University of Virginia Library


First it will be useful to consider in some detail what we know about the origin
and nature of mixed paper stocks. It is worthwhile in this connection to quote
at length Philip Gaskell's description of what is presented as a printer's typical
procedure for acquiring and using paper:

As a rule he took paper from particular lots for particular books, not from a general stock,
for it was never normal for a printer to sink capital in a stock of paper sufficient for miscel-
laneous book-printing. A batch of paper was specially ordered from a paper supplier for a
particular book, the order generally being placed and the cost met by whoever was finan-
cially responsible for printing it, be it a bookseller or the author, or the printer himself;
alternatively the printer could buy the paper and charge it out with the sheets. Enough
paper would normally be ordered at once for the whole edition (which was one reason why
casting-off was an early necessity), and as far as possible it would be supplied as a single
lot, of even size and quality and with a single watermark-design throughout.[15]

Gaskell goes on to note that although many early books are printed on more
than one paper, the various papers usually appear in extended runs, and "batches
of different papers would not as a rule be mixed," but paper left over from a
job, and/or less-than-perfect "naughty" sheets culled from quires of paper, were
sometimes used to eke out a main supply of paper.[16] Such remnants, as opposed to
runs, might be used in the middle of books where slightly defective paper would
be less noticeable, or to print cancels or other resettings (as when fewer than the
required number of sheets have been run off, or some accident spoiled part of
the run of an edition sheet.)[17] Finally, in a cryptic footnote Gaskell remarks that


Page 155
"Not all printers (etc.) bothered about uniformity of paper in a book,"[18] with no
explanation of what this might mean in terms of his generalized account of the
acquisition and use of printing paper. Is the implication that a book printed on
mixed paper indicates a lack of care on the printer's part? My own investigations
indicate that Gaskell's account requires considerable modification.

When I began to study paper bibliographically I was vaguely aware from
Greg's account of the Pavier quartos that mixed stocks of paper were sometimes
used in producing hand-press books. Stevenson in fact begins "Watermarks are
Twins" by stating that "Most early books are printed on a variety of papers. They
contain a number of different watermarks," but goes on to assert that "that simi-
lar wild mixtures" to those found in the Pavier quartos "are rare."[19] I was thus
unprepared for the wild mixture of papers I found when I began to examine cop-
ies of the three editions of Piers Plowman printed by Robert Crowley in 1550 (STC
19906, 19907, and 19907a).[20] The first edition was printed almost entirely on
remnants: in the twenty-two paper copies examined (three others are printed on
vellum) I found that twenty-four different paper stocks had been used in produc-
ing its thirty sheets (twenty-nine full sheets and two half-sheets printed together).
Individual gatherings were printed on as many as five different stocks, while only
six gatherings appeared on a single paper. There is considerably more regular-
ity in the second edition, where only eleven stocks were used—and in the third,
where there are seventeen. In the latter two editions there are several substantial
runs of paper, a feature generally absent in the first edition, and gatherings are
much more likely to appear on a single paper stock. Seven paper stocks appear
in two of the three editions, and three stocks can be found in all three. In total,
at least thirty-nine different paper stocks were used to print the three editions—
and thus at least seventy-eight individual watermarks. And nine of the water-
mark pairs had their designs repeated in a second pair of moulds (a feature of
the Pavier quarto papers, as will be seen below). But in the three Piers editions
there is no discemable difference between the care taken by the printer with the
first edition, which lacks significant runs of paper, and with the second and third.

The mixing of paper stocks probably occurred at every stage of production,
distribution, and use in the printing house. In the sixteenth and seventeenth
centuries, most of the ordinary printing paper used in England came from Nor-
mandy and Brittany. As Stevenson notes, many of these paper mills were small,
one-vat, family operations. A paper merchant or factor might make regular trips
up and down the small streams where the mills were situated, picking up what-
ever was available and, sorting by size and quality, combining it into larger units
for packing and shipping to England. Stevenson speculated on other additional
ways in which varieties might become mixed: "no doubt the piling into cargo
boats, the unloading at London, the stacking in the warehouse of the London
paper merchant and later on the shelves of the printer he sold the paper to—all


Page 156
this handling contributed to the mixture of papers we find in books. And much
might happen in the laying out of tokens for the presses."[21] At the end of a par-
ticular printing job the printer might return leftover paper to the warehouse, to
be mixed with other small batches and used in future projects. Or the printer
might have sold off such odd lots to a paper merchant who would have combined
them with other remnants similarly acquired and then resold them, perhaps to
be used for job printing or ephemera. At this stage we don't know what vagaries
of these complex processes result in some books being printed on a single paper
stock, others on slightly mixed stocks, still others on wildly mixed stocks. (All
three occur in the Pavier quartos.) How did printer or publishers decide to print
on homogenous paper or mixed paper? And why, within mixed paper volumes,
do some gatherings appear on a single stock? Only a large-scale study of all ex-
tant books printed over a lengthy period could even hope to answer these ques-
tions. But, in contrast to Gaskell's suppositions, moderately mixed paper stocks
may ultimately be found to be something of a norm, and wildly mixed stocks not
at all infrequent.[22]


Philip Gaskell, A New Introduction to Bibliography, (1972; repr. with corr. New York:
Oxford Univ. Press, 1975), p. 142.


Gaskell, New Introduction, p. 142.


Gaskell, New Introduction, pp. 142–143. For an extended consideration of the potential
bibliographical significance of runs and remnants in printed books see chapter 6, "Runs and
Remnants," in Stevenson's The Problem of the Missale Speciale (London: Bibliographical Society,
1967), pp. 71–99.


Gaskell, New Introduction, p. 142, n. 4.


Stevenson, "Twins," pp. 57, 58.


See R. Carter Hailey, "Giving Light to the Reader: Robert Crowley's Editions of Piers
(1550)," diss., Univ. of Virginia, 2001, pp. 173–322.


Stevenson, "Twins," p. 60.


For example, having examined five copies of the [1550?] edition of Chaucer's Workes
(STC 5071–74; actually a single edition with a different colophon for each member of the four-
man consortium of publishers), I distinguished sixteen different CROWN stocks mixed with two
BEAR stocks. For a book of 181 edition sheets, I consider this a moderate mix. By contrast,
in examining a single copy of John Marbecke's 1550 Concordance (STC 17300) I found fifty-six
watermark pairs represented in its 202 edition sheets.