University of Virginia Library

Aldine Octavos with Copies on Royal Paper

Numerous Aldine octavo editions survive with copies on royal paper—nearly
fifty, to my knowledge, though this figure is surely too low. Those editions known
to me are listed in the Appendix. The date 1550 was chosen as term to keep
the research programme within reasonable bounds; undoubtedly the practice of
producing octavo editions with royal-paper copies continued at the Aldine press
in the second half of the century. The list is divided into two parts because in
1535 the method of printing these copies changed, and with it their external ap-
pearance. The list is based in the main on information contained in Renouard's
Annales concerning copies which had passed through his hands, or which he had
seen in public and private collections; to this material I have added informa-
tion from recent catalogues and studies by scholars conscious of the importance
for bibliographical research of information concerning leaf dimensions. Where
possible, Renouard's information has been checked against copies now in librar-
ies open to the public, or in private hands. His few errors are recorded in the


Page 97

With one exception I have only been able to inspect personally royal-paper
copies of Aldine octavos now in the British Library. Large-paper copies are in-
dicated in the General Catalogue of the British Library by the letters "L.P.", but
this indication is not all-inclusive. I found it more expeditious to examine the
Aldines included in the books left to the British Museum in 1799 by the Rev.
Clayton Mordaunt Cracherode. In terms of quantity, his bequest, numbering
about 4,500 printed books, cannot compare with the thousands of incunables and
sixteenth-century editions which came to the library in the nineteenth century
with the King's Library (1823) and the Grenville Collection (1846), but in terms
of quality the Cracherode books are exceptional: seven of the eleven royal-paper
copies of Aldine editions which I saw at the British Library have a Cracherode provenance.[27]

Royal-paper copies of Aldine octavos can usually be distinguished by their leaf
height. The height of a sheet of narrow median is c.350 mm; thus the maximum
height of a leaf of an uncut octavo printed on such paper is c.175 mm. Copies of
Aldine octavos with leaf heights greater than that have perforce been printed on
larger paper. But only from the paper evidence of several copies can one arrive
at a clear picture of the edition in question, or at least of its paper element, since
many such Aldine octavos also have vellum copies. Thus the Cracherode copy
of the 1519 Horace (BL C.19.C.5, 210 × 99 mm) is printed on two papers, one of
which has a crossed-arrows watermark, similar to, perhaps identical with, one
of the papers found in royal-paper copies of the 1534 Themistius, and the other
the watermark of a crossbow inscribed in a circle, surmounted by a fleur-de-lys.
Both show the structural features of the royal-paper formes used to produce the
Themistius papers, that is, chain intervals in excess of 35 mm, except on either
side of the watermark, where the intervals are less. Four other copies of this edi-
tion which I have examined (leaf measurements of the largest, 165 × 99 mm)
are printed on two papers with different versions of the cardinal's hat mark, one
with and one without countermark, and with chains disposed regularly at inter
vals of less than 35 mm. The two British Library royal-paper copies of the Virgil
of 1514 (1519?) (C.19.C.2, 215 × 102 mm; G9696, 202 × 95 mm), in addition to
paper with the crossed-arrows mark, contain a second paper also occurring in
the royal-paper copies of the 1534 Themistius, with the watermark of a bilateral
anchor inscribed in a circle. However, in the Trinity College Cambridge copy of
this edition (Grylls 11.25, 152 × 94 mm) we find only paper with the cardinal's
hat mark, with characteristics similar to those of the smaller-paper copies of the
1519 Horace.

And only paper evidence can unmask copies originally printed on royal pa-
per, but trimmed so severely over time as now to resemble in dimension copies
on narrow median. Such is the case of the Cambridge Ausonius of 1517 (CUL
Sel.6.55). Despite its present leaf measurements (171 × 97 mm), its origin as a
royal-paper copy can be confidently asserted because it is printed from beginning


Page 98

FIGURE 2. Virgil, Opera, 1514 (1519?), BL C.19.c.22, pp. 112v–113r.

to end on the same crossed-arrows paper used in the royal-paper copies of the
1519 Horace and the 1514 (1519?) Virgil, while four other coppies of the edition
which I have examined are printed on two papers with the cardinal's hat mark.

The pages of the typical royal-paper Aldine octavo are characterized by the
size of the lower margin, which in the Cracherode Virgil of 1514 (1519?) measures
84 mm (figure 2). At the same time, in those belonging to editions published
prior to 1535 the outer margin is always narrow, like that of a copy on narrow-
median paper. This combination of featuers gives these royal-paper octavos a
singular appearance, judged by Renouard "trop allongée et bizarre" (p. 141).
But it can be readily explained by supposing that the same formes were used in
printing the royal-paper copies and the ordinary-paper ones. Figure 3 shows a
sheet of chancery paper printed on one side with the outer forme of an octavo
imposition scheme called by Gaskell "common" but which, adopting a terminol-


Page 99

FIGURE 3. Sheet of centripetal octavo, outer forme; Gaskell, fig. 50 (by permission of Oak
Knoll Press).

ogy suggested by the German bibliographer Martin Boghardt, I prefer to call
centripetal. This is the imposition scheme used to print Aldine octavos until
1536; subsequently, until the mid 1550s, the press used the centrifugal scheme
(Gaskell's "inverted"), with pages 1, 2, 3 and 4 on the inside of the forme.[28] In
this diagram the rectangle surrounding the sheet represents (approximately) the
outlines of a sheet of royal paper, centred on the forme. If this sheet is printed
on formes imposed thus, all the leaves will have large lower margins, but only
ff. 1, 2, 3 and 4 will also have large outer margins. The outer margins of the
other four leaves will be limited to half the distance between the pages of type in
the forme. Without some intervention in the printing house a copy consisting of
sheets printed thus would not be viable. The answer was to trim the large outer


Page 100

FIGURE 4. Cicero, Epistolae ad Atticum, 1540, BL Davis 763, pp. 184v–185r.

margins so as to reduce them to the same width as the small ones, an operation
which must have been done prior to binding, and was probably done prior to
printing, to facilitate the folding of the sheet.[29]

For editions published from 1535 onwards the procedure was different. The
royal-paper copies of these editions are as tall as those of an earlier date, but
the outer margin of each leaf is greater. While for editions prior to 1535 the
maximum outer margin width of copies in our list is 102 mm, for those after that
date it is 134 mm, a feature which gives these copies a more normal appearance
(figure 4). At the same time, from the position of the type page it is clear that


Page 101
here too the royal-paper copies were printed from the same imposition of type
as that used for copies on smaller paper. But obviously the method of producing
these large copies had changed. What this method was was demonstrated more
than twenty years ago by the Belgian scholars Jean-François Gilmont and Émile
van Balberghe. Having examined the royal-paper copy of Paulus Manutius's
Commentarius on Cicero's letters Ad Atticum (1547) which once belonged to Ren-
ouard, they pointed out that the second of the two suggestions he had advanced
to explain how this copy had been produced—namely that it had been printed
on half-sheets of royal paper—could be demonstrated by a detail which had
escaped him. This was the presence of a pinhole towards the bottom of the in-
ner margin of ff. 1, 2, 3 and 4 of each gathering.[30] Their article was illustrated
by several diagrams, one of which is reproduced here, with their kind permis-
sion (figure 5). Superimposed on the image of an octavo forme intended for the
printing of chancery paper are the outlines of a half-sheet of royal paper, with
the pins represented by black spots. Comparing figures 3 and 5, one can see
that with the second system the outer margin of the inner leaves of royal-paper
copies (in this case ff. 1, 2, 3 and 4, the press having adopted in the meantime
the centrifugal octavo imposition scheme) could be as wide as the entire distance
between the pages of type in the forme, thus doubling their width with respect to
the earlier system. It would still have been necessary to trim the sides of the sheet
before printing, but not so drastically as in the earlier system. Pinholes are visible
towards the bottom of the inner margin of the first four leaves of each gather-
ing of the three royal-paper copies of Aldine octavos from this period which I
have examined in the British Library, thus confirming the general applicability
of Gilmont's and Van Balberghe's reconstruction. In copies of Aldine octavos of
this second period the presence of pinholes provides an additional clue to their
origin as royal-paper copies. Indeed, the holes can still be visible even when the
copy has been so heavily trimmed as not to be recognizable as a royal-paper
copy from its leaf height.[31] To form a gathering of a royal-paper octavo printed
on half-sheets in the way described, a method of folding is required commonly
used in Italy in the first years of printing for quartos in eights, in which the two
half-sheets are folded together, with the sheet containing f. 1 on the outside, and
with the first fold along the short axis, forming a gathering in which the two
conjugate leaves of the second half-sheet are positioned between, and not after,
the two conjugate leaves of the first.[32]

The date 1514, that of the first instance which I have been able to document
of an Aldine octavo with royal-paper copies, shows that the practice goes back
to the days of the great Aldus. In his Supplément aux listes chronologiques Renouard
reports (p. 484) that Samuel Butler, Bishop of Lichfield (1774–1839), had once
had "à sa disposition", but without apparently acquiring it, a copy "en grand
papier" of the Euripides of 1503. If this was a royal-paper copy, it would be the


Page 102

FIGURE 5. Half-sheet of royal, superimposed on outer forme of centrifugal octavo; Gilmont,
fig. VIII.

only instance known to me of such a copy of a Greek octavo.[33] More promising
is Renouard's reference to the appearance in a sale catalogue of a copy of the
1509 Horace "de cette forme allongée par le bas, dont on connoît plusieurs autres
éditions Aldines de ces mêmes temps".[34] Many of the Aldine octavos with royal-


Page 103
paper copies published between 1514 and 1541 also have vellum copies. The in-
formation given here on this point is surely incomplete: even in the course of the
relatively limited research which it was possible to undertake in the preparation of
this article a vellum copy came to light of the 1514 (1519?) Virgil (CUL SSS.39.8),
an edition of which, according to Renouard, "on ne connoît aucun exemplaire
sur vélin" (p. 69). One of the explanations offered for the Greek royal-paper folios
of 1525–1536, namely that they were provided with large margins for the use of
scholars, is much less plausible when applied to the royal-paper octavos. More
likely is that the latter were intended as special copies for discerning readers, less
expensive than vellum copies, but still in their way extra-ordinary. It is probably
no coincidence that the change in the shape of royal-paper Aldine octavos took
place at the time when Aldus's third son, Paulus, was assuming control of the
press, leaving his Torresani relations to operate as publishers and booksellers.
And this change may, in its turn, have been due to Paulus's intention to mark his
arrival at the helm of his father's press by the presentation to important person-
ages of beautifully bound copies of his earliest publications, beginning with the
octavo Pliny of 1535–1536 and continuing with the series of octavo editions of
Cicero's works edited by Paulus, which the press published in 1540 and 1541.
For these presentation copies he may well have wanted a book less "bizarre" in
appearance than the earlier Aldine royal-paper octavos.[35]

Even from these incomplete data it is possible to draw some provisional
conclusions concerning the production of royal-paper Aldine octavos. From the
beginning many octavo editions comprised more than one type of copy. For the
earliest editions we have reliable information only of copies on narrow-median
paper and on vellum.[36] But before the death of Aldus two other types of copies
appear, those on royal paper and those on blue paper. For the 1514 Arcadia ex-
amples of all four types are known to survive. The success, or at least the succès
, of the first group of octavo classics—a risky initiative, in which Aldus
had invested heavily—must have convinced the press to continue to offer readers
alternatives to the copies on narrow-median paper, but of more than one sort.
The positioning of the pinholes on leaves 1–4 of the second series of Aldine
royal-paper octavos suggests that the same tympan, and probably therefore the


Page 104
same press, was used for narrow-median and royal-paper copies; I assume that
this practice was also adopted for the earlier series.

Despite the likelihood of further instances coming to light of Aldine octavo
editions with copies on royal paper, it would seem, on the evidence at present
available, that not all Aldine octavos printed between 1501 and 1550 contained
such copies. Given the attention devoted for centuries by scholars, collectors and
librarians to the first group of octavo editions, those printed between 1501 and
1505, it is significant that no sure information is available about the existence of
royal-paper copies of these editions. Leaving aside the Sannazaro of 1514, in our
list the editions published before 1535 form two chronologically distinct groups,
those published between 1517 and 1522 (fourteen) and those published in 1533
and 1534 (nine). Yet in the first of these periods the press published a total of
forty-two octavo editions, and it published another thirteen between 1522 and
1528, a period for which we do not know of a single example of an octavo edition
with royal-paper copies. In 1533 and 1534 it published a total of twelve octavo
editions; perhaps the three editions of this period in which royal-paper copies
have not been found (Petrarch, 1533; Pontanus, Carmina, 1533; Poetae tres, 1534)
would provide a good starting-point for further work.[37] Between 1535 and 1550
the firm, adapting itself to prevailing trends in the Venetian publishing industry,
increased substantially its production of octavo editions of vernacular works, but
these, some forty in number, are poorly represented in our list.

At the moment it is not possible to say whether the practice of printing royal-
paper copies of editions of which the main print-run was on smaller paper was
yet another of Aldus's innovations. While there are many fifteenth-century edi-
tions with vellum as well as paper copies, none is known with copies printed on
paper of different sizes. The practice seems to have originated, at least as far
as Italy is concerned, in the early years of the sixteenth century. But there is
already evidence to show that it was not exclusive to the Aldine press. At some
time between 1504 and 1517 there appeared a quarto edition of Sannazaro's
Arcadia, based textually on the Neapolitan princeps of 1504. The edition has no
date, printer's name or place of publication, but on the basis of its types has been
attributed by F. J. Norton to Bernardino dei Vitali, a printer active from time to
time in Venice between 1495 and 1539, but who also printed in other centres,
including Rome. Norton's copy, now in Cambridge (CUL Norton c.66), with leaf
dimensions of 205 × 130 mm, consistent with its having been printed on chancery
paper, uses for the most part paper with two different versions of the balance wa-
termark.[38] But in the British Library there is another copy (BL 81.h.16; figure 6),
with leaf dimensions of 278 × 202 mm, and large outer and lower margins


Page 105

FIGURE 6. Jacopo Sannazaro, Arcadia, after 1504, BL 81.h.16, sig. B1r.


Page 106
(100 mm and 111 mm respectively), printed on paper most of which has the wa-
termark of a ladder in a circle; from its size this paper is clearly royal paper. The
position of the type page in the London copy shows that the type was imposed
for the chancery paper copies and the same imposition was used for royal-paper
copies, as in the Aldine editions. Evidence has also come to light recently of the
existence towards the end of our period of royal-paper copies of an octavo edition
not printed by the Aldine press. Dr Sara Centi kindly informs me that she has
found in the Biblioteca Comunale of San Gimignano a royal-paper copy of the
Carmina quinque illustrium poetarum, a chancery octavo edition published in Venice in
1548 by Vincenzo Valgrisi. The copy has leaf dimensions of 193 × 120 mm; it also
has pinholes in the lower margin of the first four leaves of each gathering, just as in
the royal-paper copies of Aldine octavos of the same period. It is likely that other
royal-paper copies of Italian sixteenth-century chancery editions, Aldine and oth-
ers, await discovery in the many Italian libraries with sixteenth-century holdings,
most of which are without published catalogues, and without specific information,
even within the library itself, about the leaf dimensions of individual copies.


Cracherode's books can be found in the British Library at pressmarks running
from 671 … to 688 …, and from C.17 … to C.24…. For Cracherode see Oxford DNB,
vol. 13, pp. 909–911 and in the online edition; for the three collections see Harris 1998,
pp. 29–32, 207.


See Gaskell, figures 50 and 51. For Boghardt's terminology see Hornschuch,
pp. 22–25. Boghardt objects to Gaskell's terminology because it begs the question of the
historical relationship between the two schemes. His own terms are purely descriptive. Scant
documentation is available on the use of octavo imposition schemes in sixteenth-century Italian
printing; what little there is suggests that in this period Gaskell's "inverted" scheme was no less
common than his "common" scheme; see Fahy 1994 and 1995. As far as sixteenth-century Ital-
ian printing is concerned, Gaskell is wrong in suggesting (p. 106, note C) that sheets printed on
formes using the "inverted" scheme were perfected by turning them on their longer axis.


This possibility was hinted at many years ago by Foxon, p. 173, who suggested that
these large copies of Aldine octavos "are printed from the normal imposition of type on a
square sheet of paper, giving very deep bottom margins." That the sheet of royal paper was
centred on the forme is shown by the position of its watermarks, fragments of which, when vis-
ible, are found in the leaves positioned on the outside of the formes. In the royal-paper copies
of the 1519 Horace the mispagination of the outer forme of gathering m, an error not present
in the narrow-median copies, suggests that the larger copies of this edition were run off before
the smaller ones.


See Renouard, pp. 140–141; Gilmont, pp. 49–54. The copy in question is now in the
Ahmanson-Murphy collection in California (entry 45 in our list).


See entry 38 in our list, where the leaf height is only 170 mm. I am grateful to Neil
Harris for examining this copy for me.


For quartos in eights folded in this way see Hellinga, p. 11; Zappella, p. 121, fig. 33.


I have not been able to investigate a copy of the 1517 edition of Oppian, De piscibus, in
Latin and Greek, in the library of the Earl of Leicester, Holkham Hall, Norfolk, which, as the
Earl's librarian, D. P. Mortlock, kindly informs me, has a binding height of 178 mm.


See Renouard, p. 57. The sale catalogue is Catalogo della biblioteca della ch. me. di
monsignore Natale Saliceti
, Rome, 1789. I am grateful to Neil Harris for transcribing the entry
therein concerning the 1509 Horace. The relevant part reads: "Exemplar nitidum in majori,
& solidiori charta excusum, formaque oblonga praeditum from which it will be seen that
Renouard's description is an interpretation rather than a translation of the original. An ele-
ment of doubt persists because, in comparison with the octavos of other printers, early Aldine
octavos on narow-median paper could also be described as "forma oblonga"; see Needham
1994, pp. 302–305.


For the bindings see Hobson, pp. 113–116, and Brooker, passim and ff. 1–15. They
were largely the work of the greatest binder active in Venice at that time, called by Hobson
the "Mendoza Binder" because of the numerous bindings carried out for Diego Hurtado de
Mendoza, Imperial Ambassador to Venice from 1539 to 1546, and tentatively identified by him
as Andrea di Lorenzo of Verona.


In need of checking are Renouard's comments on octavos from 1501 onwards printed
"sur papier fort" or "sur beau papier fort". The results of a survey of the paper used in a few
copies of the 1501 Martial, the 1502 Dante and the 1503 Greek Anthology—all editions which
Renouard claims contain copies of this sort (indeed, in the case of the Greek Anthology, was
printed in its entirety on this sort of paper)—cast doubt on the objectivity of these remarks:
Martial, three copies examined (BL C.4.f.2; C.19.b.4; CUL Sel.6.45), twenty-four sheets, all
of the same paper, unwatermarked, but with the countermark A, placed between two chains;
Dante, four copies examined (CUL CCE.1.1; Norton d.185; Young 253; Sel.6.20), thirty and
a half sheets, twenty-nine of which are watermarked with a balance inscribed in a circle, with
a cinquefoil countermark, plus one sheet of the same paper as that used in the Martial; Greek
Anthology, four copies examined (BL 672.a.1; C.4.C.8; TCC M.12.17; M.12.18), thirty-seven
sheets, all of the same paper as that used in the Martial.


In a letter to Paulus Manutius, dated 3 November 1535, the Benedictine monk On-
orato Fascitelli writes of a consignment of copies of the octavo edition of Sannazaro's Latin
works (September 1535) sent to him at Rome by the Aldine press. The consignment contained
a few copies which Fascitelli describes as being on median paper ("in carta mezzana"), much
sought after by their Roman friends (Lettere, pp. 348–349).


On Bernardino dei Vitali see Norton, pp. 62, 160–161; Ascarelli and Menato 1989,
p. 336. The edition is also described, on the basis of four copies in Italian libraries, in Villani,
p. 16; all are on paper with the balance watermark, like the Cambridge copy. Villani claims that
the types used are identical to those of the 1504 princeps; in fact, the two fonts are different, as
can be seen by the treatment of the group "Qu", rendered by a ligature in the princeps, and by
two distinct typographical characters in the anonymous edition.