 | Notes on Recent Work in
Descriptive Bibliography |  |
PUBLISHERS' BINDINGS, ENDPAPERS, AND JACKETS
When descriptive bibliographies are conceived as records of books as
published, and
thus as contributions to publishing history, the only bind-
ings that are relevant
to the basic descriptions are those commissioned by
publishers as the outward form
in which their books were to be offered
authors, printers, publishers, and subjects (as opposed to catalogues of
specific collections) provide binding descriptions only for books published
in edition-bindings, which became common in the 1820s. Before that
time, however, booksellers (or others fulfilling the publishing function)
did apparently often place portions of some editions on sale in a form of
temporary or unfinished binding; and when such bindings can be identi-
fied (as with the boards-and-label style of the late eighteenth and early
nineteenth centuries), they would properly be included in basic descrip-
tions. (See the comments on the recent work of Nicholas Pickwoad and
Aaron T. Pratt under "Ideal Copy" above.)
Some account of the craft or custom (that is, non-publishers') bind-
ings that are
found on most books from before the 1820s may be given
in descriptive
bibliographies in the paragraph on the copies examined,
where any post-publication
features of individual copies can be recorded
(see "Arrangement" below). Such
accounts, when they exist, are usually
very brief; the place where more detailed
descriptions of custom bindings
occur is in catalogues of collections, where the
focus is on the particular
copies that happen to have been brought together in those
collections.
The historical study of the structure and decorative styles of custom
book-
bindings is a distinct field, with its own terminology and an
extensive
and distinguished literature. The best sources to turn to for basic
guid-
ance are relatively recent ones: David Pearson, English Bookbinding Styles
1450–1850: A Handbook
(2005, 2014); the chapter on "Armorials, Other
Binding
Stamps & External Features" in his Provenance Research in
Book
History (1994), pp. 97–131; Mirjam M.
Foot, "Bookbinding and the His-
tory of Books," in A
Potencie of Life: Books in Society, ed. Nicolas
Barker
(1993), pp. 113–126; Nicholas
Pickwoad, "The Interpretation of Book-
binding Structure: An Examination of
Sixteenth-Century Bindings in the
Ramey Collection in the Pierpont Morgan Library,"
The Library, 6th ser.,
17 (1995), 209–249; and
Philipp. J. M. Marks, The British Library
Guide
to Bookbinding: History and Techniques (1998). A recent
example of a great
descriptive catalogue with detailed accounts of bindings is Giles Barber's
The James
A. De Rothschild Bequest (2013). Other outstanding work of
recent
years can be found in any book by A. R. A. Hobson,
H. M. Nixon, or
Mirjam M.
Foot. (A long list of the literature of custom binding appears
in my Introduction to Bibliography [2002 revision], section
8D, pp. 239–248;
for publishers' bindings, see section 8E, pp. 248–251.)
In the strictest sense, publishers' "bindings" are not bindings but rather
casings,
since the cloth-covered boards and spine constitute a case that
is manufactured
separately and into which the sheets are fastened. But
the term "publisher's
binding" (or "edition-binding") is widely used and
"case." The description of publishers' bindings requires a means for citing
colors and cloth grains; describing endpapers also entails noting colors
and sometimes (in the case of marbled papers) the various patterns that
marblers have traditionally used; and accounting for jackets obviously in-
cludes the identification of colors. I took these matters up in three essays,
on color (1967), patterns (1970), and jackets (1971, supplemented in 2011),
including in them some suggested wordings for the relevant paragraphs
in descriptive bibliographies.
Color
"A System of Color Identification for Bibliographical Description"
was published
in Studies in Bibliography, 20 (1967), 203–234,
and reprinted
in my Selected Studies in Bibliography
(1979), pp. 139–170. My recommen-
dation of the ISCC-NBS Centroid Color Charts, emerging from an extensive
survey of
possibilities, was clearly the correct one, and those charts have
gained fairly
wide acceptance among bibliographers since then. Their use
was endorsed in Philip Gaskell's
A New Introduction to Bibliography
[1972,
1974], pp. 238–239, and, for the
book-conservation community, in Abbey
Newsletter for
December 1980. A list of eighteen bibliographies using the
charts was
published in 1990 by Craig S. Abbott in
"Designating Color
in Descriptive Bibliography: The ISCC-NBS Method in Practice"
(Papers
of the Bibliographical Society of America, 84:
119–129).
Abbott also explained some of the mistakes made in those
bibliog-
raphies in the use of the charts, the most basic one being a failure
to
understand that each chip represents the center of the color-name
block
designated by the accompanying number and abbreviated name.
Thus
bibliographers who say their matches are inexact do not recognize
that
their choice of the closest match in each case does in fact provide
an
exact match to the area surrounding the centroid chip. I made this point
in
1967, but it cannot be stated too often, considering how
frequently
bibliographers have misunderstood it. (Although Abbott objects to the
phrase "centroid number"—and the citation of a
specific number in the
form "Centroid 191"—as indicative of this confusion, I
think the usage
can be condoned as a shorthand version of "the number associated
with
a chip in the ISCC-NBS Centroid Color Charts.") Abbott further notes that
bibliographers have sometimes
misinterpreted the abbreviations assigned
to the color-name blocks and expanded
them improperly—not realizing,
for instance, that lower-case letters indicate the
"-ish" form of a color
name (so that "yG," yellowish green, is a different color
from "YG," yel-
low green). As Abbott observes, many
bibliographers have seemingly used
the charts without consulting the related
ISCC-NBS dictionary, where
provides some useful cautions for bibliographers wishing to employ the
Centroid Charts.
One of the requirements I formulated for an appropriate color-
matching system
was that there should be "strong assurance of continued
availability in the
future." Because the Centroid Color Charts were
published
by the National Bureau of Standards (NBS), I believed that they met
this
requirement. Unfortunately, however, in 1983 the NBS
discontinued their
sale, pending the results of a test of the stability of the
color chips by the
color consultant H. Hemmendinger; and
in February 1984 the NBS dis-
covered that twenty-seven of the chips
(out of 251) had shifted enough to
move from one color-name block to another.
Although the charts were
included in the NBS's 1984–85 catalogue of
Standard Reference Materials
and in the March
1984 price list, they did not appear in the
October
1984 price list and have not been available from the NBS
since then. On
6 April 1984 I wrote a long letter to Richard W. Seward of the Standard
Reference Materials
section, explaining why the charts were essential for
bibliographical work and
(recognizing that bibliographers formed a small
constituency) making the obvious
point that a color standard was needed
in many areas of endeavor. It seemed to me,
I said, that a "national
bureau of standards" could be expected to provide a
standard for color
designations. I received no reply, and three months later (on 9
July 1984)
I telephoned Stanley D.
Rasberry, chief of the Office of Standard Refer-
ence Materials, with whom
I had a cordial but unencouraging conversa-
tion. He confirmed that distribution
of the charts had stopped, and he
said that the only hope for any resumed
production (since the NBS did
not have its own color laboratory) was the receipt
of private funding.
There the matter has rested, except that over the years several
persons
associated with the development of the charts have privately sold
copies
from the small stock that remained (after the distribution of many
cop-
ies by the NBS), out of an original edition of 20,000. The most recent
of
those persons has been Nick Hale, who had been technical
director
of the Munsell Color Company at the time it oversaw the production
of
the charts (by the Tobey Color Card Company) and who had alerted the
NBS to
the problem of shifting in 1983. Hale's
possession of the stock was
publicized to the bibliographical world on 19 July
1996 by Sandra Alston
(of the Fisher
Library at the University of Toronto) through the email
list of the
Bibliographical Society of Canada. When Hale (then of Hale
Color Consultants, Inc., which is now in Naples, Florida) sent out
copies,
he included with each set a statement about the color shifts, along with
a
thorough report of the results of the 1984 measurements; and as of
June
2003, at which time about 150 sets were left, he also enclosed
a brief ac-
count of the history of the Centroid
Charts.
The significant color-chip shifts discovered in 1984—that is,
the
twenty-seven that shifted from one color-name block to another—are
as
follows: the chip for block 24 shifted to block 235; 56 to 59; 59 to 62;
75
to 78; 108 to 110; 118 to 125; 126 to 138; 129 to 131; 139 to 141; 147
to
166; 194 to 197; 195 to 199; 196 to 200; 197 to 201; 198 to 202; 200 to
204;
206 to 210; 208 to 212; 213 to 202; 215 to 229; 218 to 223; 220 to
225;
221 to 226; 225 to 243; 239 to 259; 247 to 250; and 257 to 260.
Persons
who consult descriptive bibliographies that cite centroid numbers
(and
the corresponding color names) should be aware that these
twenty-seven
numbers may not be accurate (depending on the date of the
bibliogra-
pher's work). But given the approximate nature of such citations in
the
first place, along with the fact that the only major chip shifts are to
closely
related color blocks, the shifting is not likely to render any
previously
written description seriously defective. The copies of the charts that
are
used in connection with reading these bibliographies may not of
course
precisely match those that were used by the bibliographers, but this
situ-
ation is not a serious problem. In 2003, Hale expressed the belief, based
on his long experience, that the chips
had probably changed very little
since 1984 and thus were still
usable.
Whether new descriptions, however, should be based on the Centroid
Charts is a question that one might at first think ought to be
answered in
the negative. After all, they are not readily available, and at least
twenty-
seven of the samples in them are not accurate. But it may be that there
is
no satisfactory alternative from a practical point of view. The two
main
contenders are the charts produced by the Munsell Color Company and
by
Pantone, Inc. Munsell has tradition behind it, since the Munsell system
is the
fundamental one of the twentieth century (see my discussion of it in
the
1967 essay) and is in fact the system on which the centroid sampling
of
the color solid is based. And the Munsell Book of Color
is published in two
editions, one with glossy and the other with matte samples
(over 1600 of
them). The fact that matte samples are available and
that the samples are
removable makes the Munsell Book
especially appropriate for bibliographi-
cal use (and it is indeed widely employed
for scientific as well as industrial
purposes). It is unwieldy, however, and at
present each edition costs about
$1000. It is not suitable for bibliographers to
carry easily with them,
and most will not wish to spend the money on a copy.
Libraries could
of course be encouraged to purchase copies for their
special-collections
reading rooms, but inevitably bibliographers will still find
many places
without a copy.
Pantone is more commercially oriented than Munsell, but bibliogra-
phers will be
interested in the fact that the Pantone system is frequently
used in the
graphic-design and printing industries. There is a wide variety
of Pantone
Matching System Color Guides for different purposes; most
or uncoated samples, identify each one with a Pantone number and name,
and cost in the $100–$500 range. A loose-leaf book is also available, as
well as apps for mobile devices. And there are several websites (such as
www.pantone-colours.com) that offer comprehensive showings of Pan-
tone colors, each labeled only with a Pantone number; but matching to
a computer screen is not recommended, since colors can vary from one
screen to another. (Pantone is a subsidiary of X-Rite, which offers spec-
trophotometers, but besides their costliness they go beyond the specificity
required for bibliography.) A compact and convenient book does exist,
called The Pantone Book of Color (1990), by Leatrice Eiseman and Lawrence Herbert, who provide an informative introduction. It displays 1024 col-
ors labeled with Pantone numbers and with names; the Pantone names,
however, are often fanciful—such as "dusty jade green" or "moonlite
mauve" or "apricot wash"—and are not appropriate for bibliographical
citation.
The matter of naming is vitally important, since a description must
enable one to
visualize a color without recourse to a chart (though a chip
number is there for
times when greater precision is needed). The merit
of the ISCC-NBS naming system
(used in the Centroid Charts) is that only
commonly
understood adjectives are used. The ISCC-NBS dictionary
(discussed in my essay)
enables one to convert Munsell notations as well
as centroid numbers to the
standard color names; but the use of Pantone
numbers (though it would allow easy
access to an approximation of the
colors on the internet) would not accommodate
the use of those names.
Taking all these considerations into account,
bibliographers would be
well advised to follow one of two courses. (1) Continue to
use the Centroid
Charts, since 89.2% of the chips are
probably still accurate (in that any
shifting has not moved them out of their
color-name blocks) and since the
other 10.8% can be assumed to have shifted to
such closely related color-
name blocks as not to invalidate a description (which
in any case refers to
a segment of the color solid, not a single point). (2) Use
the Munsell Book
of Color when convenient to do so,
supplementing it at other times with
the Centroid Charts;
combining the two in a single bibliography poses no
problem as long as all the
names are expressed in the ISCC-NBS standard
forms. Besides the ISCC-NBS
dictionary, another source for the equiva-
lences is a website provided by the
Texas Precancel Club, which lists the
centroid numbers and names along with the
equivalent Munsell nota-
tions and small color samples: search the internet for
"NBS/ISCC Color
System - Tx4.us", or use "tx4.us/nbs-iscc.htm". An example of a
bibliog-
raphy that uses the centroid system and also supplies the
corresponding
Munsell notation in each instance (as "perhaps more accessible") is
David Alan Richards's
Kipling (2010), which thus sets a valuable
precedent.
Patterns
The essay I wrote on "The Bibliographical Description of Patterns"
(covering
cloth grains and marbled papers, as well as the stamping applied
to the cloth for
particular editions) appeared in Studies in Bibliography,
23
(1970), 71–102, and was reprinted in my Selected Studies in Bibliography
(1979), pp. 171–202. Since
that time there has been a great deal of interest
in the description of
nineteenth-century publishers' bindings; but there
is one publication that stands
out as a promising guide for future use. It
is Andrea
Krupp's
Bookcloth in England and America, 1823–50
(2008), a
slightly revised version of an article in Papers of the Bibliographical Society of
America, 100
(2006), 25–87. This work emerged from an ongoing project
by two
conservators (Krupp and Jennifer Woods
Rosner) at the Library
Company of Philadelphia called "The Database of
Nineteenth-Century
Cloth Bindings (available online via the Library Company's
website, under
"Collections," then "Conservation Department Research on
Bindings").
Based on the Library Company's holdings, each entry in this
excellently
designed database includes more than sixty fields of data on binding
cloth
and design, binding structure, sewing patterns, and endpapers,
providing
a wealth of information that is beginning to make possible a more
detailed
understanding of the history of publishers' bindings and a more
exact
determination of the dates of undated books or successive
bindings-up.
The heart of the publication (in both 2006 and
2008) is a "Catalogue
of Nineteenth-Century Bookcloth Grains"
containing actual-size color
photographs of 125 grain patterns (plus six
variants)—far more than have
been published in previous sets of photographs for
bibliographical use.
Krupp thinks of her book as a
"'field guide' for identifying and dating
nineteenth-century bookcloth."
Each photograph is labeled with a reference code and a descriptive
name. The
basic names are the ones previously used by Carter, Sadleir,
Gaskell, Ball, and me, though some of the modifying adjectives are
new,
and "moiré" is taken as a family name, not a modifier. (Krupp claims that
I use "a different approach"—different from supplying
names—by as-
signing "letter and number codes"; but I emphasized in my essay, as I
do
again now, the importance of a standard terminology consisting of
com-
monly understood words.) The codes consist of three-letter
abbreviations
of the grain families, with attached numbers reflecting the
arbitrary order
of the photographs within each family. Thus "Ban8: Fine dotted
ribbon"
is the eighth photograph in the section devoted to the "Bands" family,
and
"Wav3: Diagonal wave" is the third in the "Waves" section. The
twelve
families are Bands; Beads; Diapers and Diamonds; Hexagonal;
Leather
Textures; Moiré; Nets and Meshes; Ribs; Ripples; Sand, Pebble,
Bubble;
Waves; and Weaves and Checkerboards—supplemented with a
"Miscel-
laneous" section and one called "Winterbottom," which shows sixteen
have appeared in reference works and thatrepresent versions of patterns
used in the late nineteenth or early twentieth centuries. This "Catalogue"
is accompanied by a "Table of Nineteenth-Century Bookcloth Grains,"
which for each grain states thenumber of examples thus far entered in the
database and their date range (extremely useful information) and then, for
any grain illustrated in one of seven previously published sets of photo-
graphs (Carter, Sadleir, BAL, Tanselle, Gaskell, Ball, and Winterbottom),
gives the corresponding notations. There is also a separate set of color
photographs called "Catalogue of Nineteenth-Century Ribbon-Embossed
Bookcloth," consisting of 122 examples (plus six variants) of abstract, flo-
ral, and geometric patterns that are "larger than 6 mm. per repeat."
The result is a reference work that surpasses in various respects any-
thing
available before—though any of the earlier reference works can
still be used,
preferably with Krupp numbers also cited (and her
"Ta-
ble" makes it relatively easy to give such correspondences). But
inevi-
tably questions may be raised. Although Krupp has
not tried to create
a logically comprehensive system of classification (simply
listing families
of patterns in alphabetical order), one may still question the
inclusion
of a "Miscellaneous" section. And the "Winterbottom" section
results
from her belief that merging Winterbottom's late
nineteenth-and early
twentieth-century versions of patterns into the sections
displaying earlier
versions would be misleading. But its presence shows that the
"Catalogue
of Nineteenth-Century Bookcloth Grains"is not meant to
be limited
to the 1823–50 period (the dates in the title of the book), even though
the
bulk of the examples in the "Catalogue" (and "Database") are from be-
fore
1869, reflecting thework thus far done with the holdings of the
Li-
brary Company. The insertion of the Winterbottom examples in their
proper
places, however, would not mislead anyone since their dates (like
the dates of all
the other examples) are included in the "Table"—and
perhaps all the dates should
be given in the captions to the photographs
as well. Taking this step would make
clearer the way in which Krupp's
"Catalogue" can
gradually be expanded in the future.
The same thoughts are raised by Krupp's segregation of
ribbon-
embossed patterns into a separate catalogue, using a different style
of
codes for patterns (and for variants) and providing no names. She
rec-
ognizes that "another option would be to give each pattern a name (or
a
code) and integrate them fully into one large table…. There is
certainly
much more work to be done in that area than the amount of attention
I
have given them here" (p. 16). It is worth noting that when she
generalizes
about the number of patterns, she does include ribbon-embossed
exam-
ples in the totals. (One could argue, as I did in note 16 of my 1970 essay,
ally for each book; yet ribbon-embossed patterns, being a characteristic of
cloths prior to their use on individual books,can logically be distinguished
from the frames and lettering stamped into the cloth for those books. I will
add here that the basic cloth weave can be a distinguishing feature even
for embossed cloth and should not be overlooked when examining the
embossed pattern.) Incorporating the "Miscellaneous," "Winterbottom,"
and ribbon-embossed examples into the main body of the "Catalogue"
would be a first step toward making Krupp's system (however arbitrary) a
more welcoming and appropriate framework for future expansion.
The most important discussion of Krupp's work is B. J. McMullin's
"Patterned Book Cloth: A Review Essay," in
Script & Print, 32 (2008),
163–175. After a
historical sketch of the development of bibliographical
thinking on the subject
(containing some trenchant observations), McMul-
lin brings up the basic problem raised by the proliferation of
examples
shown by Krupp: should the set of photographs
be indefinitely expanded
as additional grains and variants are noted that do not
precisely match
any of those already illustrated? The fact that some patterns are
provided
with an additional photograph or two labeled "var1" or "var2"
implies
that such expansion is expected. It is not clear, however, why
additional
photographs would necessarily be preferable to added words for this
pur-
pose. As McMullin notes, the situation is
epitomized in the "diaper" cat-
egory, where the illustrations are for ultra-fine
diaper (defined as 20–22
units percentimeter), fine (10–15/cm.), medium (8–9/cm.),
coarse and
smooth (4/cm. each), and (anomalously)"5.5 diaper." The precedent
set
here would require more photographs for diapers that fall between 15
and
20 units per centimeter, or between 4 and 8 (except for 5.5). But a
series
of illustrations accomplishes nothing more than is achieved by simply
using
the word "diaper" followed by the measurement. A photograph of
diaper
would still be needed, but additional photographs serve no purpose
(and
indeed can be a complication, since the closest match at one time
might
no longer be the closest match after the insertion of
morephotographs).
The "Beads," "Nets," and "Ribs" categories offer exactly the
same situ-
ation, but in all categories the adjectives of degree ("coarse,"
"fine," etc.)
do not cover all the possibilities.
Such omissions occur, of course, because the photographs represent
the books thus
far examined for the "Database"; but a comprehensive
"field guide" to a class of
artifacts cannot be constructed in that way
because, even at the price of
continued revisions, it will always be unsat-
isfying in certain situations. The
"Database" itself is a great accomplish-
ment; and as it grows—incorporating, one
hopes, records from many
collections other than the Library Company's—it will
become even more
bindings can be drawn. And it is appropriate that a field guide should be
a by-product of such a database. But the relation between the two needs
to be further thought through. In one respect the field guide is bound
to grow as the database does, because a field guide should illustrate all
distinct patterns, however rarely used some of them are. But in another
respect it should not grow, for if it attempts to register variations of dis-
tinctpatterns it will become more bewildering than helpful. Deciding
which patterns are "distinct" may to some extent be debatable, but it is
not nearly as intractable a problem as defining variations within them,
since the latter are nearly infinite in number. A classification scheme, by
definition, should deal with diversity not by reproducing it but by pro-
viding a path through it. Contrasting the classification of bookcloth with
that of color helps to illustrate the problem. As I pointed out in my 1970
essay, color is anatural phenomenon with definite limits and a continuum
within those limits; one can thus scientifically sample that continuum at
spatial intervals. Cloth patterns do not offer this opportunity, for it is
not possible to determine "centroids" within their variations. One can of
course subdivide patterns according to measurements of their repeated
units, as Krupp has done in a few instances (though her measurements are
not based on regular intervals); but, given the essential difference between
patterns and color, the result would not provide a necessary visual aid,
as centroid colors do.
Since the large number of photographs in Krupp's
"Catalogue" is not
entirely an advantage, bibliographers may well find it
preferable to limit
themselves primarily to her examples inwhich the name of the
grain
has no adjective or measurement attached and then to add
whatever
modifiers seem necessary on a given occasion—as in "coarse (8/cm.)
Rib
(cf. Krupp Rib3)." The fact that Krupp's photographs are actual size al-
lows one to know(through direct
measurement of unit-repeats in cases
where a measurement is not stated) the norm
fromwhich "coarse" and
"fine" variants depart. (The use of "cf." in my example
seems desirable,
though it differs slightly from my 1970 suggestion. Even when the
specific
variant is pictured in Krupp. it might be
better to cite only the Krupp
photograph of the basic
pattern, both for uniformity and to allow for
possible future revisions in Krupp's "Catalogue.") Bibliographers could
also do their
matching against the photographs I have published or those
provided by Gaskell (see below), since the smaller number of examples
in
those sets (36 and 29 respectively) do after all include the great
majority
of patterns found on nineteenth-century books and since the basic
names
given in them are the same as those used by Krupp.
This last point is not crucial, however, because Krupp's
"Table" al-
lows one, when consulting (or writing) a bibliography employing any of
grain in the "Catalogue"—and to find it online. Indeed, it is the presence
online of the "Catalogue" (along with the associated "Database") that
makes Krupp seem the most convenient reference for bibliographers to
cite, despite its limitations. One hopes that the "Table" will be added to
the online resources, since without it one still has to turn to the printed
book (or article) to make conversions between Krupp's code and earlier
ones and to learn the date range of each grain; and the measurement of
repeated units in every grain should be stated, since the pictures on the
website are no longer actual size.Presumably the online version will be
continuously revised to incorporate new grains (as the database grows)
and perhaps to rectify some of the flaws in the printed versions. Most
bibliographers and users of bibliographies are likely to have a portable
electronic device at hand as they examine books, or will not find it incon-
venient to make sure that they do.
Bibliographers should be aware of two other very large databases of
bindings that
allow for searching by a particular cloth grain; although
they give one access to
many images of bindings, neither is as appropriate
as the Library Company's for
bibliographers to use as a basic reference
in identifying the cloth on a book in
hand. One is "Publishers' Bindings
Online 1815–1930: The Art of
Books" (on the website of theUniversity
of Alabama Library and the University of
Wisconsin—Madison Library,
the partners in this project). It does provide (under
"Research Tools") a
"Gallery of Book-Cloth Grain Patterns" showing fifteen basic
grains la-
beled with standard names; each image can be enlarged and offers a
link
for searching the database for examples of books bound in that cloth.
The
notes on each book include the color and grain of the binding but do
not
come close to providing the amount of technical detail about binding
and
endpaper structure that the Library Company database does. Its emphasis
is
not primarily on bibliographical research (there are artistic, historical,
and
literary "galleries," each with essays and teaching tools), but it does
offer a
"gallery" of bindings by decade, accompanied by commentary.
The other database is
the British Library's "Database of Bookbindings"
(on the Library's website, under
"Catalogues," then "Bookbindings"),
which has excellent technical notes, including
the identification of cloth
grains bystandard names. But since it covers the whole
history of book-
binding as represented in the Library's collections, it contains
a relatively
small selection of publishers' cloth bindings, and it provides only a
verbal
(not a pictorial) guide to cloth-grain names for purposes of searching.
In addition to Krupp. there are three other significant
post-1970 treat-
ments of cloth grains. Two years after my essay, Philip Gaskell's
A New
Introduction to Bibliography
(1972, 1974) appeared, containing a section on
"Publisher's Cloth in Britain and
America" (pp. 238–247). Besides giving
the ISCC-NBS color system), Gaskell provides a classification of grains
and a set of photographs. He follows ingeneral my system of classifi-
cation and the naming of individual grains, though he leaves out one
("weave"), adds four ("beaded-line," "crocodile," "crackle," and "frond"),
and moves two to a different family ("bubble" becomes one of the "sand-
texture" grains, and "cord" becomes "ribbed-morocco" under "leather-
texture"). As for associated codes, Gaskell simply assigns each photograph
a number according to its place in the sequence of illustrations in his book
as a whole (beginning with "Fig. 84"). Thus his codes (unlike mine) do not
reflect a classification scheme; butthat fact is of little significance since
the order of the photographs (unlike Krupp's arbitrary order) does follow
his classification and brings related families together.
Sensibly he provides only a single photograph for each basic pattern—
that is,
only for grains whose names do not have adjectives of degree at-
tached to them.
He understands that grains "notably fine or coarse," for
example, can be specified in words, without the presence of
photographs
(even though, as in all systems, the norm represented by each
photograph
hasnot been precisely determined). My 1970 set of photographs
does
include a few fine, coarse, diagonal, and moiré examples for
illustrative
purposes but without suggesting that such photographs are necessary:
I
made clear that adjectives should be attached (as needed by the level
of
precision required) even when there is no matching photograph—and
that,
when still greater precision is desirable, a bibliographer may have
to supply one
or more additional photographs. The only thing I would
now add—whether one is
using my system or Gaskell's or Krupp's—is
that the greater precision sometimes needed to distinguish,
for example,
separate bindings-up (or secondary bindings) can often best be
achieved
by providing a measurement. On the subject of measurement, I
should
perhaps say that when a published photograph is intended only for
gen-
eral matching of a type of design (as opposed to serving as a record
of
the binding of a specific book), one does not need to know whether
the
photograph is actual size or not; but a measurement scale attached to
a
photograph (as in Krupp's basic catalogue but not her
catalogue of
ribbon-embossed patterns) is nevertheless helpful in determining when
a
match has been made.
Thirteen years after Gaskell's book, Douglas Ball published Victorian
Publishers' Bindings (1985), a thoughtful and well-documented
study that
includes a chapter entitled "The Decoration and Graining of Cloth
and
ItsBibliographical Importance" (pp. 24–31). In that chapter, Ball
points
out the potential (but not always conclusive) role of precise grain
identifi-
cations in the process of distinguishing and dating separate
bindings-up;
names of grains. But the most useful parts of his book for descriptive
bibliographers are two of the appendixes. The first, "Identification of
Cloth Grains" (pp. 123–129), besides pursuing in more detail the limita-
tions of some ofthe previously published photographs and the variations
in assigned names, offers descriptive comments on the basic families of
patterns. One of the emphases throughout—and properly so—is on the
importance of measuring the frequency of repeated elements in designs
in order to give precision to any modifying adjectives and the norms
they relate to. He suggests approximate norms for rib grain (11 repeats
per centimeter), ripple (25+/cm.), dot-and-line (perhaps 2.5–3.5/cm.),
checkerboard (perhaps 4/cm.), diaper (6/cm.), and bead (5.5/cm.). Then
he offers a particularly helpful discussion of the morocco family (pp. 126–
128), dividing those with "linearly directional" ribs into ribbed morocco
(with basically continuous ribs and converging channels, 8–9/cm.), cord
(with ribs separated by cross-creases, 8–9/cm.), and parallel cord (with
channels that generally do not converge). These measurements may be of
use to bibliographers in deciding when to add such adjectives as "fine";
but those adjectives can merely be suggestive, and precision can come
only with appended measurements. The second appendix, "Further Data
on Grains" (pp. 130–142), consists of two parts, the first being a table of
"Dates and Frequency of Use" (pp. 132–136), which assembles informa-
tion drawn from some 2600 entries in the first volume of Sadleir's XIX
Century Fiction. The data given here, when combined with those in Krupp's
"Table," form essential context for discussions of dating. The other part
of the appendix, "Some Additional Grains" (pp. 137–141), describes and
illustrates seven grains not dealt with in previous sources—but now incor-
porated into Krupp's guide. (Bibliographers will also wish to know of the
remarkable amount of detail in another appendix, "Nineteenth-Century
Edition Binders' Signatures," pp. 168–192).
The other of the significant post-1970 books is William
Tomlinson and
Richard Masters's
Bookcloth 1823–1980
(1996). Primarily a detailed
and
illustrated history of the bookcloth industry, it contains much
information
not otherwise available, drawing on the knowledge and materials
that
Tomlinson acquired in a sixty-year career in the
field, part of the time
with the Winterbottom Book Cloth Company of Manchester, a
promi-
nent supplier of bookcloth after 1853 and the dominant one
from 1891
to 1980. There are two chapters of particular interest for
bibliographers,
"Description of Bookcloth Qualities: (pp. 86–107) and
"Identification of
Designs (Grains)" (pp. 108–123), the first supplemented with
thirty-six
tipped-in cloth swatches and the second with thirty, dating from
after
1891. The first group is instructive in allowing one to feel
some of the
variety was available (some fifty qualities are described) cannot be put
to direct use by bibliographers, who deal with cloth attached to covers.
The second group, showing patterns ("designs"), turns out not to be more
useful for identification than good photographs. In any case, the late date
of all the samples means that this book cannot offer an appropriate gen-
eral standard for bibliographers—along with the fact that only arbitrary
letter codes are used (though they were employed, with some variation,
in the Bibliography of American Literature). What the book does accomplish
is to make clear to bibliographers how inexact all their descriptions are
when compared to what was necessary in the cloth trade: in 1948, for
example, Winterbottom's advertised "41 Qualities, 71 Designs and 688
Colours, equal to almost 50,000 different Effects." (Despite this profu-
sion, twentieth-century books in general use rather plain cloth and do
not display the variety of patterns found on nineteenth-century books.)
The authors recognize that bibliographical description does not normally
require the same precision as that employed by a cloth supplier,but they
add that using cloth as evidence for dating may well require something
closer to it.
In my 1998 review of this book (Printing History, 19.1: 39;
reprinted
in my Portraits and Reviews [2015], pp. 345–347),
I cited an example of
the kind of information it supplies that is especially
valuable for bibliog-
raphers: sometimes it was more economical to cut the cloth
in a "two-
way" fashion rather than "one-way," with the result that the cloth
pattern
might run vertically on some copies of an edition and horizontally
on
others (and thus the two directions may not signify separate bindings-
up).
A footnote to the Tomlinson-Masters bookappeared in 2002 in the
form of an article
by Willman Spawn and Thomas E.
Kinsella ("The
Description of Bookcloth: Making a Case for More
Precision," Papers
of the Bibliographical Society of
America, 96: 341–349), which describes the
William
Tomlinson Bookcloth Collection at Bryn Mawr College and lists
the
forty-three pattern books it contains (dating from the mid-1920s to
the 1980s),
plus Winterbottom's 1958 code book. The authors make the
point (previously
observed by Tomlinson and Masters)
that "established
bibliographical conventions do not account for the known
complexities
of bookcloth"; they grant that descriptive bibliographers may not
usually
need to go beyond those conventions, but they state (what is clearly
true)
that book history in general would benefit from research in such
archives
as exist, for the purpose of discovering the manufacturer, date, and
trade
specifications for the cloth found on particular books.
An additional work displaying cloth grains, but obviously not a can-
didate for a
bibliographical standard, is Geoffrey Wakeman's
Nineteenth
Century Trade Binding (1983). Its second volume
is a folder containing a
mens, plus four pages showing seventy-one rubbings of Winterbottom
designs. Support for Krupp's system comes from another publication:
Carlo Dumontet's "Nineteenth-Century Bookcloth Grain Classification
and the Special Collections Cataloguer" (Papers of the Bibliographical Society
of America, 104 [2010], 105–112), which recommends the insertion of her
system of naming into the thesaurus of controlled vocabularies for cata-
loguers, especially the "Art & Architecture Thesaurus Online" (on the
Getty Research Institute's website) and the Rare Book and Manuscript
Section's "Binding Terms" (on its website). The latter is perhaps now
superseded by the online "Language of Bindings" thesaurus (established
through Nicholas Pickwoad's Ligatus project at the University of the Arts,
London). Some of the problems Dumontet outlines might be alleviated by
the adoption of my suggestion (above) of making the unmodified terms the
standard, with adjectives and measurements added in specific instances;
thus a full verbal description, with as much modifying detail as necessary
and with a Krupp reference (including "cf."), could appear in MARC
field 563, and the unmodified form (with the possible local addition of a
Krupp code) in field 655.
Besides describing cloth grains, the bibliographer must also record the
designs
(abstract and pictorial), frames, and lettering that are stamped
(in blind or
gold) into the cloth for individual editions. My suggestions in
1970 are probably
sufficient for most occasions; but a proposal for greater
detail has been offered
by Gene G. Freeman in "Descriptive Standards
for
Publisher's Bindings: Preliminary Notes" (Trade Bindings
Research Newslet-
ter, 5 [June 1992], 3–10). Freeman recommends dividing
each cover into
twenty-four "cells" (six rows of four) and each spine into twelve
(six rows
of two) and then estimating the percentage of the area in each cell
filled
withdecoration or lettering. An example is "1ABCD—Author's
name,
centered, 1A and 1D < 10%." But surely, if "centered" is not
enough,
it would be far more accurate simply to measure the distance
fromeach
end of the name to the edge of the cover. He also thinks that a
diagram
showing the "estimated fillfactor" of each cell "should be included in
a
descriptive bibliography" because the "distribution of decoration over
the
book's surface" can help to attribute unsigned bindings to particular
designers.
Few people, I imagine, will agree with this point, and I see no
reason for
bibliographers to adopt either of his recommendations. The
same issue of this
newsletter contains another short article with a promis-
ing title, David B. Ogle's "Uniform Notation for Describing
Decorative
Trade Bindings" (pp. 13–15); but the proposal (assigning a five-letter
code
to each book, the first two letters abbreviating the color and the last
three
the type of design) is only intended as a means for collectors to
locate
their books—though presumably such a code, if searchable in a library's
two issues of the newsletter ("Further Binding Description Comments," 6
[September 1992], 4; 7 [December 1992], 6) responds to Freeman's and
Ogle's articles by arguing for "a simple account, specifying cloth texture
and color, and describing the design [in words]."
Bibliographers who wish to place individual designs in a broader his-
torical
context have a number of well-illustrated sources to turn to. Al-
though the
leading American historian of publishers' bindings, Sue
Allen,
did not live to produce the comprehensive work she had planned,
her
publications are always worth seeking out (some are not easy to
locate).
Especially helpful are her Victorian Bookbindings: A
Pictorial Survey (1972,
1976), with its succinct history and 241
illustrations on microfiche,and
American Book Covers
1830–1900
(1998), an accordion-folded brochure that
includes
a time-line and twelve photographs of grains. (She also wrote an
essential
article, "Machine-Stamped Bookbindings, 1834–1860," which
includes a list of
American binders' signatures, for the March 1979 issue
of Antiques,pp. 564–572.) The equivalent of a time-line is provided by
the
"Keys to Identifying Covers" ofeach decade from the 1830s to the
1890s
in Calvin P. Otto's
Onlyin Cloth (1998).
Other books with many excellent illustrations are Eleanore
Jamieson,
English Embossed Bindings,
1825–1850
(1972); Ruari McLean, Victorian
Publishers' Book-Bindings in Cloth and Leather
(1973); Ellen K. Morris and
Edward S.
Levin, The Art of Publishers' Bookbindings, 1815–1915
(2000);
and Edmund M. B. King, Victorian Decorated Trade Bindings, 1830–1880:
A Descriptive
Bibliography (2003). One other work that should be men-
tioned, even though
it contains few illustrations, is Robert Lee
Wolff's
five-volume Nineteenth-Century Fiction: A
Bibliographical Catalogue (1981–86),
simply because it provides such a
large sampling of books with binding
descriptions (nearly eight thousand entries).
Wolff says he was "guided
by Sadleir's practices," and
he generally does use Sadleir's names for
grains.
Illustrated studies of individual designers also exist: on Margaret
Armstrong (by Charles Gullans and
John Espey, 1991), for example, or
on John Feely (by Sue Allen in a 1994
Clark Library pamphlet, Decorated
Cloth in America), on
Frank Hazenplug and Sarah
Whitman (by Gullans in
the same pamphlet), and
on Alice C. Morse (by Mindell
Dubansky, 2008).
Samples of the work of nineteen designers are listed
(but not illustrated)
in an exhibition catalogue from California State University at Fullerton,
Decorative
Approach to Trade Cloth Book Binding (1979).
When describing the ornamental designs stamped on grained cloth
for individual
editions, bibliographers should be aware of a small number
of bindings, used for
only about a decade after 1835, in which the cover
and spine designs were embossed
on the cloth before it was fastened
to the boards and spine strip rather than (as
was usual) being stamped on
surface, whereas stamped designs are pressed below the surface.) Because
the embossed designs were placed on the cloth in advance, the positioning
of the cover and spine elements obviously determined the size of book-
block that could be accommodated; sometimes when the match was not
exact, the designs are off-center on the finished book. These bindings are
discussed, and thirteen designs illustrated, by Andrea Krupp and Jennifer
Woods Rosner in "Pre-Ornamented Bookcloth on Nineteenth-Century
Cloth Case Bindings" (Papers of the Bibliographical Society of America, 94
[2000], 176–196). When one encounters a binding of this kind, one should
compare it with the Krupp-Rosner illustrations; and if it is different, one
should report it, in order to increase the inventory of the relatively small
number of such designs. (Occasionally an embossed designwas applied
to the cloth on an assembled binding case, but those instances can be
identified by the factthat the intaglio plate flattens the cloth grain and
thus leaves an edge mark.)
My 1970 essay also covered the patterns of marbled papers, used both
for covering
binding boards and for endpapers. Because every piece of
hand-marbled paper is
unique, any matching that a bibliographer does
can only be approximate (no matter
how large the sample), and the twelve
illustrations I provided, divided into two
classes and given their tradi-
tional names, are perhaps sufficient (along with my
suggestions for addi-
tional verbal modifiers). But if one wishes to make
comparisons against a
larger selection, one should turn to Richard J. Wolfe's great book, Marbled
Paper: Its
History, Techniques, and Patterns (1990), which contains (following
p. 186)
one hundred ninety-two excellent color illustrations, each assigned
a number and
(in a supplementary list) a name (the same traditional name
I had used, with many
alternatives and with useful historical notes). In-
stead of attempting a
classification of patterns, Wolfe arranges the
samples
according to "their development and manufacture in various
geographi-
cal locations," and thus types of patterns are not kept together. But
the
advantage is that the bibliographer who uses these samples in
conjunc-
tion with Wolfe's discussion of them (in "The
Evolution of Marbled Pat-
terns," pp. 179–192) has a basis in some instances for
the approximate
dating of papers. Although that possibility is obviously most
useful in
connection with pre-1820 books (before the period of publishers'
bind-
ings), nineteenth-century publishers did often make editions available
si-
multaneously in cloth and in "half-calf" with marbled covers, which can
be
difficult to identify. In any case, the wide availability of the Wolfe book
(along
with its unquestionable authority) makes it an appropriate refer-
ence for
bibliographers to cite when describing marbled paper.
Other sets of color illustrations of marbled patterns have been pub-
lished
(usually in conjunction with marbling instructions), but none
of marbled specimens in Anne Chambers, The Principal Antique Patterns
of Marbled Papers (1984); fourteen good color photographs of "very basic
patterns" in Iris Nevins, Traditional Marbling (1985, 1988), photographs
that are also included in her catalogues of papers available for sale;
ninety-three pages of illustrations in Barry McKay, Patterns and Pigments
in English Marbled Paper (1988); and sixty-seven large color illustrations
in Einen Miura, The Art of Marbled Paper (1988; trans. 1991). Attempts
at classification are less common. Phoebe Jane Easton has a chapter on
"Nomenclature of Marbled Patterns" in Marbling: A History and a Bibli-
ography (1983)—reprinted in Ink & Gall, 7.2 (Winter 1993), 7–9—which
summarizes my system. The same issue of Ink & Gall also includes Carina
Greven's "The Development of a Standard Nomenclature for Marbled
Paper" (pp. 10–13), which includes a classification based on the "evolv-
ing nomenclature" developed by the Belgian-Dutch Bookbinding Society.
This system, however, simply enumerates seven types of patterns, in over-
lapping categories: (1) pebble or stone; (2) drawn (using a tool), subdivided
into ten patterns, such as curl, feather, and peacock; (3) combed (actually
a type of "drawn"); (4) shadow (or "Spanish"), usually made on pebble
(and when not, they fall into the sixth group); (5) fantasy (either abstract or
pictorial); (6) combination (of two or more patterns); and (7) overmarble
(marbled more than once). The object here is not primarily to provide
a classification scheme (and this is not a satisfactory one) but rather to
establish an international terminology (with equivalents in several lan-
guages listed). This project is now fully reported in Elly Cockx-Indestege,
Sierpapier & Marmering: Ein Terminologie voor het beschrijven van Sierpapir en
Marmering als boekbandversiering (1994).
In addition to marbled paper, nineteenth-century publishers often
used
floral-patternedpapers for endpapers, and the basic article on this
subject is
Sue Allen's "Floral-Patterned Endpapers in
Nineteenth-Century
American Books" (Winterthur Portfolio,
12 [1979], 183–224). Allen divides
the patterns into eighteen categories (which go
beyond the strictly floral):
(1) seaweeds; (2) maidenhair fern and related
designs; (3) sprigs and tiny
bouquets; (4) small flowers and leaves connected by
stems; (5) tightly fitted
flowers and leaves; (6) large, realistically drawn
flowers; (7) large, real-
istically drawn flowers showing depth and volume; (8)
"Eastlake" style
(rigidly stylized) flowers; (9) small geometrics; (10) mechanical
and imita-
tive; (11)designs on two axes; (12) Renaissance style (scrolled or
curling
leaves, sometimes with flowerheads or coiling tendrils); (13)
publishers'
initials and trademarks and similarly organized units; (14) oriental
mo-
tifs and arrangements; (15) fairy-tale patterns; (16) butterflies and
birds;
(17) typefounders' combination borders; and (18) "modern" explosive
de-
signs. (The two most commonly used styles werenumbers 4 and 8.) She
teen categories) and a "Catalog of Selected Nineteenth-Century Ameri-
can Endpapers," which contains comments on each of the illustrations
(plus a list of the binders represented). Bibliographers would do well to key
their endpaper descriptions to her illustrations, as a way of indicating the
general style of pattern and placing it in a larger context. But they may
also wish to supply their own illustrations. An example of a bibliography
that provides illustrations of decorated papers—in this case twenty-two
of those used by Knopf on covers—is Richard J. Schrader's bibliogra-
phy of Mencken (1998). Books that display decorated papers—such as
Annette Hollander's Decorative Papers and Fabrics (1971), Henry Morris's
Roller-Printed Paste Papers for Bookbinding (1975), and Decorated Paper Designs
… from the Koops-Marcus Collection (which includes some wonderful full-
page illustrations of marbled patterns; 1997, in the Pepin Press's delightful
"Design Series")—are not of particular use to descriptive bibliographers
unless they show close relatives of the papers being described.
The whole range of decorated paper (however produced) is concisely
classified by
Henk J. Porck in "Characterization of Western
Handmade
Decorated Papers: Development of a Standard Terminology" (in Looking
at Paper: Evidence & Interpretation, ed. John Slavin et al. [2001], pp. 196–
201), where six
categories are outlined: monochrome (dyed in the pulp or
painted), metal (gold or
silver), marbled, paste, block-printed, and relief-
printed (embossed).
Because my 1970 essay, besides dealing with patterns, includes (near
the end of
the first section) asample of a complete binding description, I
wish to call
attention to a feature of the binding process not mentioned
there: the pencil
marks that frequently appear near the beginning and/
or end of the printed
gatherings. Collectors and bibliographers have of-
ten noticed them, but they have
rarely been commented on—though
Roger E. Stoddard
included an illustration of one in Marks in Books, Il-
lustrated
and Explained (1995). His item 10 (p. 9) shows what he calls a
"binder's
mark"—in this case a "4" penciled in the gutter at the foot of
the last page of
the penultimate gathering. The first person to investigate
such pencil marks in
detail is Robert J. Milevski, who reported on them
(with
illustrations) in "Marks in Nineteenth-Century Trade Bindings"
(The Book Collector, 60 [2011], 41–56). He identified two categories:
"sew-
er's marks," which usually appear in the bottom margin of the last
page
of the last gathering (or the last page of an inserted publisher's
catalogue);
and "production marks," which usually appear in the lower margin
(or
sometimes one of the side margins) of the first page of the second,
third,
and/or fourth gatherings. In both cases, they take the form of a
letter,
a number, or a squiggle. Their exactfunction is not clear, but
Milevski
offers some possibilities: those at the end could identify the person who
those at the beginning could identify the collator and/or signify that all
gatherings were in order and any insertions were in place. More will be
learned about them as more are reported and examined; and bibliogra-
phers should note them in their binding descriptions, just as they would
naturally mention that better known kind of evidence of the binding op-
eration, the binder's ticket (with specific copies indicated as sources in
both cases). Milevski observes that the pencil marks are "direct evidence
of the human hand in the mass production of books in this era" (p. 47);
as such, they deserve to be recorded for further study.
Jackets
My comments on the rationale and procedure for including dust-
jacket
descriptions in a descriptive bibliography were first published as
the third
section (pp. 109–115) of "Book-Jackets, Blurbs, and Bibliogra-
phers" in The Library, 5th ser., 26 (1971), 91–134; they were
supplemented
on pp. 58–60 of "Dust-Jackets, Dealers, and Documentation" in Studies
in Bibliography, 56 (for 2003–04 [2006]), 45–140.
These discussions were
reprinted, with slight revisions, in my Book-Jackets: Their History, Forms,
andUse (2011), pp. 24–30 and 43–46.
Two recent descriptive bibliogra-
phies that can be singled out for their careful
descriptions (and photo-
graphs) of jackets are David
Supino's
Henry James
(2006, 2014) and
David
Alan Richards's
Kipling (2010). For others, see "Introduction to the
Field"
above, showing that jackets are now receiving much more attention
than
they once did. Another indication is Paola
Puglisi's
Sopraccoperta (2003);
and in 2016 Mark
R. Godburn published Nineteenth-Century
Dust-Jackets,
a narrative history that contains many more illustrations of
nineteenth-
century jackets than I provided and deals more fully than I did with
con-
tinental Europe and with unprinted jackets. Among the kinds of boxes
in
which books have been published are those that take shapes related to
the
subject matter of the books. Eugene Umberger has recently studied
this
phenomenon by writing about books on tobacco housed in such forms
as
cigar or cigarette boxes (see "Detachable Book Coverings (That
Aren't
Dust-Jackets)," The Book Collector, 65 [2016],
91–98).
My 2011 book includes a list of almost 2000 examples of pre-1901 Brit-
ish and
American publishers' book-jackets, boxes (slip-cases, boxes with
lids, etc.), and
other detachable coverings for books (such as envelopes or
overall wrappings). It
is limited to items that have printed text or deco-
ration on them, and the bulk
of the entries give some indication of the
extent and nature ofthe printed matter.
My list was not intended to be
a census but simply to report the examples I had
learned about during the
previous four decades. It is extensive enough, however,
to give an idea
of the growth in the use of book-jackets during the nineteenth
century.
consult it as an initial indication of which publishers were using jackets at
a given time, and for what kind of books, as a way of gauging whether the
particular books they are describing were likely to have been published
in jackets or boxes. If any of those books themselves happen to be pres-
ent in my list, the bibliographer will thenknow where to find a copy in a
jacket or box (when my source was an institutional library) or at least can
be sure that a jacket or box once existed (when my source was a dealer's
catalogue or private collection).
Since the publication of my book, a large number of additional ex-
amples have
been reported tome (and some further ones are listed in the
appendixes to
Godburn's book); I hope that they will eventually form a
supplementary list posted
on the website of the Bibliographical Society
of the University of Virginia, thus
increasing the body of evidence on
which bibliographers can draw. The University
of Virginia Library itself
now holds by far the largest collection of pre-1901
jackets: in 2014 it pur-
chased the collection of the dealer Tom Congalton (the
largest collection
reported in my book), and two years later it bought the second
collection
that Congalton formed, bringing the total of its holdings to well
overa
thousand volumes. Every bibliographer who is describing a book from
the
1830s onward (or indeed one ofthe annuals or children's books from
the
preceding decades) should make an effort to find out whether it
originally
appeared in a jacket or box and, if so, to locate such a copy.
| Notes on Recent Work in
Descriptive Bibliography | |