Arlington House: Historic Structures Report (Phase II) (Appendix 6)

ARLINGTON HOUSE

Historic Structures Report
Phase II

APPENDIX VI

NORTH WING PHYSICAL INVESTIGATION

NORTH WING PHYSICAL INVESTIGATION
THE ARLINGTON HOUSE
GEORGE WASHINGTON MEMORIAL PARKWAY

by
S. Elizabeth Sasser

Technical Advisor
James S. Askins

Denver Service Center
Williamsport Preservation Training Center

July 1985

I. INTRODUCTION

This report details the findings of a program of limited destructive investigation in the north wing of the Arlington House, conducted in May 1985, by the Williamsport Preservation Training Center. The purpose of the investigation was to attempt to resolve questions relating to the origins and evolution of the north wing, raised by architectural features uncovered during structural stabilization work (1980-1981), and research for the Historic Structure Report (Architectural Data Section, Phase II, Arnest and Sligh, 1983). Extensive fabric removal in the course of the structural and investigative work revealed unanticipated evidence of a complex and largely undocumented history of construction changes.

It was determined on the basis of this evidence that the north wing existed in some form as a freestanding structure prior to completion of the central block and loggia; that the originally constructed first floor elevation was approximately five feet higher than the existing; and that there were major changes in openings, partition arrangement, and fireplace locations. However, integration of this evidence into an evolutionary sequence and chronology of construction is complicated by the obliteration of much early fabric in the implementation of the design program for the mansion as a whole, and in subsequent phases of alteration, repair and attempted restoration. Primary architectural evidence usually found at floor and ceiling level was destroyed during the Department of the Army program of restoration (c. 1906-1929). Also, the continuity of architectural finishes was severely disrupted by general repairs and the introduction of plaster patches over expanded metal lath at all partition/wall intersections. Therefore, the extant evidence relating to early periods of construction is fragmentary and isolated in context. The current lack of sufficient historical documentation to support positive identification of period features, particularly those relating to the interpretive period (c. 1861), further complicates the interpretation of physical evidence.

In December 1984, at the request of the Park, the Williamsport Preservation Training Center conducted an evaluation of the physical evidence uncovered in the previous phase of work. After four weeks of onsite investigation by Jim Askins, Chief, WPTC, assisted by Historical Architect, Elizabeth Sasser and Preservation Trainee, Barry Caldwell, it was concluded that the evidence as it was then understood was insufficient to determine the original extent and form of the structure, or the evolution and dating of interior partition arrangements. There fore it was recommended that additional, limited destructive investigation be completed, with the following objectives:

— To establish the original building limits and siting of the north wing in its earliest configuration as a freestanding structure. This was felt to be a critical indicator of whether the building was a pre-existing structure of unknown function and appearance, adapted for use, or was constructed by GWP Custis as an element of the design concept for the mansion as a whole, as attributed to architect George Hadfield.

— To determine the nature and location of any fenestration predating the existing, and to establish whether existing fenestration represents original construction.

— To further establish the location and sequence of partitions predating the existing, and aid in developing other sequences in finish construction.

This proposal forms the basis for the architectural investigation detailed in this report. It should be noted that this is not a comprehensive treatment of conditions in the north wing. The present work was completed in conjunction with the revision of the Historic Structure Report, Architectural Data Section, for which it is to serve as an addendum, dealing only with the issues outlined above. A number of questions relating to the dating of the loggia, bathroom, and the nature of interpretive period cooking arrangements were beyond the scope of the present investigation. Finally, the author gratefully acknowledges the assistance and many kindnesses shown by the entire staff of the Arlington House, particularly Staff Curator, Agnes Mulliins, during the course of the investigation.

METHODOLOGY

The investigation procedure consisted of removal of an 8 inch wide band of plaster down to brick, extending completely around the interior of the perimeter wall of the north wing. The 8 inch width of the plaster band represents 3 courses of brick, which was felt to be the minimum necessary to identify construction joints or major changes in masonry rhythm. The height of the plaster band was set at approximately 90 inches above the existing floor, or 30 inches above the previous, higher floor level. This was done to intersect potential infilled opening relating to the earlier floor elevation. As the investigation proceeded, the height of the plaster band was adjusted upward slightly in several locations to avoid disturbing significant concentrations of extant wallpaper, and to maximize the potential for encountering infilled openings.

Plaster removal was completed one coat at a time on each wall. Representative samples of each plaster type were labeled with the area from which they were taken and accessioned by the Park into their permanent artifact collection. Several areas of plaster showing significant concentrations of wallpaper fragments were left in situ. Where incident wallpaper fragments were encountered within the confines of the plaster band, they were photographically recorded before removal. Every attempt was made to remove them with the entire thickness of plaster backing intact. These also were labeled and given to the Park to be accessioned.

PRESENTATION OF PHYSICAL EVIDENCE

The format chosen for the presentation of physical evidence was adopted for two reasons. First, to simplify discussion of the features by placing them in a visual context. The drawings were intended to highlight key features by reducing the mass of superfluous detail. Despite the limitations of recording a three dimensional space in two dimensions, it is hoped that the photographs will assist the reader in making an independent evaluation of the narrative analysis accompanying the drawings. Also, in the event that future treatment of the structure renders the physical evidence inaccessible, this approach provides at least a partial graphic record of existing conditions.

GLOSSARY OF PLASTER TERMS

series—an application of plaster in one or more coats, of which the outermost is intended as a finish coat in a specific period.

coat—a single layer application of plaster over a surface.

rough coat—the first coat in three coat plaster work. The term is used here in preference to the contemporary term "scratch coat" due to the fact that no raked base coats were found in the north wing.¹

brown coat—The second coat in three coat plaster work. It should be noted that this term is not a reference to plaster color.

finish coat—The third and final coat in a three coat plaster series, forming either the final finish condition for the wall or a substrate for paint or wallpaper.

scored coat—A plaster surface which has been hacked with a tool such as a plasterer's axe after the plaster has set, to form a key for adhesion of subsequent layers of plaster.

1. Harley J. McKee. Introduction to Early American Masonry. (Washington, D.C.: National Trust for Historic Preservation, 1973), p. 86.

North Wing first Floor Plan (click on image for a PDF version)

II. PHYSICAL EVIDENCE

NOTE: Letter designations indicating a sequence of plaster series apply only to the page on which they are shown, and do not represent a general sequence throughout the north wing.

1 West wall RM 104 Door D1-104

2 West wall RM 104. Detail of lintel at north jamb, Door D1-104

3 West wall RM 105 north of chimney breast.

4 West wall RM 105 south of partition.

5 West wall RM 105 North face of chimney breast approx. 100" above existing floor level.

6 West wall RM 105 intersection of chair rail at chimney breast

7 West wall RM 105, chair rail intersection at north face of chimney breast (plan view).

8 West wall RM 105, chair rail intersection at north face of chimney breast (elevation view).

9 RM 105, north face of chimney breast at west wall.

10 RM 105, north face of chimney breast at west wall.

10A

11 West wall RM 105, face of chimney breast.

11A

12 West wall RM 105, detail of infilled joist pockets at chimney breast.

12A

13 West wall RM 105, face of chimney breast, detail of plaster return.

13A

14 West wall RM 105, central portion of chimney breast.

14A

15 West wall RM 105, face of chimney breast, lower middle quarter of exposed masonry feature.

15A

16 West wall RM 105, face of chimney breast, lower quarter of exposed masonry feature.

16A

17 West wall RM 105, face of chimney breast, upper middle quarter of exposed masonry feature.

17A

18 West wall RM 105, face of chimney breast, upper quarter of exposed masonry feature.

18A

19 West wall RM 105, plaster detail at west face of chimney breast.

19A

20 West wall RM 105, plaster at west face of chimney breast.

20A

21 West wall RM 105, south face of chimney breast.

21A

22 West wall RM 105, wall adjacent to south face of chimney breast.

22A

23 West wall RM 105, south half of Window W1-105 head.

23A

24 West wall RM 105, north half of Window W1-105 head.

24A

25 West wall RM 105, between window W1-105 and door D1-105.

25A

26 West wall RM 105, detail of plaster patch at partition scar.

26A

27 West wall RM 105, north middle quarter of door D1-105 head.

27A

28 West wall RM 105, north quarter of door D1-105 head.

28A

29 West wall RM 105, south quarter of door D1-105 head.

29A

30 West wall RM 105, south middle quarter of door D1-105 head.

30A

31 West wall RM 105, detail of wallpaper fragments.

31A

32 West wall RM 105, between door D1-105 and cheek wall.

32A

33 RM 105, southwest corner.

33A

34 RM 105, south cheek wall.

34A

35 South wall RM 105, (north wall of central block), door D2-105.

35A

36 South wall RM 105, north wall of central block.

36A

37 South wall RM 107, (north wall of central block), western third of plaster band.

37A

38 South wall RM 107, central third of plaster band.

38A

39 South wall RM 107, eastern third of plaster band.

39A

40 RM 107, masonry detail at southeast corner.

40A

41 RM 107, southeast corner.

41A

42 RM 107, western edge of south cheek wall.

42A

43 RM 107, south cheek wall.

43A

44 East wall Rm 107, north of cheek wall.

44A

45 East wall RM 107, south of window W1-107.

45A

46 East wall RM 107, exposed brick at south jamb of window W1-107.

46A

47 East wall RM 107, between window W1-107 and north partition wall.

47A

48 East wall RM 107, below window W1-107.

48A

49 East wall RM 106, detail at partition/wall intersection.

49A

50 East wall RM 106, between window W1-106 and south partition.

50A

51 East wall RM 106, detail of infilled masonry opening.

51A

52 East wall RM 106, south of window D1-106.

52A

53 East wall RM 106, north jamb at window W1-106.

53A

54 East wall RM 106, south jamb at window W1-106.

54A

55 East wall RM 106, detail of infilled masonry opening.

55A

56 East wall RM 06, between window W1-106 and north partition.

56A

57 East wall RM 106, between window W1-106 and north partition.

57A

58 RM 106, northeast corner.

58A

59 East wall RM 106, detail of plaster edge.

59A

60 East wall RM 106, detail of chair rail ledger.

60A

61 South wall RM 106A, detail of chair rail.

61A

62 North wall RM 106, detail of chair rail.

62A

63 East wall RM 105, detail of cornice and neck moulding shadow.

63A

64 Northeast corner of RM 106A, detail showing continuity of chair rail under partition.

64A

65 East wall RM 104, north jamb of window W2-104.

65A

66 East wall RM 104, south jamb of window W2-104.

66A

67 Northeast corner RM 104, masonry detail.

67A

68 Northeast corner RM 104.

68A

69 Northeast corner RM 104, detail of wood nailer block.

69A

70 Northeast corner RM 104.

70A

71 Comparative detail of formed in places wood nailer block.

71A

72 Comparative detail of wood nailer block driven into place.

72A

73 North wall RM 104.

73A

74 North wall RM 104, east of window W1-104.

74A

75 North wall RM 104, west jamb at window W1-104.

75A

76 North wall RM 104, east jamb at window W1-104.

76A

77 Cut nail shown in situ in photograph 78.

77A

78 North wall RM 104, detail of machine made, cut nail found embedded in mortar joint.

78A

79 Northwest corner RM 104.

79A

80 Northwest corner RM 104, masonry detail.

80A

III. SUMMARY OF FINDINGS

The evidence detailed in the preceding section provides support for the theory that the north wing, while initially constructed as a freestanding unit, was designed as an element in a concept for the mansion as a whole. Bonded brick at three out of four perimeter wall corners and continuity in construction indicate that the shape and orientation of the existing structure is original. The window openings in the east and north walls appear to have been constructed to approximately their existing dimensions, although the first floor, at its original height, spanned the middle of the openings. Apart from the unexplained brick jambs on the east wall of Rm. 106, the absence of infilled masonry openings relating in a more conventional manner to the original floor level suggests that the north wing facade was designed to fit the formal and aesthetic requirements of the completed mansion. Of particular interest is the fact that no infilled openings were found on the west wall. The absence of any openings original to construction of this wall serves as circumstantial evidence of intent to add an additional architectural element to that elevation (i.e., The Loggia). The apparent conflict between exterior appearance and interior division of space is consistent with the assumption that G.W.P. Custis built the north wing (c. 1802) while still a bachelor, both as a residence and as a place to store his collection of Washington artifacts.¹ The extant physical evidence relating to the original interior finishes — chair rail, baseboard, and wallpaper — is diagnostic of residential use. However, based on the physical evidence suggesting that the north wing facade was intended to serve as an element in a larger design concept, it seems probable that this arrangement was regarded as an interim measure.

As finances permitted and the requirements of family life dictated, Custis continued development of the design scheme. The diary of artist William Birch recorded a visit to the Arlington House, probably in September of 1805. He noted that Custis "Had built the two wings of his Capital House."² Custis' daughter, Mary Ann Randolph Custis (b. 1808), related in a letter that the family lived in the two wings of the house for a number of years before the central block was completed.³ Unfortunately, no documentary source has been located which refers to the interior arrangement of either of the wings during this period. Although the location of several pre-floor reduction partitions has been determined from the physical evidence, there is insufficient information on which to base a complete interior plan for the earliest construction period. Also, it is probable that at least one major change in interior arrangement occurred prior to the reduction in floor height.

Although no definitive date can be assigned to the change in floor level on the basis of the available evidence, the first major traces of a post floor reduction room arrangement cannot predate the post 1835 cut nails found driven into the brick of the north and south perimeter walls. The principal extant feature of this phase is a room at the north end of the wing, approximately 15'—6" x 22'—0", characterized by the cornice and neck molding shadows presently visible in Rms. 104, 105, and 106. Again, the physical evidence is insufficient to determine the dating of this arrangement or its layout in the remainder of the wing.

The partition arrangement following this phase represents the existing condition in the north wing. There is presently no definitive means of establishing a date of construction for the existing partitions. Because the partition/perimeter wall intersections have been cut full height and patched over expanded metal lath, the continuity of partition and perimeter wall plaster series cannot be established. While it is possible that the present partition locations represent the 1861 interpretive period condition, there is no compelling evidence in support of this conclusion. The presence of post 1870 machine made lumber in door openings and the continuity of lath and plaster adjacent to these openings provides circumstantial evidence that the partitions themselves may post date the interpretive period.

1. Charles W. Snell. Historic Structures Report, Historic Data Section, Vol. 1 (Unpublished Manuscript). March 1982. pp. 11-12.

2. IBID., p. 19.

3. IBID., p. 23.

4. IBID., p. 24.