University of Virginia Library

Shakespearian Dated Watermarks
Allan H. Stevenson

ON a perfect morning in July, when I might have been bowling on the greens at Arroyo Seco, I made a minor Shakespearian discovery. In the Huntington Library reading room I was reëxamining the Jaggard-Pavier quartos of 1619. As all the world knows, Sir Walter Greg brilliantly proved, in 1908, that these quartos, variously dated 1600, 1608, and 1619, had all been printed in the latter year; and the most persuasive part of his evidence lay in the complex mixture of watermarks distributed through them.[1] Yet history records that conservative-minded gentlemen like Sir Sidney Lee and Alfred H. Huth were not readily persuaded.[2]

On this day forty-two years after,[3] I was searching the Huntington set of 1619 quartos for evidence of twin watermarks, and finding what I sought: most of the marks pictured by Greg were turning out to be pairs themselves or members of pairs.[4] I had also found several new marks to add to his list of twenty-seven. Thus, while leafing through the Church copy of Sir John Oldcastle (dated "1600"), I was not much surprised to note a new pot in sheet F. It was not a prepossessing fellow, for it looked more like a pot of 1600 than of 1619, having a simple form, apparently little superstructure, and a malformed base. But in a narrow band across the throat was an inscription. I could make out something like a 6, a sort of lopsided u, and what might be a tall, thinnish X. Puzzled, I moved on, hoping for another view of this new thing. But I had none.

Half an hour later—such is the timing of coincidence—I stopped similarly in my sidelong perusal of the Church copy of Henry V (dated "1608")—again at sheet F. Here was another pot not in Greg's gallery. This

one looked more like a Norman pot of about 1620. It had a candle-like top above a number of lobes, a fleur-de-lis on its chest, the initials PD on its bowl, and below them certainly—a date. The figures 1 and 6 were clear enough. Then came a space where only a 1 could go, and of which only a stub remained. And beyond that a figure that might be a 7 or 9, the top being slightly obscured by type. The base was also obscured, but that hardly mattered. The inscription on the bowl apparently had once read PD/1617 or 1619. This pot appeared in no other sheet of Henry V. Only at F1.4, with the date on F4.

Breathless, I reopened the Oldcastle, at sheet F. The inscription on the pot throat might also be a date. Indeed, it looked much like one. 1608. The 1 was a heavy mark just inside the left line of the neck. The 6 was a 6, with a curved top. The u-like form was the lower part of an o. The X was an 8 with the ends squared by the lines of the band. As the paper was thickish and the imprint of the wire-design none too sharp, I regarded it doubtfully; yet there seemed to be no other way to read what was on the band. It was no attempt at ornament; and from the reverse side it made no sense. The top of this pot (with the date) was in F4, and the base and part of the bowl in F1. And on the bowl parts of two letters were visible, deep in the quarto crack. They might be MO or MJ—or BM or LM reversed.[5] But they merely added to the mystery of this curious pot.

Briquet when summoned brought forward several analogies of dated pots, but none with a dated band high on the throat.[6] Churchill, on the other hand, offered in his first pot one remarkably similar to the PD pot.[7] Here were the same candle-like shoot above clustered lobes, a similar fleur-de-lis throat, a similar PD on the bowl, and below it 16, apparently a year date with both final digits missing. Whether this is the same pot with the 7 or 9 lost, or its twin, or a mark from a related set of moulds I do not know: such are the vagaries of tracings; and Churchill does not show the position of the chains. His note merely indicates that he found it in a sheet (of manuscript?) which he dates 1618.[8] This fact may be one reason for preferring the reading 1617 on the Henry V pot, though at first I preferred 1619. The sketches I made show a figure with an open top and a curved back. Let us call it a 7.

If it seems strange to the reader that one or two figures should drop out of a watermark, he should note that they must have been formed of tiny

pieces of wire and sewed at either end to the band wires across the belly of the pot. The constant wear on moulds in use and the brushing at the end of the day tended to distort the wireforms and even to break off bits of wire. In this instance it looks very much as if the 1 had come unsewn at the top and subsequently broken off near the base; and the final digit may have been distorted into ambiguity. Just two weeks later, in examining the plates in Raleigh's History of the World (Jaggard, 1621), crown folio, I noted a pair of shields, quarterly a fess RGD and three lions (style of Heawood 576), with the year 1610 in the base point of one shield and 16:0 in its twin. In this instance too the second figure 1 has dropped out, leaving sewing-marks to show where it had been.[9] No longer could I doubt that the gods provide coincidences by way of empiric proof.

Both dated Shakespeare pots are situated mainly between chains. The 1608 pot fits within a chain-space of 23-24 mm. Its style is that of Heawood 3548-54. Its crest has five points with a trefoil on the center one and single or double circlets on the others. The neckband is 6 mm. high and the date 9 mm. wide. There is not enough of the bowl to measure; but the misshapen

base measures 6[18]0 reading up. The handle is in normal position, at the left.

The 1617 pot is centered at top 9.5:9.5 between 19 mm. chains. The leafed shoot above (if that's what it is) is unlike that in Greg 17 or Heawood 3545, in that it has sprouts above and leaves farther down. The five lobes that show resemble those in Greg 6 or Heawood 3565. The handle is to the right, outside the chain. The bowl with the precious date measures 1[18:1] up. The base is hard to make out. The accompanying illustrations of these two pots are mere rough sketches.[10]

Friends at the Huntington who viewed these pots generally conceded the 1617 date: there could be no doubt of it, except for the ambiguity of the 7 or 9. When shown the 1608 pot Dr. James McManaway shook his head slightly in puzzlement; while others said it looked more like a date than anything else. Obviously a further example of this tantalizing pot had to be found, if possible. And the PD pot was not yet complete, for its neck was lost in the quarto fold.

None of the set of nine Pavier quartos at the Huntington yielded further instances of these pots; and the Library had disposed of its so-called duplicate of Henry V. So I turned to the Clark Memorial Library in Los Angeles. Its copies of Oldcastle and Henry V had only the pots and shields illustrated by Greg. A month later in Austin I examined the Oldcastle of 1619 in the Wren Collection: it had no dated pots. In September in New York I saw the Lenox, Morgan, and Rosenbach copies of Oldcastle and the Morgan copy of Henry V; and later in Chicago the Newberry Henry V; but still I found no pots with dates.[11] Recently Dr. Giles Dawson has kindly checked the Folger Shakespeare Library copies—an undertaking, for the Folger has seven copies of each! Yet his patience has produced no pots of 1608 or 1617, no jugs that resemble my sketches.

Thus the conviction grows that these are "intrusive" watermarks. The 1608 mark appears just once in seventeen copies examined by Greg, Dawson, and myself; and the PD/1617 mark once in fifteen. According to the Pollard-Greg theory the 1619 quartos were printed on a job-lot of miscellaneous papers supplied by a London paper merchant.[12] I have suggested that such mixtures may have resulted normally from the manufacture of paper by a cluster of papermills in Normandy (or elsewhere) and its periodic collection by a paper factor.[13] Yet neither hypothesis, nor both taken together (which may be the better view), quite explains these rare dated

pots. They imply the inclusion of an odd quire or so in the paper laid out for an edition-sheet. Either the paper merchant has thrown in the remnant of a ream, or Jaggard has swept out his stockroom. In laying out paper for the press, the outside or "cording" quires were commonly put aside and replaced by good inside quires from other reams—or by quires culled from cording quires.[14] Further, extra quires were needed for filling out short quires, for proofs, and for replacement of "naughty" sheets. And sometimes for copies allowed in part-payment of the printer's work. Occasionally, no doubt, the warehouse keeper recovered a quire or two from some odd corner or pile of waste—and thus filled out the count for the press.

If the 1608 date can be accepted, it looks as if we have an instance of such an odd quire turning up in Jaggard's stockroom (or his paper merchant's) ten years after it was sold out of France. This interval is in accord with delayed uses noted by Briquet.[15] If the other date is 1617, there has been a proper interval after manufacture; and it may be that Jaggard had used some reams of this paper in recent books. If the date is 1619, the year of issue, the small amount used is less likely to have come from a broken ream; but it might indicate a mixture of paper from more than one vat in the papermill—though then we should find other PD marks among the Jaggard-Pavier quartos. A 1619 date would have interesting implications for the time of printing.[16] A slow search among pot folios and quartos of 1608-10 and 1617-20, particularly among Jaggard's books, may ultimately answer the questions posed by these dated Shakesperian watermarks.

The papermaker PD may be represented by some of the Pot PD watermarks in Briquet's great work and Edward Heawood's recently published collection of Watermarks,[17] but his identity is unknown. He might belong to the Debon or the Durand family of papermakers, or some other.

The significance of this small Shakespearian discovery is not what it

would have been forty or so years ago. At that time the ways of the New Bibliography proved difficult for men of the old school; and the arguments of A. W. Pollard and W. W. Greg from printers' and papermakers' ornaments, though they look good now, provoked an amiable controversy of two years' duration—which was smothered in 1910 by William J. Neidig's evidence from photographic measurement of title pages.[18] If the New Bibliography had discovered in the Capell or Huth copies (say) what I have now happened on in the Church copies of Oldcastle and Henry V, it would have been saved some of the joys of creative bibliography, though it might have convinced with one thwack schoolmasters in the upper forms. For anyone at all can see that a Sir John Oldcastle dated 1600 on its title page but apparently containing a 1608 watermark may not have been printed in 1600; and that a Henry V dated 1608 but certainly watermarked 1617 (or 1619) was definitely not published before the latter year. Indeed, so elementary a demonstration may be useful even today, for some boys have not yet been promoted to the upper forms.

On the other hand, inasmuch as a paper war was waged and in effect won forty years ago, I, on a marvelous July day in 1950, might well have gone bowling on the smooth Arroyo Seco greens. The shade of Shakespeare might have been better pleased, for bowls was his favorite game.[19]

Notes