on either side of the Well House about midway between the south end of the structure and the south end of the structure and the south end of the wing and their heads are one block course (about 6 ½") below the bed mold of the cornice. Their size, including the frame and sill, is 2'-9" by 3'-9 3/8".

DESCRIPTION OF THESE

Each opening contains, within a single frame, both louvers and, behind these, a wood casement window. The louvered areas are, unlike the louvers of the door, undivided, and they have a depth of about 143 3 ½". The louvers (wood slats) are 4 ½" wide and a trifle more than 3/8" thick and they slope outward at an angle of about 38°. They are cut off, at the front, flush with the faces of the jamb and head pieces. These slats like the ones of the door are fixed.

PRECEDENT FOR LOUVERED OPENINGS

louvered openings of the kind described above were used frequently on outbuildings in eighteenth century Virginia. The dairy of the Archibald Blair property, for example, has a band of wood slats for ventilation purposes running almost without interruption around the upper part of the wood walls directly beneath the huge plastered cove. These are old but have been repaired. The old privy of the Coke-Garrett building group also has louvered openings similar to those used on the Well House. In both of these cases, the louvers, as was customary in colonial times, are open. The casement windows used with the louvers of the Well House are unauthentic and were added for reasons of present-day utility, viz., to close the Well House interior of weather and changes in the temperature.

CASEMENT WINDOWS USED TO SEAL OFF INTERIOR

The four-light casements are set about 1/8" back of the louvers. The east window swings in but the west one is fixed because of the proximity of the sidewalk elevator which would be liable to strike an inswinging casement. The casements are divided by muntins into four glazed areas. Since the sash are 2'-2" wide and 3'-1 7/8" high, the panes which result are considerably larger than the typical glass, size (8" x 10 ¾") of the first floor windows of the Tavern. The rails, stiles and muntins have been decorated with the quarter round moulding characteristic of the other windows of the building (see p. 71).

EXTERIOR TRIM LIKE THAT OF WEST WINDOWS OF BRUSH-EVERARD HOUSE

The exterior trim of the two openings is about 4" high and is composed of the cyma reversa backband, fascia and beaded termination 144 characteristic of so much eighteenth century window and door trim in Williamsburg. It is, however, unlike the exterior trim of the other windows of the house which, because it receives shutters, is of the block type. The trim of the Well House resembles, in the sequence of moldings composing its profile, the old trim of the west first floor windows of the Brush-Everard House.

SILLS OF LOUVERED OPENINGS

The sills are of the molded variety. They are about 4 ¼" high at the face of the external sheathing and at their ends they line up with the outside edges of the frames. The profile consists, from top to bottom, of a half round, a ½" band and a cove, ending in a fascia. These sills are of the same general type as those of the first floor windows of the Tavern but, since the openings are smaller, they are much lighter in weight and lack the bead in which the others terminate (see p. 72 for precedent for this sill type).

INSECT SCREENING

Between the louvers and the casement window which operates is a bronze mesh insect screen which comes into use when the window is open. This is a purely utilitarian feature without eighteenth century precedent.

FOUNDATION BRICKWORK; BRICK DRIP

The ground in the immediate vicinity of the Well House slopes toward the south so that while about 5" of its foundation brickwork is exposed at the points where the latter strikes the south foundation wall of the wing, 7" of brickwork shows beneath the south face of the Well House. This brickwork is laid up in English bond and has the characteristics of the brickwork used elsewhere in the foundations of the Tavern (see pp. 36, 38, 41, 42). Brick paving completely surrounds the building so that ground gutters or drips are unnecessary.

STONE STEP AT ENTRANCE

A stone step, 3'-11" long (the width of the door frame), 1'-0" wide and 4 ½" high, has been placed on the ground directly beneath the Well House door opening on the south side. This is an old piece of salvaged stone selected from the Warehouse pool of old materials. Single stone steps such as this, and flights of several stone steps, for that matter, were not unusual in eighteenth century Virginia (see p. 54 for a Williamsburg example). There was no stone in the immediate Tidewater area so that stone for steps and paving had to be brought in from the outside and it came chiefly from England. The stones most frequently found in the excavations in Williamsburg have been the English Purbeck and Portland stones.

REASONS FOR DIFFERENCES BETWEEN EAST AND WEST ELEVATIONS

The west elevation of the wing has the same basic pattern as the east elevation (p. 130). The differences between the two are the result of two circumstances, viz., the fact that the south porch leanto, which strikes the east elevation, eliminates the north window which, probably, would otherwise have existed on that elevation and the presence at the north end of the west elevation of the upright basement bulkhead (pp. 119-123) which brings about the shortening of the north window of that facade by one row of glass panes.

SURVEY OF FEATURES OF WEST FACADE

The basic form which, except for the modification caused by the bulkhead, is exemplified in the west front of the wing, comprehends the following: a wood wall about 25'-0" long, topped by a two-sloped gambrel or mansard roof covered with asbestos cement shingles (pp. 32, 33). The lower slope of the roof is, penetrated at two points by dormers (p. 131). Below these and on their vertical axes are two first-floor windows. One of these is a 21-light window near the southwest corner which has its counterpart in the same relative position in the east wall of the wing (p. 132). The other window is near the north end of the facade and this, since the peak of the bulkhead roof rises to a height of almost 6'-6" above grade, is necessarily shorter than the south window by one three-light row of panes, which makes it an 18-light window. Both of these windows in their detailing and in the possession of slatted or louvered shutters, are similar to the two windows of the east and south facades, respectively. No further comment on these windows seems necessary except this, that if precedent for the 18-light window is needed, it can easily be found in still existent eighteenth century buildings. To mention some examples -- there are windows three lights wide and six high on the first floor of the Brush-Everard House; on the first floor of the Wren Building and on the first and second floor 147 of both,Westover and Carter's Grove. All of the other features of the west facade of the wing -- the main and upper cornices, the weatherboarding and the corner board, as well as the foundation brickwork -- are similar to the corresponding features of the east facade (see pp. 131-133) and, for that matter, also of the south front.

EXTERIOR WOODWORK

WOODS USED ON EXTERIORS ARE NOT THOSE EMPLOYED IN EIGHTEENTH CENTURY

The exterior woodwork, i.e., wall covering, trim, doors windows, etc., of the King's Arms Tavern and Alexander Purdie House and of the outbuildings appertaining to these might have been of yellow pine or poplar or a combination of the two of these. Since, however, the lumber obtainable today from these two varieties of tree is inferior in quality to that available in the eighteenth century other woods which hold up better than these in exterior use were substituted for them. The shingles, in all likelihood, would have been of cypress, but as we have seen (pp. 33, 34), asbestos cement shingles were used in their stead.

LISTING OF WOODS USED & THE EXTERIOR FEATURES MADE OF THEM
SAME WOODS USED FOR INSIDE WORK AS IN OLD DAYS

The following list gives the woods used on this project and the exterior building parts made from each:

Red Gulf Cypress

Weatherboarding; cornice; rake and corner boards; frames, bars and sash of basement windows; wood latticework under porches; members of the south porches and of the north porch of Purdie's House and sills of the Well House.

Heart Long Leaf Yellow Pine

Door sills and wood porch floors, skirting, steps and step carriages.

Northern White Pine

Doors and door frames, and window frames, sash, trim and shutters.

153

Comb Grain Heart Pine

Pulley stiles and parting strips of windows

Douglas Fir

The framing lumber of the building.

INTERIOR WOODWORK

In the case of the interior woodwork in the public spaces of the Tavern, the House and Barber Shop and in the two Kitchens, the same woods were used for the various features as would have been used in the eighteenth century since, within doors where the problem of exposure to the weather does not exist, our present day woods are satisfactory. Chief among these woods are yellow pine, walnut, poplar and gumwood. We will not discuss here the uses to which each of these woods was put since the woods will be mentioned in describing the various parts of the interior.