Oxford’s History of Science Museum displays a marble Holy Table, inventory 15449, made in the mid seventeenth century from a 1659 engraving. Dee’s brightly painted wooden original is lost. Sloane MS 3188 nevertheless preserves more than a label for that missing object: the 28 April 1583 action records its border, reading direction, character orientation, and central field in stages. Those decisions reveal the table’s function and let the later witnesses be tested rather than simply trusted.
- Evidence type
- Object and reconstruction
- Date
- Mid-seventeenth-century marble copy
- Anchor record
- History of Science Museum · inv. 15449
The action turns a square into a reading surface
Whitby’s transcript of the 28 April 1583 action moves from geometry to sequence. A square receives a border and a smaller square at its center. The side opposite Dee and Kelley is called “in fronte Tabulae.” The reported instruction begins on the right of that side and moves toward the left, then continues around the remaining sides. After supper Dee asks which way the heads of the characters should face, and the recorded answer turns them inward toward the center. These are functional decisions. The Angelical characters form a route whose starting side, direction, and inward-facing orientation depend on the observers’ position. The manuscript also distinguishes a planning sketch from the object: a six-inch drawing has a half-inch border, while the intended table border is one inch. The sketch can encode sequence without being a scale plan of the vanished furniture.
The table organizes an action without acting as a speaker
Within the recorded arrangement, the table establishes a bounded and directional field for the viewing apparatus. The large wax Sigillum is assigned to the center beneath the show-stone, and smaller seals support the table’s feet. Dee prays and records; Kelley reports what he sees or hears; figures within the account are given speech. The table’s role is architectural and textual: it positions the stone, repeats marked supports above and below, and places characters around a central field. It does not itself generate a transcript or remove the need to distinguish recorder, scryer, reported figure, and later annotation. This division of roles explains why a furniture reconstruction cannot recreate an historical action by appearance alone. A painted border may follow the diagram while changing direction, proportions, material, or the relationship between the reader and the in-fronte side.
Worked reading: one table instruction becomes the 1659 engraving
Follow one instruction through the witnesses: face the in-fronte side, start at its right, proceed left, and turn the heads of the characters inward. In the manuscript transcript those clauses arrive at different moments. The side and direction are named during construction; character orientation is settled only when Dee asks a later question. The action therefore preserves a decision process with pauses and an unresolved detail. Casaubon’s 1659 engraving compresses that sequence into one regular square border with four groups of twenty-one positions. A viewer sees the completed circuit immediately. The printed heads face toward the interior, so the plate can be checked against the recorded answer instead of accepted only because it is old. Yet the engraving was issued seventy-six years after the 1583 action. Its even spacing, uniform line, and monochrome surface belong to print and cannot prove the lost table’s paint texture or exact proportions. Oxford’s marble copy moves the design through a further change: a durable stone object made from that printed witness. It preserves the print’s reception, not an independent view of Dee’s wood.
Manuscript, print, marble, and wax answer different questions
Sloane MS 3188 is the closest witness to the staged instruction and records reading logic. The 1659 engraving shows how a later editor and engraver stabilized the result for print. Oxford inventory 15449 shows that a seventeenth-century maker converted that print into marble; the museum says it may have belonged to William Lilly, so that ownership remains possible rather than certain. The British Museum wax discs are separate surviving objects. They help reconstruct the proposed vertical arrangement but are not fragments of the painted tabletop, and their curator likewise leaves Dee-lifetime manufacture uncertain. The lost original, printed plate, marble copy, and wax group can therefore be related without being collapsed. A modern image that stacks them into an “original altar” erases precisely the evidence that matters: surface changed from painted wood to ink to marble, proportions were regularized, two of four small supports survive, and the manuscript’s construction dialogue became a finished diagram.
Continue through the manuscript design and 1659 printed witness →