U/.

',is«,"' - •• :
,:v, ' : Ì,
• .
•

^•.&?':;:.^l#ï^!;^'-f;''!i^i^;^:';.;^

Wis«;, -vi '',:.'

;?^^?;î^^I'^!'

•/:.;:

w-l'

m\;.':M':':.'

Univ. OF
Toronto
llBRARV
^mmiQ LisT^uiì
ISISÄ
.:%
 /

ST. ANDREWS UNIVERSITY

PUBLICATIONS, XIV.

PALAEOGRAPHIA LATINA

PART

EDITED BY
Professor W. M. LINDSAY

Published for St. Andrews University by

HUMPHREY MILFORD
OXFORD UNIVERSITY PRESS
London, Edinburgh, Glasgow, Copenhagen, New York,
Toronto, Melbourne, Cape Town, Bombay, Calcutta, Madras,
Shanghai
1922
FCINTED IN ITALY
 PREFACE.
Under the title '
Palaeographia Latina '
I propose to issue
a journal of Latin Palaeog^raph}-, particularl}' of Latin book-
script until the middle of the eleventh century. In this ven-

ture I have the help of Dr. M. R. James. Provost of Eton

College, and Dr. J.
Cunningham, Member of the St Andrews
University Court.
The journal will be cosmopolitan. Articles in French, Ita-

lian, German will be as welcome as those in English. Will

writers of articles and senders of books (or articles) for review
(or short notice) please direct them to the address : Prof. W.
M. Lindsa}', The Universit}', St Andrews, Scotland, and write
in the corner of the envelope the words '
Palaeographia La-
tina '.
I hope that two parts will appear each }'ear.

The publisher is Mr. Humphre\' Milford, Oxford Univer-

sity Press, Amen Corner, London (E. C. ; and the price of
each part is. for the present at least, five shillings.

The first part, printed in Rome, I dedicate to my dead

friend. Abbé Paul Liebaert. His teacher. Padre Ehrle, has
helped not merel}' by giving advice on its form and contents
but by piloting it through the press.

W. M LlND-SAY.
 /

CONTENTS OF PART I.

The Letters in Early Latin Minuscule (till c. 850) by W. M. Lindsay.

Some Early Scripts of the Corbie Scriptorium by the late P. Liebaert.

Plate I (Latin letter-forms).

Plate TI (the Corbie en-script) Paris, Bibl. Nat. lat. 13349,
fol. 8\
Plate III (the Corbie Leutchar-script) Petrograd, F. v. I, 6, fol. 1'.

Plate IV (the Corbie Maurdramnus-script) Amiens, 6, fol. 1\^.

Plate V (the Corbie Maurdramnus-script) Petrograd Q. v. I,

16, fol. 81\

 THE LETTERS IN EARLY LATIN MINUSCULE
(till c. 850).

BY W. M. LINDSAY.

Wattenbach's account of the developement of Latin script

was very useful in its day. Here I attempt to bring it (or
part of it) up to date. For the form of the letters I refer the
reader to the facsimiles in Sir E. M. Thompson's Introduction
to Greek and Latin Palaeography (a book which every reader
of this journal is sure to have) since the onl}' satisfactor}* plan
is shew the form actuall}- used by a scribe in a word.
to
Where this book failed me, the Keeper of the MSS. in the
British Museum came to my aid and gave me spare copies
of as many suitable Palaeographical Society plates as he could
lay his hands on. From these (only six, alas!) 1 have cut out
the words photographed in the appended Plate I (nos. 1-20).
Where neither of these two means was
fell back possible, I

on Wattenbach's plan and imitated the letter-forms with m}'

own hand (nos. 21 sqq.).
In writing this account I have sought to help (1) palaeo-
graphers, b}' suppl}ing suitable names under which this or
that form can be referred to (e. g. cursive Insular e ), and ' '

(2) Latin scholars, by shewing what letters and ligatures of

letters were most easih' mistaken by medieval transcribers.
B}- the symbol Introd.
'
is meant Sir E. M. Thompson's
'

book; by Ir. Min. and Wei. Scr.

' '
m}- two booklets (now
' ',

published by Mr. Milford) on Earl}- Irish Minuscule and on

Early Welsh Script by Not. Lat. my Notae Latinae. The
;
' '

others will be easily understood (e. g. Rev. Bibl. for the ' '

Revue des Bibliothèques).

 .

8 W. M. Lindsay

A.
1 In the minuscule script of our period we may distinguish
the '
uncial type (Plate I 21) from the
' '
half-uncial '
type (PI.
I 22). Both of these types have man}^ varieties, and both ap-
pear side b}' side in many scripts.
/ I. The '
Merovingian variety (PL I 23)
uncial '
type. A
attaches the loop to the shaft of the letter by a short stroke
(see Rev. Bibl. 24, 21). An Insular (i. e. Irish or Welsh or
A.nglosaxon) variety (PI. I 24) turns the loop into an isosceles
'

triangle (see '

Ir. Min. '
pi. 7) Half-wa}' between the '
uncial
and half-uncial types stands another Insular form, the high-
' ' '

backed a, resembling a of the earliest cursive (until the 4.*''

'

or 5.^^ centuries; see Introd. facs. 101, 104 and pp. 336,337),
' '

and for us — as for medieval transcribers liable to confu- —

sion with (^ (cf. d at the end of the 6.*'' last line of facs. 97).
A mistake like dzco for avo should suggest an Insular exemplar
\ with '
high-backed '
a (cf. Zeitschr. Celt. Phil. 9, 306).
II. The '
half-uncial '
type. The circle may be closed or
' open (at the top). Of the closed form the onl}^ variety which
calls for notice is that resembling the ligature oc and productive
of mistakes like doctus for datas. It was perhaps in the Be-

neventan minuscule of the subsequent period that the danger

of this mistake was most acute.
2. The open form demands a fuller treatment. It is the

only form recognized b}^ Visigothic minuscule, in which / (see

below, s. V.) had in our period a shape like the closed form
of a. And it is a characteristic feature of all early minuscule.
The rough-and-ready rule for dating Caroline minuscule is well
known — that a predominance of open a indicates a date ' '

not later than the very open and fairly large

9.*'' century. A
^z, with its two tapering horns, is one of the outstanding fea-

tures of an earl}- t3pe of Corbie minuscule which I call in

these pages the Corbie en-t3^pe The two curves become
' '.

angles in a Corbie MS. of « 7 cent- » at Petrograd (Q. I 14

Gregory's Homilies), and this angular a is the characteristic ' '

feature of the Laon az-script and gives the script its name
(' Introd. '
facs. 125; cf. Rev. Bibl. 24, 15 sqq.). The two curves
get knobs at the top in other (') script, so that a resembles
(*) There is no ground for calling this a which resembles cc '
Mero-
vingian '.
 The Letters in Early Latin Minuscule 9

cc and a mistake like sue ceduo for sua edito becomes possible.
In the Corbie ab-script only the second curve gets a knob,
and a resembles ic [datus resembles die tus see Introd. facs. ;
'
'

s^28 and Rev. Bibl. 22, 405 sqq).

3. All these modifications are precautions against the
danger of confusing "
open
This confusion in a '

a with u.

transcript hardly gives a satisfactory' clue to the exemplar.

For all minuscule scripts in their early stage use open ^7, " '

and often the only real dirterence between '

open '
a and 7t

is that the former ends in a (slighth') upward curve while the

latter does not (see below, s. v. '
U' '). Perhaps the danger
was greatest for transcribers of a Visigothic or Insular mi-
nuscule exemplar; although a MS. in which it might be called
an inevitable pitfall for anv transcriber, Carlsruhe Reich. 57
Isidore's Etymologies, seems lo have been written at X'erona
(in a unique type of script) in the 8.*'' centur\' (see Holder in
'
Mélanges Châtelain "

). The danger is often increased b\' the

presence of another of the many (^z-types in a neighbouring
syllable. I noticed in a Milan MS. (B 159 sup.), written at
Bobbio c.words sepulehrum parare (in an apparently
750, the
contemporar}' minuscule entry on fol. ISS') written with the
'
open type in the first and the uncial in the second s\'l-
'
' '

lable of parare. Nine out of ten transcribers would write

pzcrare^ since the open a oi pa bears far closer resemblance
to the n of sepulehrum than to the a of ra.
Cursive script of all parts of the Continent loved to
4.

write a (in open form) above the line and connected with the
following letter. Sometimes the a stands directl\- above the
following letter (see the ligature ag in the last two lines of
'
Introd. facs. 114; it might be mistaken 'or ga).
'
More often
it stands high to the left or (e. g. the ligature ad in line 2 of

'
Introd. "
facs. 115; the ligature at of the second last line of
'
Introd. '
facs. 131) fairly high. This cursive feature disap-
pears from most book-hands about (*) the end of the eighth
century. But not from \'isigothic (e. g. in Madrid Acad. Hist.
25 Isidore's Et\"mologies. of. 946 A. D., at; in Madrid Toi.
10. 25, of ^-02 A. Ü., at, am, Escurial a I 13, foil. 1-187,
ar-, in
of 912 A. D., sometimes as. etc.). The resemblanc: of an so
(^) In a Flavigny MS. of 816, Leyden Seal. 28, haec often shews it. In
Oxford Bodl. 849, of the year 818, a.jr of •
pax '.
10 W. M. Lindsay

written to a common form of Zis mentioned below ^s. v. '

Z ').

The suprascript a of e. g. a Rheims MS., Le}'den Voss. Q. 60

(" 8-9 cent. "). of small size and angular (instead of curv^ed)
perhaps deserves mention. Its at ^with the branch of t ' "

wanting) might also be mistaken for Z.

5. But Insular scribes save space by writing a un open (*)

form) not above but below the line and connected with the
preceding (not the following) letter, e. g- na (see PI. I 1 7ia- '

sonis '). This subscript a is not uncommon in the Conti-

'
'

nental minuscule of centres like Murbach (e. g. Gotha I 85

Canones Murbacenses, with ma. 7ia), Cambrai (e. g. the half-
uncial Cambrai 441 has it in »la on fol. 76^). etc., where In-
sular script was also practised. '
Suprascript '
a in the Insular
minuscule of our period I have in the Book of found only
Armagh (along with subscript a), where it generally shews
' '

the closed form, but occasionali}- (see Introd. facs. 13", ad ' "

in the 5.*" last line of col. ii) the open.

6. Of the earh" histor\- of the ligature ae a few statistics
'

have been published (b)' U. Robert in Mélanges Jul. Havet '

1895). but onlv a few. The different methods of expression

mav be classified as follows, although the boundary-lines be-
tween this and that varietv are occasionally effaced in practice.

1. with both elements equally preserved. L^ncial ae be-

comes as in PI. I 25, and half-uncial becomes as in PI. I 26,
Both, but especialh- the first, are confusible with a ie ligature

(§ 115).
2. ^vith the e reduced, the earl\' treatment of e in ligatures
(§ 29). For example, in the Bamberg Gennadius we find the
form of PI. I 27.
3. with the a reduced, either (1) as in PI. 1 1 "
^z^quora ',

or (2) as in PI. I 28.

4. with a appended in the form of a mere cedilla or loop
to c ; either (from the second variety of no. 3) as in PI. I 29,
or (from the Merovingian form of <z,- Ì5 1) PI. I 30 (as in the
Laon az-script), or (from the first variety of no. 3?) PI. I 31,
or else (with a conventional mark) PI. I 32.
The fourth method is found as early as the half-uncial
Basilican Hilary of the >'ear 509. The first variety of no. 3
(') Seldom closed (e. g. '
Ir. Min. '
pi. v., nia in the 1 }^ last line of
col. i and the 10, "> last of col. ii).
 The Letters in Early Latin Minuscule 11

is affected in the Irish half-uncial of the Cathach (written pro-

bably b\' Saint Colnmba's own hand) the Book of Durrow, the
Garland of Howth. For the Merovingian variety of no. 4 "
"

may be cited the uncial Treves Gospels, written in 719; for

the third variety of no. 4, the Salaberg'a Psalter: for the first,

the Maihingen Gospels (written probabl\- at Echternach about

'
740), and so on. The cedilla will be seen at the end of' Introd.
facs. 129 '
superbi^i^ ".
The Moore Bede (written apparenti}-
at Le Mans, c. 737) has both the second variety of no. 3 and
the third of no. 4 (see "
Introd. '
facs. 142, line 10). The long
loop appendage (seen in facs. 130, line 9) appears often in the
Book of Kells.
The sometimes identical with the second va-
oe ligature,
riet\- of no. even rarer in the minuscule of our period
3. is

than is the diphthong oe in the Latin vocabular\-.

B.

7. In minuscule we form (like our b)

find only the half-uncial
not the uncial (like our B) although there are one or two :

examples of the latter: e. g. Oxford lat. theol. d. 3 (of unknown

provenance) and Cassel th. O. 24 (written at Fulda when Ser-
vatus Lupus was a monk there) offer it occasionalh. Some
scribes of Anglosaxon half-uncial add a downward curve at the
top of the Ò (on the right), e. g. one scribe of the Lindisfarne
Gospels (also in Durham A II 17 and the Salaberga Psalter).
But since the same addition appears at the top of / and h (in
the Salaberga Psalter), it should not be called a trace of
the upper round of uncial B. No more should the branch
which shoots out from the middle of the shaft in the cur-
sive of the Frankish Empire. It is rather a mere connec-

ting link between ò and the following letter. Sometimes it

is arched: and this is the usual form when ò is connected

'

with a following letter in the Luxeuil t3-pe (see Introd. '

facs. 124, line 6 bet '). More often it is a straight horizontal

'

stroke, and ò with this appendage is a characteristic feature

of the Corbie ab-script (see Introd. facs. 128. line 2 bi '. ' ' '

'
bo line 4 bu ') as well as of the N. E. France minuscule
'. ' " "

of Autun 20, Montpellier Ville 3, and is common in the Me-

rovingian bookhand of London Harl 5041. Berne 611, St Gall
214. The Corbie ab-scribes often leave an interval between
 12 W. iM. Lindsay

the main part of the lower curve of the letter and the dot
or knob which this curve should end. A gappy h of this
in ' '

kind might sometimes be mistaken for t.

8. Open b (with the round at the base of the letter not
' '

continued back as far as the shaft) is a cursive usage which

was not expelled from bookhand for some time- Of Bene-
ventan minuscule we are told (Loew Ben. Scr. p. 134) that ' '

b is more often open than closed in the ninth centurw For

other minuscule, statistics have not vet been collected. Cer-
tainly the closed form is no sure criterion of a late date in
our period, for (to mention onl}- one piece of evidence) the
'
ancient minuscule of the marginalia of the Bologna Lactan-
'

tius shews both the closed and the open forms of the letter.

/ 9. The shaft of b (as of /) was straight in half-uncial. In

Insular half-uncial however the shafts of b and / bulge out to
the left (see '
Introd. '
facs. 135-136, 140-141); and this Insular
peculiarity (which often found its wa\' into minuscule) appears
V on the other side of the Channel and characterizes what De-
lisle(in his L'Evangéliaire de Saint Vaast d'Arras ') called
'

'.
the '
calligraphie franco-saxonne
/ 10. In cursive the scribe often began the letter b, or any
long-shafted letter, with an upward hair-line on the left, so
that the shaft of b (as of /, d, h) resembles a whip (with the
thong on the This cursive feature appears even in
left side).

^^the Irish half-uncial of the Ussher Gospels (see Introd. facs. ' '

134, line 3 Ii, line 4 /) and the Veronese half-uncial of Verona

59 Vigilius Tapsensis, etc. It was not expelled from minus-
cule until the latter part of our period (e. g. Stuttgart H. B.
VII 39, written at Constance 811-839, and Paris 2123, of 795-816,
offer examples of the whip '-shaft). '

^
11. Another cursive freak was to break the continuitx' ot
the shaft of b (and of /) near the base and make a sudden
turn to the left. The letter b so written seems to end in a
hook. This hooked b was greatly favoured in the Corbie
' '

\ ab-type (see Introd. facs. 128) and (with a similar /) in the

' '

Merovingian bookhand of London Harl. 24143, Berne 611.

It often appears (usually with a similar /) in the earlier mi-

nuscule of our period, e. g. Epinal 68 Jerome's Letters (of 744) ;

Einsiedeln 27, foil. 1-24 Ascetica ("8 cent. "); Cassel th. O. 10
Augustine, etc. (Fulda, 8 cent. "). ''
 The Letters in Early Latin Minuscule 13

12. But the most violent contortion in cursive script o(\

the regular form of ó and / is that extraordinär}' sinuous type .

of shaft seen in '

Introd. '
facs. 126, line 6 a./ìaì The sinuo-
' '.

sit}' is not ver}' pronounced in this specimen taken from Lon-

don Harl. 5041 (Merovingian), but we find \eritable '
cork-
screw patterns on other pages of this MS.
'

13. In calligraphic minuscule a short horizontal stroke so-

metimes caps the shaft of (also d, /, //), especiall}' in one fi

species (e. g. Cambrai 624) of that class designated b}' Traube

(' Palaeographische Bemerkungen p. 1) between Gallic half- ', '

uncial and minuscule and assigned to the 7.^^ and S.*"" cen-
',

turies.

14. In cursive script c was usually higher (often much

higher) than ordinar}- letters. A '
high '
(or highish) c is not\
at all uncommon in the minuscule book-hands of especiall}' the
earl}' part of our period (Munich 14437, of the year 823, is a
fairl}' late example from Continental script; for a later from
Insular see '
Introd. '
facs. 138). Some scribes emplo}' it only
(or b}- preference when c is connected with a following letter '

^especially o). The temptation of later transcribers to substi- \

tute e may be
appreciated from a glance at the e-ligatures in, ^^
e. g., facs. 139 of Introd. In a St Bertin MS., St Omer 15
' '

Hieron}-mi Breviarium, high c ends at the top in a knob' '

turned to the left, and the same antenna type (commoner ' '

in e; q. v.) appears in Laon 68.

15. The lower part of c (as of e, cursive script \
etc.) was in
often separated from the upper. This fashion of writing c with
two strokes of the pen (see Introd. facs. 131, line 8; facs. 134, ' '

line 1) produced in man}' types a knob-projection at the back

of the letter; see Introd. facs. 128 for examples from the
' '

Corbie ab-type, in which this formation is normal; facs. 124,

line 8 for an example from the Luxeuil type; facs. 125, line 8 .

and line 9 co for examples from the Laon az-type.

' '

*
16. In other types the upper part of high c took the ' '

form of another c (as in PI. I 3 unicum 4 sacrifi^ia '). This ' ', '

variety is usuali}' called broken r, although it seems better

' '

to assign this general name to any c which is written with

two separate strokes of the pen, and to call this special va-
 14 W. M. Lindsay

riet\'double e. The double

' '
a feature of Italian script,
' '
^ is

first minuscule cultivated

of that Italy before the su- in North
premacy of Caroline minuscule (§ 64), then of its successor
the Beneventan script (of South Italy). But it is found else-
where too in the eighth centur}- (e. g. Cambrai 619 Ca nones
Hibernenses, written at Cambrai 763-790; Ep in al 68 Jeromes
Letters, of 744) At St Gall it persist till after our period (e. g.
;

St Gall 46, of 872-883).

^ 17. Of
former universal prevalence a trace has been
its

left in the ligature ct which, we ma\- sav, ever3-where in Con-

tinental script exhibits the double c form icf Introd. facs. '
'
' '

^152, line 5 docti '); although divergences are not unknown,

'

the most notable being a Bobbio type recumbent t (§ 108) ' '

which ma}' sometimes be mistaken for x or ox or ex (PI I

33), e- g. Vat. lat. 5763, Wolfenbüttel Weissenburg. 64, Mo-
dena O I 17, all three MSS. of Isidore's Etymologies, written
at Bobbio in the beginning or middle of the 8.*'' centurv.

^ Of other ligatures of ^ with a following letter the com-

18.

monest is CO (PI. I 34), found in the Corbie ab-script and in

s^ some 8.'*" cent. MSS.. e. g. Fulda D. 1 Codex Theodosianus

(written at Constance\ Verona 33 Augustinus de Agone Chris-
tiano (written at Verona). In Luxemburg 68 Echternach, " 8-9

cent. ") it appears in the cursive entry contuli but not in the ^

text. The letter becomes a mere circle (as does / in Spanish

ligatures; see below, s. v.) in the rare ligature cc (PI. I 3vî),

found in P^pinal 68 (written at Murbach in 744), e. g. o^^asio \ '

in Milan L 99 sup. (written at Bobbio, perhaps about the same

^. time), in the Milan Maximus (along with a similar ce, etc.;

V '
Introd. "
facs. 114, line 6), etc. On the ligature ec see § 111-

D.
19. The uncial type (PI. I ?ß) is as common as the half-
uncial (like our d) in most of the minuscule of our period and
in some scripts (e. g. Beneventan) evt^n commoner. It was a
convenient way of writing the letter without lifting the pen,
and was quite at home in cursive script. The marginalia in
'
ancient minuscule ',
e. g. of the Bembine Terence, use it freely
(in '
open '
form) along with a disjointed variet}^ of the half-
'

uncial letter (PI. I 37), and this pair (uncial and '
disjointed
 The Letters in Early Latin .Minuscule 15

half-uncial d) appear in several early Insular MSS. of Bobbio

(e. g. the Naples Charisius), Fulda (e. g. Cassel theol. F 22), etc.
20. Insular scribes who used the '
high-backed '
variet}- of
\
a (see ^ would find dangerously similar the half-uncial d
1) ' '

in its regular shape (*). A shape which (partially) escaped the

danger was one appropriate to d when connected with a pre-
ceding letter, a very open d with the left side of the round
bent backward at the top (' Introd. facs. 12v5, line 4 od; facs. '

128, line 3 ed). This shape was found convenient by Insular

'

scribes, and in their hands often suffers that peculiari}- Insular '

modification, the insertion of an angle at the middle part of

the round. (On a similar treatment of the letter q see below,
s. v.). facs. 137, taken from the Book of Armagh,
'
Introd. '

'

shews uncial d emplo}'ed throughout, with the ligature

' ' '

form of half-uncial d used occasionali}-, not merely in the

' '

ligature ed (col. i, line 1) but also independently (col. ii, line 4,

line 9). In the word '
audierunt '
(col. i, line 6) the a has the
'
high-backed '
form, the d (as usual) the '
uncial '
form. In
'

facs. page from the Ley den Priscian, only the uncial
138, a '

form appears, although in other pages of this MS. the half- '

uncial especially the 'open' variet}) is not unknown.

'

21. Open d (like open ò, /, g) is a test of the age of

*
' ' "

Latin MSS., but, until statistics have been collected, cautiorr,'

is necessar}' in appl}ing d appears (often along
it. '
Close '

with '
open ') in many MSS. which undoubtedly belong to the
S.**" e. g. Wolfenbüttel Weissenburg 81 (of the year
century,
772), Cologne 98 and 9L Vat. Pal. 187, London Add. 18304,
and is normal (except when the letter is connected with a
preceding letter) in the Laon az-t}pe. the Corbie ab-type, etc.
22. A better criterion for distinguishing the older MSS.^
of our period is the projection of the shaft of half-uncial d ' '

below the line. We find it in Merovingian book-hand (and,

"
of course, cursive), in the Laon az-type. the '" N. E. France
minuscule, the Corbie ab-type, and other pre-Caroline scripts^
evolved from cursive, e. g. Epinal 68 (of 744 Cambrai 619 .

(of 763-790), sometimes in the North Italian minuscule of Milan

Trivulz. 688 (written at Novara before 800), etc. It ma\- be

(*) In the Bobbio Irish minuscule of Florence Ashb. 60, in which the
regular shape of d is used, the reader finds himself continually confusing,
e. g., quid and quia.
 16 W. M. Lindsay

/ taken as evidence that a MS. is not later than the 8.'^ centur}-
(if we except its traditional survival in the Corbie ab-t} pe to

\ the earl}' part of the 9.*^).

23. '
Open '
d
half-uncial form) was, when the
(of the ' '

curve terminated a knob at the top, danoferoush' like c/.

in
A cursive variet}' (like Greek minuscule Delta) of the uncial ^
'

form was often dangerous!)' like pre-Caroline a. It had a slight

curve to the right at the top (see Introd. p. 337, the Amherst ' '

>^ Papyrus d. This 'Delta' variety, so frequent in the cursive

of all countries (e. g. in the subscriptions in the Basilican Hi-
finds its way into book-hand, e.
lar\'), g. occasionalh' in Co-
logne 41 (of Hildebald's time), in the Bangor Antiphonar}-
(Irish half-uncial of 680-691), the group of MSS. at Munich
written for Princess Kisyla (Charlemagne's sister), the rude
uncial of St Gall 912. When the slight curve is prolonged
in a hair-line down half the length of the letter or more (as
in the cursive Irish script of the Book of Dimma), the resem-
blance to cursive o ceases and the letter approximates to that
variet}' already- cited from the Bem]:)ine Terence marginalia.
For other modifications of the shaft of d see above, § 10, ^ 13.

E.

24. lanciai e (PI. I 38) mav be compared

to an open mouth
with a tongue. our e) makes the upper
Half-uncial e like
half resemble an archer's bow, but half-uncial scribes often
retained the uncial form (made taller than ordinary letters)
'

when e was connected with a following letter (see Introd. *

facs. 98, line 4 et er, line 5 em, line 7 er, line 8 ex, line 9 ep,
line 10 e_^, line 13 ef). This tallish uncial e in ligatures (the ' '

tongue of the e making the first part of the following letter)

is a feature of most minuscule scripts of our period. Perhaps
it catches the eye more in an early t}'pe of Corbie minuscule

than in any other. The sloping tallish e (ending at the top

in a tag like a beetle's head) in ligatures like ew, the very
open a with tapering horns), the maiuscule 7i (like our N)
are the three characteristic features of this script, which 1 style
\ the '
Corbie en-type '.

Ç 25. To the '

beetle-head '
variet}', so dear to the early
scribes of Corbie, some other scribes (e. g. the Cambrai scribes
of Cambrai 619, the Treves scribes of Munich 28118) prefer
 The Letters in Early Latin Minuscule 17

an antenna variety, with the top curved backwards into a

' '

knob, and (like the antenna shape o( c ^ 14) the letter is

'
"
,-

often sloped upwards so as to tower above the other letters.

The hiyh antenna e is affected by the Micy scriptorium,
'
'

bat is found in other quarters too (e. g. London Add. 18332,

a MS. of Carinthia). i\.ll scribes, when the '
bow '
of '
half-
uncial e was accidentally
'
with ink, were prone to pre- filled

serve the identit}' of the letter by means of a long (or lontrish)

stroke sloping upwards. But even when the bow remains '
'

intactwe find this excrescence occasionally (e. g. it is a feature

of a Rheims MS., Leyden 114), sometimes ending in a knob
to the left, e. g. in Le3^den Seal. 28 (written at Flavigny),
sometimes in a knob to the right, e. g. in Boulogne 66 Aug-
ustine contra Cresconium (written at St Bertin).
26. These elongations of e are an inheritance from cursive
script, in which e (like ^; i^ 14) was often exaggerated in
'

height and length. Cursive often substitutes for the bow '

of half-uncial e a long upward-sloping narrow loop, and perhaps

the cursive of North Itah' ma}' be singled out for this pecu-
liarity: Examples
be seen in Introd. facs. 114, where
will * '

^
however the length and narrowness of the loop are not so
marked as in other early specimens of the Bobbio scriptorium
(e. g. the Naples Gesta Pontificum and Wolfenbüttel Weissen-

burg. 64).
27. The
loop was the result of writing elongated uncial '
'

e without lifting the pen. The pen forms the loop by coming
down from the top of the mouth of the e to the point where ' '

the tongue has to be made. Another one-stroke method

' '

was Beginning at the right end of the upper curve of

this.

the C (the open mouth ') the pen nearl}- completes the C,
'

but when it reaches the lower curve it makes that curve turn
to the left, instead of to the right, and finishes its course b\'
inserting the tongue (PI. I 39). This '
eight figure '
e, although
found elsewhere (e. g. in the early cursive of Lombardy), may
be called par excellence the Insular cursive form. From In-
sular cursive it found its wa}- into Insular bookhand and was
sometimes written conventionall}- like a long sloping narrow
/ see PI. I 5 '
versus '). The '
conventional shape is a cha- '

racteristic (with a long and narrow ,^ and a ver}- open q,-

§ 87) of an early type of x^nglosaxon minuscule (e. g. Petrograd

2
 IS W. M. Lindsay

Q. I 15), The '

figure eight '
shape lis much more common
(see Zentr. Biblw. 1908, p. 290; Zeitschr. celt. Phil.
9, 306), and

appears even presumably an Insular importation) in the Con-

tinental script of such MSS. as Cologne 40 (written at Cologne),
Paris nouv. acq. 1575 Eugippius (written at Tours). When
the two loops of the 8 are, as frequenti}' happens, left incom-
plete it might be mistaken for .<r.

28. The reserve process was also possible. The pen

o
might begin at the right end of the lower curve of the C and
make the upper curve turn to the left (PI. I 40). This is a
favourite method with Continental scribes in such ligatures as
ei (e. g. in London Cotton Nero A ii, written at Verona), eo
(e. g. in Berne 611, of a Merovingian type), and, above all, ^/.

On fol. 47'" of the Cologne MS. just mentioned, Cologne 40,

'
et retro '
has the first e written in this Continental cursive
fashion, the second written in the Insular cursive fashion. An
earl}' example of ex with this be found in the
sort of e will
'
ancient minuscule '
marginalia of the Vienna Hilar\-. But
this '
Continental cursive '

e is also known in Insular script,

e. g. in the Book of Mulling eCy ei, eo, ex.
29. In cursive script e is often broken into two parts. In
the word '
gestarum '
in '
Introd. . facs. 114, line 1, the lower
part is united with the head of the ,^ and the upper part (the
'
bow ") is added separately
line 3 etc.). (cf. pra^, line 5 patr^m,
On a treatment of c see above, § 16. This fashion
similar
of writing e (and c) with two strokes of the pen produced in
man}' types a knob-projection at the back of the letter, as
will be noticed (both in e and in c) in Introd. facs. 128 (the ' "

Corbie ab-type)- This projection is more exaggerated in an

S.'** century MS. from Fulda library, Cassel theol. O. 5, than
in any MS. which I have seen. The e of et in this MS. sug- ' '

gests at the first glance the letter R, the ^ of es suggests B. ' '

30. The term broken c we proposed to extend to any

' '

minuscule c written with separate strokes of the pen, instead

of restricting it (as is usually done) to a particular variet}'
re-sembling one c standing on another. There is a form of e
analogous to this double c. It is the e so characteristic of
' '

Beneventan minuscule (see Introd. facs. 120) and of its pre- ' '

decessor, the North Italian, and is the favourite formation with

many scribes.
 The Letters in Early Latin Minuscule 19

V
Beneventan and North Italian the two parts of the
31. In
letter are equal in size, and the letter is not so tall as the
analogous c. Elsewere the loop set on the top of the c-like
"

substructure is often greatly reduced in size (see the ^ of et '

in line 3 of Introd. facs. 123 for a not very pronounced spe-

" '

cimen). Again Insular scribes (^and others) love to expand

the upper half at the expense of the lower in such ligatures
as C7i (PI. I there is hardly a letter of
41), ei (PI. I 42). In fact
the alphabet (see »5 110) capable of being written in so many
forms and with so varied a manipulation of the pen. On the
ligatures ae. oe^ et, eo, ex see § 6, 110. Ill, 133.
32. Of the rest the leaf-shaped ep ligature is the most^
worthv of mention (see' Introd. "
fase- 114. line 6). though con-
fined to the ancient minuscule of the marginalia of the Bo-
' '

logna Lactantius and of Paris 12214 Petrograd Q I 4. etc., —

and to the cursive types of the Paris Avitus. the Milan Hege-
sippus, Lucca 490; etc., the Corbie en-t}pe of Paris 4403'^ Co-
dex Theodosianus (also in the Corbie ab-type of Donaueschin-
gen 18, e. g. fol. 159""). Initial O is sometimes written in this
"
leaf '
form.

33. We ma\- convenienth',

if not quite correcth. distinguish

(1) the *
our F, but with the upper branch
half-uncial '
(like
arched), (2 the minuscule (like our f, but ver\- rarely stand-
'
'

ing on the line), (3) the looped (PI. I 43) varieties. The ' '

'
minuscule is a rounded and narrowed half-uncial the loo-
' * ' '

ped '
a '
minuscule "
written without lifting the pen. Perhaps
the normal position of all three (or at least of the first) is with
the (lower) branch resting on the line, and the normal posi-
tion ot .c But even where s is
often coincides with that of /'.

allowed to stand on the

is nearh' always projected line, f
below the line more or less. Some scribes, e. g. Pinguine
in the St Gall Priscian, actually place the (lower) branch itself
below the line. The looped is a cursive variety (e. g. in
' ' '^"

the marginalia of the Codex Claromontanus, in Lucca 490

sometimes), and in the book-hand of our period is usually
confined to ligatures (e. g. fr and fu in Zürich Stadtbibl. C
12). Often the loop does not quite reach the shaft of the
letter (e. g. /^ in Cassel theol. Q 10 and London Harl. 5041,
 20 W. M. Lindsay

fi in in Liege 306), occasionali}- resembling an

Epinal 68, fr
open p (e. g. fe. fn in Milan C 105 inf.).
'
'
(^)
/^ 34. Another cursive variety is split-backed /" (see In- ' ' '

trod facs. 114, line 2 for a not ver}- pronounced example),

'

found (along with a similari}- formed s) in the earlier or less

calligraphic Insular script of our period, e. g. the Book of
Mulling, the Book of Dimma, Oxford Digb}- 63 (of c. 850;
'

see '
Introd. "
facs. 144). In the Laon az-t\pe (see '
Introd.
facs. 125) the right side of the split is fantasticali}- curved
backwards; in and (facs. 124) are
the Luxeuil type both / jr

often similari}- treated. The crooked stem of minuscule / • '

in many MSS. of our period ma\- be a relic of this Mero-

j. vingian affectation.
^ 35. The branch of /' was a convenient connecting4ine with
a following letter. When the letter was /, dropped / was ' '

convenienti}- suspended to the tip of the branch, thus dropp-

ing below the line if the branch rested on the line. When
the branch and appendage were carelessly written together
(with a curve instead of an angle), h might be mistaken for
/. had another wa}- of writing /f, by means
Insular scribes
of the cursive Insular form of / or rather longa a long si- '
i
',

nuous stroke. This stroke straggles down through the top oi

the / as far as the line or below: the branch of the / is
omitted (see Zeitschr. celt. Phil. 9, 304 sqq. for details). In a
few earl}- specimens of Insular script written on the Continent
(e. g. the Irish script of \'ienna 16 and the Naples Charisius.

both written at Bobbio; the Anglosaxon script of London

Egerton 2831, written at Tours) we find sometimes a third
expression, the looped form of f with the / appended (as
• '

a curve) to the lower end of the loop, the point where the
loop touches the shaft. This third expression was discarded
b}- Insular scribes. It was too dangerousl}' like a common
Insular expression of si (see Introd. facs. 138, line 7 dioni.^a).
' '

! In the phrase figurain simul on fol. 4r of Vienna 16 (or in

such a word as significattcin in the Naples Charisius) the only
difference between ti and si is that the loop touches the shaft
in the first combination but not in the second. In an earl}-

(') When the foUawing letter abutted on '

half-uncial '
/, a close p might
be suggested and. e. g., fa miscopied as pa.
 The Letters in Early Latin Minuscule 21

specimen of the Anglosaxon script of St Bertin. Boulogne

63-64, the difference is effaced, for the scribe often writes fi
(when he uses the looped /) in ' '
the same careless way as
he writes fr. without taking pains to bring the lower end of
the loop as far as the shaft. Similari}- in the Anglosaxon
half-uncial of an Echternach (?) MS., Paris 9382 occasionalK-
(e. g. magnificentia fol. 82""), and in the Anglosaxon minuscule

of an Echternach MS.. Paris 9527 (e. g. inter /fciant fol. 66')

36. Continental scribes (not Visigothic, who do not use
'
looped /) indulged themselves with this free-and-easy way
'

of expressing fi (the cursive yinit of the uncial Corbie Go- " '

spels, fol. 7^ is an early example) and abstained from writing

si in combination. Therefore a Continental scribe can express
/z preciseh' as an Insular scribe express si^ and confusion of

words like sinit and unit would be inevitable in the transcript-

ion of Continental script b}' Insular copyists or of Insular
script b\' Continental. I have not found in an\' home insular
MS. this combination put to the Continental use of express-
ing fi. The instances abready cited from the Insular script of
Continental centres seem to be alien intrusions.
37. The f {s. etc.) of the Maurdramnus script of Corbie
has a knob-like protuberance at the back of the letter. We^
may call knob-backed f (PI. I 44). In Cassel theol. O. 5
it " '

there is sometimes actuall}- a small loop at the back, so that

fa (fol. 67') might even be mistaken for qua (with suprascript
arch for 7t ,-
see below, s.v. '
U ').

G
38. The and (Continental) minuscule
uncial, half-uncial
t}'pes bear a rude resemblance to the Arabic numerals 9, 5,
3, except that the circle of the 9 is open on the right, the
top-stroke of the 5 is prolonged to the left, and at the top of
the 3 there is a short stroke to the right. Spanish minuscule
adopted the uncial (*), Insular minuscule the half-uncial (^) form.
An earlier uncial type, merel}- c with a tag added to the lower
(^j Spanish scribes often express z by the half-uncial g-form, e. g. the
Toledo Bible (Madrid Tel. 2, 1).

(^) The uncial type I found occasionally in the Anglosaxon minuscule

of Gotha I 75 Sedulius. In Insular half-uncial (e. g. the Salaberga Psalter,
the Lindisfarne Gospels, the Douce Primasius) it is often found.
 22 W. M. Lindsay

end of the curve, was liable to be confused with ci (when

written with dropped i)\ so the tag- was lengthened to
'
'

N. prevent the confusion. reading MSS. like Milan L 99 sup.

In
Isidore's Et3^mologies (Bobbio, ''
mid. 8 cent. "), Le\'den Voss.
Q 63 Gregor}- of Tours (^^8 cent."). Vati Pal. 554, foil. 5-12
(Anglosaxon, 8 cent "), one finds some difficulty at first in
'"

distinguishing ci from g (e. g. adficiat from adfigat). And

early MSS. of Bobbio use an abbreviation of cjim, the letter
c followed b}'^ a sinuous vertical stroke (cutting the tail of the

c) which might easily be miscopied as g (see Not. Lat. p. 41). ' '

So might a rarer Insular cum S3'mbol, c followed b}' a 7-

' '

mark (Ibid. p. 360).

"O 39. Some specimens of Continental minuscule use
early
all three t3'pes2^ Berlin th. F. 354 Gregory's Moralia (written

apparently at Corbie, mid. 8 cent. ") the Maurdramnus Bible

''
;

at Amiens (written at Corbie, 772-780) ; Paris nouv. acq. 1575

Eugippius (written at Tours, " beg. 8 cent. "); Oxford lat. theol.
d 3 Commentar)' on Pentateuch (of unknown provenance) St ;

Gall 70 Pauline Epistles (written at St Gall, Verona c. 760);

42 Gregory's Pastoralis Cura (written at Verona) Oxford Jun. ;

25, foil. .1-59 Ethici Cosmographia (written at Murbach, ''8

cent. "); Berne 611 -|- Paris 10756 (written, in part at least, be-
fore 721), and so on. All three appear in the cursive margi-
nalia of the Codex Claromontanus (Paris g^rec. 107).
40. The use of two out of the three is extremely commonj
Indeed we may say that only Visigothic and Insular minuscule
'
are restricted to one form (also Beneventan, although the
half-uncialform appears in the Bamberg Cassiodorus). The
freedom of choice does not surprise us since we find the un-
cial form encroaching on half-uncial script as early as the
Vienna Hilary (written before 560); while the half-uncial form
is frequent in the later specimens of uncial script: e. g. Mu-

nich 6278 Gregory's Moralia (written at Freising) Paris 9561 ;

Gregory's Pastoralis Cura (written at St Bertin); Paris 4884

Victor of Tours. In the text of the uncial Treves 36 Prosper
(written at Treves in 719) the uncial form is used, but in the
apparenti}' contemporary marginalia the half-uncial. That the
Carolingian minuscule of Brittany makes frequent use of the
Insular type of g^ and the Carolingian minuscule of Aquitaine
freel}' admits the Visigothic t}pe is natural enough. But it
 The Letters in Early Latin Minuscule 23

would be a great mistake to say that the use of the 9-form

always points to Spanish influence, or even that the use of
the 5-form is confined to Continental centres where Insular
script waF. formerl}- or still practised.
41. Just as a developement of the 9-form'^
the 5-form is

(the upper part of the curve being replaced b}' a straight ho-
rizontal stroke), so the 3-form is a developement of the 5-form
(the straight horizontal stroke being curved down on the left).
The rough-and-read}- rule for dating- Carolingian minuscule,
that the earlier specimens keep both bows of the 3 open while
the closing of the upper or the lower bow or of both indicates
a later date, justifies itself b}- the argument that the open
form is least removed from the half-uncial parent type. That
the rule has many be seen from a glance at
exceptions will
'
Introd. facs. 154 and 155 (of the years 821 and 823, with the
'

upper bow, and sometimes also the lower, closed) and facs.__
158 (of about 840, with both bows open). Scribe A of the
Ada Gospels closes both, scribe B keeps the lower open but
sometismes closes the upper. The MSS. written at Cologne
in Archbishop Hildebald's (794-818) shew both the open form
and the form with the top bow closed and also the parent
5-form (with the horizontal stroke curved). The group of Mss.
at Munich which were written for Princess Kisyla, Charlema-
gne's sister, shew the same trio.
^
42. But the substitution of a circle or loop for the upper
bow of the 3-form is older than Carolingian minuscule. It was
a feature of cursive script, the result of writing the 5-form
without lifting the pen. Beginning at the right-hand side of
the upper stroke (the cross-beam) the scribe curved that stroke
dowwards on the left and, without lifting his pen, continued
the curve into a circle or loop until he reached the centre of
the upper stroke, from which point again he proceeded in a
donnward direction with the rest of the letter. A bizarre Ita-
Han t3'pe of g illustrates the procedure (PI. I 45). It appears
in the cursive addition (foil. 117-118) in Verona 38 (written at
Verona) in Vat. Barb. 679 Cresconii Canones (written at
;

Farfa); iu Carlsruhe Reich. 222 Apocalypse, etc. (b}' the first

of the scribes, presumably an Italian) also, slightl}- altered,
;

in Milan C 105 inf. Hegesippus (written at Bobbio), etc. With

this slightly altered form ma}' be compared a Merovingian type
 24 VV. M. Lindsay

with the top-Stroke sinuous throughout (see I 6 *

gessit ', 7 "
sin-
^ilariter ').

43. While Carolingian minuscule adds, as the last element

in the letter, the small stroke to the right of the top of the
3, Italian begins with the complete top-stroke, starting from
its right-hand end (PI. I 46); and this Italian t3'pe is also the
prevalent type in Autun 20\ written at Au-
Burgund}- (e. g.
tun; Montpellier 84, written at Couches). The upper stroke
may be sinuous as well as straight. Not merely the comple-
tely closed upper part of the letter but also the shortness of
the lower part distinguish it from the Carolingian type (with
longer sweep of the lower bow of the 3).
^ 44. In the Corbie ab-script too, a script of Charlemagne's
time, the lower {«art is kept equal in size to the upper or so-

metimes less. The upper is a neat oval, completel}' or almost

completely closed, and the upper stroke is treated either in
Italian (and Burgundian) or else in Carolingian fashion (see
•
Introd. facs. 128).
'
This large oval head appears also in
MSS. like Petrograd O. I 4 Cassian's Collationes (Corbie en-
type), Paris 17371, foil. 1-153 Jeronie on Jeremiah (written at
St Denis, 793-806), etc.
45. Another wa}' of writing the halt-uncial form without
lifting the pen was to begin at the left side of the top hori-
zontal stroke and not continue the stroke past the point at
which the pen had to start on its downward course- This
produced a form which might be compared to a flat-headed
(numeral) 3, but which ma}- more conveniently be termed the
z-variety (resembling Greek minuscule Zeta) since it un-
doubtedU' was sometimes liable to be confused with the letter
z b)' transcribers. It is known to the half-uncial scribes of all

countries, and was the form favoured in the half-uncial pas-

sages of Tours MSS. from Alcuin's time (see Introd. facs. ' '

132), from which it found its wa) often into Tours minuscule.
At Tours (and elsewhere too) the top horizontal stroke is fairly
long and the whole letter is fairl}- broad; in strong contrast
to a form affected by some narrow
Insular scribes, a ver}'
type of letter which begins well above the line and descends
well below: e. g. in the marginalia (apparently written by
Boniface) in the Codex Fuldensis; in the marginalia of the
 The Letters in Early Latin Minuscule 25

Douce Primasius (see '

Ir. Min. '
pi. iv.) : in part of Petrog^rad
O. I 15 Theologica Varia '
beg. 8 cent. '), etc.
46. In ligatures (especially ^^n, _^r) the half-uncial form^
became like our S.
Indeed the ligature j^i exactl}^ reproduces
our printed S, the i forming the tag at the top-end of our
letter. This ligatured form is sometimes used independently
by scribes e. g. in the Ussher Gospels (see ' Introd. facs./
:
'

134. line 6); in Milan D 23 sup Orosius ('Insular' half-uncial

of Bobbio) and in some minuscule specimens of Insular script.
;

Confusion with in transcription is conceivable.

.t Both the
half-uncial MSS. close the under curve of the letter into a
loop (similarly the half-uncial Durham A II 16 and B II 30 in
^«, joro). But minuscule Insular £- which terminates in a loop'^
below is not found before the ninth century (occasionali}' in
the Book of Armagh of 808) and may generally be taken as^
evidence of a date later than our period (e. g. the Bamberg
Scriptores Historiae Augustae).
47. Lasth' must be mentioned a curious disjointed Me- ' '

rovingian variety (PI. I 47), in which a curved stroke starts

from the foot of the completed letter and is continued to the
line or above. I have found it in Berne 611 -j- Paris 10756, St

Gall 214 +
St Paul (Carinthia) frag., London Add. 24143; also
(in a modified form) in London Harl. 50^1, foil- 79 sqq., a MS.

already cited for its corkscrew shafted /, â (i^ 12).

' '

H.
48. The minuscule of Burgundy can be recognized by the
characteristic form of /i. The shaft (sometimes slighth' curved)
slojes downward from left to right and is (like the shafts of
other letters in Burgundian minuscule) not '
clubbed '
(i. e.

thickest at the top) but '

pillared '
(i. e. thick throughout).
But /i with a sloping shaft (straight or slighth" curved) is
not by any means unknown in other types of script. It ap-
pears in the Veronese cursive of Verona 163 Claudian and Ve-
rona 33 Augustine de Agone Christiano; in Cologne MSS.
frequenti}'; in the Swiss minuscule of St Gall and Chur and
'

in Einsiedeln 157; in the Corbie ab-type, and so on. (' Introd.

facs, 129, line 5 shews an example from an 8 cent. MS. of ' '

Murbach).
The Insular '
autem '
symbol was treated b}' scribes as k
26 W. M. Lindsay

with a appended, and is subject to the same modifica-

'
tail '

tion of form as the letter itself.

49. Correctors often add h (when omitted b}' a scribe) in
a conventional form, a small suprascript (Greek) rough breath-
'

ing (see Introd. p. 64). This is usuali}- termed the dasia

' '

(the Greek for rough breathing. In most MSS. of our period

' "

it has clearl}' been added by a corrector after our period.

But not in all. In a MS. written at Soissons c. 700, Brussels
9850-2, the correction (which occasionali}' takes the form of a
small //) seems contemporary. And in a Freising MS., Mun
ich 6228 f' 8 cent.) ", where it is very frequent, it seems to
come from the scribe himself; also in Lucca 490 (of c. 800),
e. g. fol. 7' menest//ei, fol. 20^" cleant//es, etc.
The paragraphus (the sign of a new paragraph) in an
' '

earlv Tours MS., Paris nouv. acq. 1575 Eugippius (also in

Paris 13386, etc.), takes a shape indistinguishable from the
'
dasia and would cause trouble now and then to transcri-
',

bers. And the same mark sometimes pla}s other parts e. g. ;

in a MS. later than our period (Vat. lat. 1570 Virgil with
Commentar}) s\'ntax-mark ', used to indicate the construct-
*

ion of the sentence, sometimes have this form. Thus the

dasia-mark stands above the Preposition and its Noun in the
line (fol. 34"") frigidus obstiterit circinn praecordia sanguis.
:

A transcript might conceivabl}' shew circJuwi praecliordia.

1.

50. I.
'
/ ionga '.
To give the letter i some individualit}-
the Romans This longa plays many
often lengthened it.
'
i
'

parts in the MSS. of our period, but never indicates the long
quantit\ of the vowel, as in Roman inscriptions of the earl}^
Empire. It also takes man}- forms: (i) it ma}- stand on the
line and be elongated to the height of letters like â, d, /,• (ii)
it ma\- be projected below the line (whence our j) to the depth

of letters like p, </, while its upper part stands on a level with
the shaftless letters; (iii) it ma}- be elongated at once above
and below (e. g. very often in the Corbie ab-type see In- ;
'

trod. facs. 128, last line, for a poor example).

'

51. It is only in cursive (especially the earlier cursive) that

we find elongation i whatsoever, without distinction.
of an}-
This promiscuous use ma}- be illustrated from the ancient mi- '
 The Letters in Early Latin Minuscule 27

nuscule '
of the scholia of the Bembine Terence (with e. ».

nimlum). Practically all scrioes who use '

i long-a "
follow (per-
haps often unconsciously) certain limitations and in many ty-
pes of script there are definite rules enforced on writers.
52. The rules for Beneventan and Visigothic minuscule
have been investigated (b}' Loew Stud. Pal. "). In both these '

scripts '
i longa '
is used(l) for initial (^) /, ,2) for j (i. e. conso-
nantal i)\ so that a South Italian or Spanish scribe writes In,

Igitur, the only difference being that a

lam, malor, alo, etc-,

South Italian calligrapher avoids i longa before a shafted ' '

letter (e. and especially in ille ') more

g. in '
ibo ', "
ipse ', *

than a Spanish calligrapher. This Beneventan and Visigothic

'
longa is often hardly distinguishable from / and has been
i
'

mistaken for it not mereh' b\- Caroline minuscule transcribers

(who never use longa ') but b}' modern editors. Thus aio
'
i

becomes alo, ìuaiiis becomes malus. On the other hand it

preserves the intelligent transcriber or editor from the danger
of confusing iu^ with vis, etc. ("').

53. For other scripts than Beneventan and Visigothic the

usage has still to be investigated (see below.). The few sta-
tistics which have been collected and published seem to re-
quire revision, for some collectors, after the}' have found In,
Inde, lus, lustus, lungo in the opening pages of a MS., will
hasten to pronounce that ever\- initial i is so written whe- ' ";

reas all that these five examples prove is that initial / be-
fore n and before ii is elongated by this scribe. The possible
uses of longa are for
'
i
"
:
—
(1) any initial i. The long form would be a handy in-
dication of the beginning of a wordf
(2) initial / before ii. The long form would prevent in
from being mistaken for m. In the rude half-uncial of the
Cathach, a Psalter written probably by St Columba, the Irish
saint writes longa (or rather longior for it is onl}' slightly
*
i
'
" '.

higher than shaftless letters) in the word (or opening syllable)

(') In a phrase like ad insiclam there is not necessarily initial i. The

scribe may visualize it as one word adinstdani (like ad/ero).

(-) Loew 1. c, p.
13. cites Tac. Hist. 4, 48, 10 Hegatorum vis adolevit),
where waver between vis and ius. Since our unique MS. of the Hi-
editors
stories is in Beneventan script, its reading is undoubtedly vis. For ins would
inevitably be written with longa by a South Italian scribe. '
i
'
 28 W. -M. Lindsay

in whenever he uses the minuscule form of n (In. but ix). It

is true that one might expect to find longa before n in ' '
i

the middle of a word also, in order to keep. e. g., vifio res

from being mistaken for iiniores but there seems to be no ;

instance in the book-hand of our period (except delnde. prolu-

de, etc., which may be otherwise explained).
(3) initial i before ?^ The long form would prevent iii
from being mistaken for ui (e. g. ins for vis). Here too the
extens-ion to in in the middle of a word is conceivable; and
certainh' elus, conlunctus, etc are not unknown in the book-
hand of our period. The collector of statistics must seek evi-
dence whether these do not rather (as in Beneventan and Yi-
sigothic minuscule) belong to the following class.
(4) i
z). both
(consonantal
initial (e. g. lam, lovis, lungo'

and medial Mala, malestas, malor, malus).

(e. g. For it is
not conceivable that a scribe who used longa (as we use '
i
'

/I to indicate the consonantal sound of Latin / should content

himself with indicating this sound at the beginning of a word

onl}-.
54. .A.more purposeless use ot longa is rather a fea- "
i
"

ture of cursive and, when it appears in books, indicates an

earh' script which has not yet discarded cursive features:
(5) after letters like r, / and (the '
half-uncial '
t}-pe of) g
all of them ending in a thrust to the right. We find rl. tl

gl (some or all) in the *

ancient minuscule
of the marginalia "

of such MSS. as the Bologna Lactantius and L\-ons 253'" Au-

gustine's Civàtas Dei in the half-uncial of such MSS. as the
:

Cassel Hegesippus; in the Corbie en-type not gi, for the

'
half-uncial t}pe of g is not used etc. (For an example from
'

the earl}' minuscule of Tours, see '

Introd. '
facs, 130, line 3
'
altaris ').

It seems reasonable to connect with this fashion that cursive

Insular fashion of writing ri, ti, etc., of which I have given
details in Zeitschr. celt. Phil. 9, 304. The subscription of the
Stowe St John's Gospel fragment offers in five lines no less
than eight examples of ri written with a long vertical sinuous
stroke following the r. A common Insular form of the liga- //

ture (see below, s. v. '

T seems to superimpose this cursive
*)

'
i longa '
on the body of the /. And the less common Insular
ri ligature (or rather connexion ') does the same (see PI. I
'
 The Letters in Early Latin Minuscule 29

8 '
scriptum '). And just as in the ancient mi-
scriptam ', 9 ' '

nuscule of, let us sa}-, the marginalia of Vat. Reg. 886 (e. g.
'

fol. 73" armatus in) longa often transects the wide upper
'
i
'

curve of a preceding s, etc., so the Insular scribe often writes

/Î (especialh' in the word in Explicits) with this cursive ' '

finit

Insular i longa straggling down through the wide upper

' '

curve of the f (see above, s. v. F '). '

55. In the Cassel Hegesippus longa after / is often "

i
'

projected below the line (usuali}' curved like our wanting j

the dot). This was the fashion which gained most universal
acceptance (see Introd. facs. 153, line 6). We find it all
' '

through our period in all t}pes of Continental script and in

some utilized to distinguish the assibilated pronunciation (e. g.
in natio) from the iinassibilated (e. g. in ileitis). See below, s.
V. ^
T •).

56. The analogous expression of r/ (on that of gi see

below) usuali}' appends to the elevated branch of the r a long
sinuous vertical curve like Insular cursive longa (e. g. in '
i
"

the ancient minuscule marginalia of Lyons 523 ''* and of the

' "

Codex Claromontanus). This I'i ligature (see Introd.' facs. '

120, line 3) is a feature of all Continental script throughout

our period, but not of Insular. Its rare appearance in Insular
script proves the MS. to have been written in a Continental
scriptorium (e. g. Munich 6298. written at Freising). An In-
sular scribe writes, as a rule, an ordinar^• r followed b\' an
ordinar}' i. A short variet}' of the rz ligature might be mist-
aken for n.
57. II Subscript
In uncial script id is often expressed
i.

b}" prolonging the second upright of u downwards in a straight

line, so that ui resembles a ver}' open q (see below, s. v. ' '

'
U '). Similari}' ni. etc occasionali}' in the ancient minuscule' '

of such MSS. as the Basilican Hilar}'. In minuscule script /

prefers a comma-form similari}' appended (^). It is a constant

feature ot Insular script, and a frequent feature of the Con-

tinental of scriptoriums under Insular influence, e. g. Freising,

Literally
(^) subscript i (i. e. set directly under the centre of the
' '

letter)appears in a Vatican MS. written at Treves in 810, Vat. Pal. 1448-

and in another, Vat. Pal. 212, of uncertain provenance (also Treves?). These
two MSS. sometimes mi by ;/ and
e.xpress ni, ;« with a comma written (at
a very slight inter\'al) under their centres.
30 W. M. Lindsay

Murbach; but from which Insular influence

also of others
seems remote (e. hi in the North Italian minuscule
g. mi^ fti.

of Vercelli 202 Isidore's Etymologies; mi. ni in the Lons-le-

Saulnier Bede, written at St Claude, Jura, 804-815, a MS. which
shews no Insular abbreviation; in Beneventan minuscule, etc.).

It is commonest in mi, ni\ next in hi\ while (.minuscule) ;«" is

perhaps confined to Insular script (also appearing in the Con-

tinental of a Murbach MS., Colmar 39, e. g. fol. 103'' quia '). '

Peculiar too to Insular script are the rarer bi (e. g. the Book
of Mulling and the Book Berne 671, of Armagh, for Irish;
for Cornish; Petrograd O. XIV 1, for Anglosaxon), di (e. g.
the half-uncial St Chad Gospels, for Welsh?; Milan C 301 int.,
for Irish; St Gall 761, for Anglosaxon). xi (e. g. the half-un-
cial St Gall 51 and the Le3'den Priscian, for Irish; while the
Anglosaxon half-uncial of the Maeseyck Gospels exhibits ra-
ther dropped than subscript i, and the Anglosaxon mi-
' ' ' '

nuscule of Paris 9527 appends a long curve to the upper

right end of the x). The bi ligature might be mistaken for
//. The di (e. g. in St Gall 761) is indistinguishable from the
ti ligature.

58. Insular too is the comma appended to the right end

of the curve of / in // (also in the Continental script of Mu-
nich 210. from Salzburg). This comma, when in small size,
is hardU' distinguishable from the flourish in which the curve

of Insular / sometimes ends (e. g. in the Freising Anglosaxon

script of Munich 6297) and perhaps that is wh)- // in Insular
;

script usually prefers '

dropped to subscript i. (Continental
' ' '

script, we have seen, as sometimes Insular, appends the i to

the right end of the branch, not of the curve, of /, a practice
from which arose a common Continental and Insular form of
the ti ligature; ^ 115).
59. The Insular method of writing si (see above, s. v.
'
F ') may perhaps be classed with subscript So too may ' '
/.

that curious expression minuscule

of peri (e. g. in the Irish
of the Book of Mulling, the Book of Armagh, etc., and the
Nothumbrian of Vat. Pal. 68), with a curve appended to the
of the Insular abbreviation-symbol for per {p with a
' " '
'
tail
'
tail '). A transcriber might be pardoned for mistaking it for
a common Insular abbreviation of pits (c*. *
Introd.' facs. 127,
line 1 '
bus ").
 The Letters in Early Latin Minuscule 31

60. Like Insular wavers between sub- //. so Insular // •

script i (e. g. in the Naples Charisius and the Book of üim-

"

ma, for Irish; in the Moore Bede and the glosses of Petro-
grad F. I 3, part z, for Anglosaxon) and dropped /. A cu- ' "

rious cursive ai (with the suprascript form of a) in Epinal 68

(of 744) resembles z (,e. g. m(3;/oribus The appendage of i .

to suprascript u which so often expresses ui (in qui etc.) ' '.

in the minuscule of Bobbio, St Gall. etc.. is mentioned else-

where (s. V. '
U *).

Dropped i. It has been alread\- remarked (s. v.

61. III.

•
C ") when written, as so often in all kinds of script,
that ci.

with dropped i (i. e. with the letter beginning much nearer

'
'

the line than usual) was dangerously like the uncial tvpe ' "

(sickle form) of g when the handle of the sickle was not *

"

sufficientU' long. The Insular «'but not exclusiveh- Insular)

ligature fi (with the / appended to the right end of the branch
of f) also exhibits what may be called '
dropped /. while the "

Continental fi-ligature is identical with the Insular si-ligature

(see above, s. v. '
F '). '
Dropped '

/ is common after / from

the earliest times (e. g. in the uncial of the Bologna Lactan-
tius. etc. ; in the half-uncial of the Basilican Hilarv. etc. : in
the '
ancient minuscule '
of the marginalia of MSS. like the
Basilican Hilary and often in the word contuli in their sub- ' '

scriptions) and in most types of script. (For an Anglosaxon

example, not a very good one. see 'Introd.' facs. 142 line 3).
It is sometimes mistakable for b. alii resembling abi. etc.

62. Dropped and subscript i are. as we have seen,

'
"
' "

not always to be distinguished. They ma\' often convenientlv

be included under the wider term enclitic /, where the letter '
'

is treated b\- the scribe as a mere appendage to the prece-

ding letter. The S-form assumed bv the Insular ligature gi

has been already mentioned (s. v. G "). Merovingian gi ap- '

pends i in a much longer curve to the head of the letter (see

^
Introd.' facs. 124. line 2; 125, line 1). The same long ap-
pendage (attached in similar fashion to the open 3-form of^^)
appears in MSS. like Epinal 29 (of 744); and Beneventan gì
preserves this early practice. In ei (see above, s. v. "
E ') the
same (usually) long curve (or straight line) is appended to
the right of the '
tongue '
of the e in Insular as well as most
Continental scripts ; although, what we might expect in Insular^
 32 W. M. Lindsay
'
subscript '
I (as in //) appears in the half-uncial of Durham
A II 16.
63. When the '
cursive '
Insular e (the '
figure eight '
form)
'

isemployed, the / merel}- continues the curve of the '

tongue
and might be ignored b}'^ a transcriber sometimes. Similari}'
when the '
cursive '
Continental e (the reverse of the Insular)
is employed; the ligature being then dangerousl}' like the
Continental ligature et. This Continental ei appears also in ' '

Insular minuscule, e. g. in the Book of Mulling (see above, s.

V. "
E '). Curiousl}^ enough, a Freising scribe of Munich 6262
854-875 A.D.) writes /z' (in '
viginti ',
'
ubertatis ', etc) preci-
sely in this fashion (see below, i? 119).
r^ 64. These ligatures ei^
fi^ gi, li. ri, ti became obligator}'
in Beneventan developed stage which
script (especiall}' at the
came after our period); and, as we shall see (s. v. T ), one '

particular form of the // ligature (resembling a reversed Greek

minuscule Beta) became obligator^ for the assibilated sound
of //. These rules merel\" reduced to s\'stem the practice
already prevailing in Italian minuscule, for in North land
Central?) Italy we find a script hardly (if at all) distinguishable
from the earlie^r stages of Beneventan. We
find there also a
scri{3t hardl}' distinguishable from Caroline minuscule, which
attained great beauty at Verona when Pacifico (d. 846) was
head of the scriptorium. The presence of longa seems a '
i
'

good criterion for distinguishing this Caroline t\pe of Italian ' '

minuscule from the Caroline minuscule of France, although

the minuscule of the North Italian scriptoriums (Verona, Ver-
celli, Novara, Nonantola. etc.) has really not \et been inve-
stigated.
..-— 65. We
have seen that a mistake like malus for maiiis^
alo for aio suggests an exemplar in Beneventan (or rather an}'-
Italian) or Visigothic minuscule. Suggests not proves ' ', ' '.

For, in the first place, the letters i and / (also /) of Rustic

Capital (and some Uncial) are easil}' confused at any part of
a word. And, in the second, the more that any species of
minuscule retains of its cursive origin, the more readily it
admits longa '
Three Merovingian types of book-hand,
i
'. ' '

namely the Luxeuil type, the Laon az-t}'pe, and the Corbie
ab-type, all abound with longa although it is perhaps
'
i
'
;

onl}' in the Luxeuil type that maiiis., etc. could be misread

 The Letters in Early Latin Minuscule 33

as malus, etc, and even there not ver}- often. The use of 'i

longa '
lor j is a feature too (but seldom a constant feature)
of the oldest minuscule of many writing-centres, such as Tours
(e. g.London Egerton 2831;, Corbie e. g. Paris 4403 and '

12239), Murbach (e. g. Epinal C)8, of the year 744), Cologne

(e. g. Cologne 210 and 98). B}- the time of Charlemagne these
centres dispense with '
i longa '
altogether (the calligraphic
Tours minuscule of Alcuin's or PVedegisus' time, the Corbie
minuscule of Maurdramnus' time, the Cologne minuscule of
Hildebald's time).
66. A mistake like lacto for iacto is much weaker evidence.
For the use of longa for initial i (or /) was more widel}"
'
i
'

prevalent than its use for j (medial as well as initial). Most

prevalent of all was its use for the initial of in and any scribe ;

who recognizes longa at all is sure to write In, Inpono,

'
i
'

etc. (probably also exinde, etc. and Impono, etc.). Practicall}-

every specimen of Insular minuscule offers In (often closely
resembling hi), and most add other examples of initial
'

longa '
i

such as lungo. Ita. At some Continental centres of Insular

script we see (in their Continental script especialh) a conflict
between the Insular use and what we may call the Conti- '

nental disuse of
'
longa For example, in a Fulda MS. '
i '.

written before 847 by two scribes. Vat Reg. 124, the one
avoids longa altogether, the other writes it always as the
'
i
'

initial of in and sometimes of other words such as ita \

' ' '

In the Lorsch MSS. at the Vatican Library which shew the

work two or more scribes we often find the same conflict.
of
67.Another use to which longa was occasionally put '
i
'

has still to be mentioned, for the second of two neighbouring

i's; e. g. in Cologne 210 exilt, filil. Sometimes in the j-form,
e. g. Verona 59 filij (without the dot). These would escape
the danger of confusion with n. The second unit of iii (xiii,
etc.) or the second and fourth of iiii (xiiii, etc.) are often si-
milarly treated. (On the Spanish ligature zV, see below, s. y
V. 'T').

L.

68. The '

foot '
of this letter usually substituted in minu-
scule a curve for the angle of the majuscule form, and often
(e. g. in the earlier specimens of Anglosaxon) this curve is

3
34 W. M Lindsay

continued right under the following letter (or letters). In- '

trod.' facs 114 shews in line 6 an example from an early MS.

of Bobbio of this practice where the foot is not curved ' '

and in line 9 (' facile ") an occasional earl}- type (with arched
'
instep ') which might be mistaken for //, a t\pe perhaps com-
moner in uncial than in minuscule. When the curve of mi-
nuscule / is reduced to a minimum, the letter is easily confu-
sed with the long form of i (see above, s. v.). When an or-
namental tag is added to the straight or curved) foot, /might
be mistaken for the ligature // (with that dropped i seen in ' '

line 3 of the facsimile just mentioned).

69. A
kind of dropped / appears in earlier MSS.. such
' '

as Epinal 68 of the year 744), London Add. 11878 (of the

Luxeuil t3'pe). Paris nouv. acq. 1507, in ligatures like el^ fi^
the head of the / being attached to the tongue of the e, ' '

the '
branch '
of the /, and its '
foot '
prolonged in a sinuous
sweep under the line. On certain modifications of the shaft
of / (and the other shafted letters), e. g. the *
corkscrew shaf-
ted '
/ of some Merovingian MSS., see above, s. v. '
B '.

M, N.
70. While the majuscule form of
is rare in the minu- m
scule of our period form sometimes in Epi-
(e. g. the capital
nal 68 of the year 744; the uncial form, i. e. with rounded
and not angular shoulders in the Corbie ab-t3pe of Petro-
' '

grad F, 111), the majuscule form of 7i (angular both in uncial

and in capital script) is common all through our period. Its
special frequency in the Corbie en-script makes it a prominent
feature of this t\pe it is much affected in the calligraphy ot
;

Tours, etc. and, in fact, it may be said to be one of the

;

marks which a palaeographer uses in dating Caroline minu-

scule. But until statistics are collected, this test of date must
be used with caution. For example, it is of common occur-
rence in so late a MS. as the Corbie Gregorian Sacramentar)'
of 972-986 (Paris 12052). And we must allow for caprice of
individual scribes. The Ratisbon scribes of Munich 14468
written at Ratisbon in 821) do not use it; those of Munich
14437 (written in 823) do. In the Cologne MSS. of Hildebalds
time I have not noticed it, although it is common in the
earlier Cologne 98, Cologne 210, etc.
 The Letters in Early Latin Minuscule 35

71. This uncial and capital n (like our \) is used exclu-

sivel}' inContinental half-uncial script, which thus escapes the
danger of having n and r confused by transcribers (see below,
s. V. R '). The form resembling our n should therefore be
'

called the minuscule rather than the

' '
half-uncial form al-
' '

though the analogous form of m is rightl\- called half-uncial

'
"

(being, in fact, the uncial form reduced in size and with the
first curve turned into an upright). Insular half-uncial scribes
admit both the minuscule form of n and the other, and some
(e. g. in the Lichfield Gospels) affect variety- where n is re-

peated, e. g. a«no, «on.

72. The rough-and-ready rule for dating the minuscule of
our period, that m and ii do not end in a slight upward turn
to the right (a great departure from the uncial form of w)
does not hold for all scripts, e. g. Merovingian book-hand of
various types and Beneventan (from the earliest stage Me- .

'

rovingian examples of this late form will be seen in 'Introd

' '

facs. 124, 125, 126, 128; Beneventan in facs. 120 (which should
however be dated 874-893 and not 812 '). Insular calligra.
' ' *

'

phers (and some others) often so exaggerate this uncial '

feature of minuscule ni (and n) that the final stroke is brought

in a sweep under the line till it reaches a point immediatelv
below the middle shaft of the ni (or e\en further to the left).
The\" do this especially, but not onl}'. at the end of the line.
And Insular half-uncial script, especially at the end of a pa-
ragraph or a page, sometimes indulges in a freak-form, with
this final curve fantastically elongated (see Introd.' facs. 135,
'

end).
73. There is an early, a very earh* form of « (of the
majuscule shape) which closeh" resembles (capital) m^ and is
due to writing n without lifting the pen. The middle stroke-
instead of going down to the foot of the second upright, ma-
kes a slight dip and joins the second upright at the top (PI.
I 48). appears sometimes (nearly in this form) in that part
It

of the Book
of Mulling to which the St Moling (d. 696) sub-
scription is appended, and strongly confirms the tradition
about this early specimen of Irish minuscule. It appears also
in the early cursive of Verona 3 and fol. 403"^ of Verona 1 :

in the quarter-uncial of Rome Sessor. 55 Augustine in the

' '
;

corrector's entr}- on fol. 110 of the Basilican Hilar \- (correc-

36 W. M. Lindsay

ted in 509-510); in the "ancient minuscule' marginalia of Vat.

Reg. 886 and the Bologna Lactantius, and so on.
Quite a contran' formation of 7i (of majuscule shape)
74. ' '

is employed in the Corbie Leutchar-t}'pe of Berlin th. F. 354

(by the first scribe, whose n resembles his //), in the Anglo-
saxon half-uncial of London Ro\-al '1 E vi etc. The middle '

stroke is horizontal (usualh' on the line) and begins from the

foot of the first upright (from its centre, when the first upright
is projected below the line). Some Anglosaxon half-uncial
scribes make this type of n broader than the normal form.
The normal (majuscule) form is however occasionalh' fantasti-
cally extended breadth (often wifh a reduction in the height
in
of the letter), especiall}- when a scribe found that he had left
himself too much space for his task and wished to extend his
writing so as to fill the whole of the last page of a quire (e.
g. ;/ome;/ in the Anglosaxon half-uncial of London Royal 1
E vii).

75. More often a scribe desired to save space (and secure

correct syllable-division) at the end of a line. In such a case
Insular scribes (especially in half-uncial MSS.) write n at the
end of a sideways (like our numeral 3). The write m si-
line
milarh', but with an additional curve (see Introd.' facs. 135,
'

line 1, end); while urn is expressed by capping this long spiral

with a suprascript u, am with a suprascript a (e. g. Durham
A II 17; although suprascript a is a great rarit}- in Insular
script). Sometimes scribes write ii above the line, in majuscule
or minuscule form, or even in that tachygraphic form which
exactl}- resembles our z and in the Anglosaxon half-uncial of
;

the Lindisfarne Gospels we find occasionally stilted n^ with ' '

the bod\- of the letter (in majuscule form) above and the first
upright (like a stilt) prolonged down to the line.
76. Details of these contrivances for reducing the space
occupied b}' w and n will be found in Not. Lat.' pp. 342 sqq.,
'

as well as of another contrivance which is usually (but un-

scientifically) called an abbreviation-s3'mbol for m and n^ a
' '

suprascript stroke which in ancient minuscule

'
(e. g. the
'

scholia of the Medicean Virgil), in majuscule from the earliest

times, in the calligraphic minuscule of Verona (and other
centres) is often not suprascript but treated like an ordinar}'
 The Letters in Early Latin Minuscule 37

letter ('}. In the minuscule of our period it is, we may say,

only Spanish scribes who '
abbreviate "
«. so that intelligent
transcribers would never be in dancfer of misinterpreting the
symbol unless the\' had a Visigothic exemplar. And even
Spanish scribes often distinguish, adding a dot (or a shorter
stroke) above (or under) the abbreviation-stroke when m and
not n was symbolized.
77. It seems hardl\' worth while to give details of the
forms assumed by the abbreviation-symbol of w in the mi-
"
'

nuscule of different countries and writing-centres, since man\'

scribes do not confine themselves to one form and most scribes
make no distinction between the s\-mbol of m and the general
abbreviation-stroke. Insular (especially Irish) scribes are an
exception. The normally reserve a stroke curved backward
(or hooked) at each (or one) end for w see below, s. v. 'U'):
so that we
can alwa\-s be sure what an Irish scribe means by
(let us say) a with a stroke above. With the m '-stroke it '

will mean am. With the ordinar^" abbreviation-stroke it will

mean ant. The peculiari}' South Italian m-stroke, like a small
numeral 3 (see Introd.' facs. 120, from a Monte Cassino MS-
'

of 874-893, not 812 ") hardly came into use till after our pe-
*'

riod. The other shapes, so far as the}' call for mention, ma\'
best be mentioned in an account of the general abbreviation-
stroke itself. (On the ligatures no, on, ns, nt see s. vv. O ' ",

^S\ 'T'.).
78. A curious variety of n (the majuscule shape) found in
the Merovingian minuscule of London Harl. 5041 (as in the
Bobbio uncial of Milan B 159 sup. perhaps deserves notice
(PI. I 49). It might suggest at to transcribers (cf. § 114),
Tarnen might be mistaken for tameat and altered to timeat
or tumeat.

O.
79. Just as the letter / was made more observable bv
being elongated (' longa '), so in Merovingian cursive the
i

letter o takes the form of Greek minuscule Delta, a form often

(*) \. e. put after the preceding vowel. It is, I think, wrong to speak
of this as a device to eke out the length of the line and not leave a blank
space at the end. For the scribe often puts the m-stroke in the margin,
outside the line.
 38 \V. M. Lindsay

assumed b}" the letter d in Insular minuscule (see above, s. v.).

In the book-hands of Merovingian t3pe, e. g. the Luxeuil

script, the Laon az-script, the N. E. France " script, the ''

Corbie ab-script, we find this Delta '-form of o much in '

evi-
dence. For an example from the last named script, see '
In-
trod.' facs. 128, line 1. And it occasionally shews itself in
other script of the Epinal 68 (of the year
8"". century, e. g. in
744). In the Merovingian book-hand of Berne 611 it is some-
times indistinguishable in ligatures from Insular cursive e (§27);
or resembles er^ etc

4f Subscript o is, like subscript /, a feature of Insular

80. ' ' ' '

script, although (unlike i) it is never found in Continental.

Oftenest after ni (PI I 50), n^ h, and the ligature ti (see be-
low, s. V. '
T '), hut also after ;/, /, d (e. g. in the Lindisfarne
Gospels), / (e. g. in Durham A II 17). In the Welsh Liber
Commonei attached to the tongue of cursive Insular e
it is ' '

to form eo. Suprascript o is, in our period, almost as rare

'
'

in Insular script as suprascript a. It appears in two speci-

" "

mens of Anglosaxon script which ma\' come from the same

scriptorium (P2chternach i), viz. the Maihingen Gospels (middle
of 8.*'' cent.) and Paris 9382. The>- offer os (PI. I 51), o>\ om.
on, usuali}- so written. In Irish it shews itself at the ver\-
close of our period in the St Gall Priscian (written in Ireland,
probably in 845), where one of the scribes, Finguine, writes
om, etc.. as in PI. I 52. It is, like 'subscript' o, unknown to
^ Continental book-hand.
-^ 81. The nearest approach to 'subscript" o in Continental
script is the combination ro (see '
Introd. '
facs. 114, line 4;
119, line 10), in which howerer the o does not normally come
below the line. This combination (the branch of r usually-
high and angular) is a prominent feature of most Continental
minuscule (not Visigothic nor Beneventan) of the 8.*'' century ;

and indeed till the end of our period, e. g. b}' one scribe of
Paris 11533 Bible (Corbie, of 850), in Oxford Laud, misc. 120
(written at Würzburg, 842-855), in Liege 306 (written at St
Trond in 834). One scribe of a Murbach MS., Oxford Jun. 25,
foil. 1-59 ("8 cent.") writes it quite like 7W. Itappars in the
Anglosaxon minuscule of Tours in London Egerton 2831.
Of other combinations in which o is attached to a pre-
ceding consonant the commonest are co ('Introd' facs. il4,
 The Letters in Early Latin Minuscule 39

line 7), to (facs. 126, line 7, line S), fo,eo, go, fio. Even Insular
scribes occasionali}' write mo (and fio) as in PI. I 53, instead
of with '
subscript '
o.

These combinations are 8.*** century usages and indicate

that Continental book-hand has not yet fully discarded the
practice of cursive. The o often terminates in an upward
X stroke which suggests the Merovingian '
Delta '-form.
82. The combination os^ in '
ligature '-form (PI. I 54) or as
a mere '
union a device of uncial and half-uncial
'
(PI. I 55), is
scribes for saving space at the end of a line, and is similari}-
employed b}' some minuscule scribes, e. g. in Munich 4547,
a MS. written for Princess Kisyla, Charlemagne's sister. Of
the combinations in which o precedes a consonant the most
universal is or (see below, s. v. R '), in which the r is usually "

sacrificed to the o, as much as fi in a common uncial form ot

ofi (PI. I 5Ó . On the rare ligature oe see § 6.

P.

83. The round of p in minuscule script (as in half-uncial

and the later stages of uncial) stands on the line; and when
the lower part of the round does not reach as far as the shaft,
we have the '
open '
form of the letter (see '
Introd. '
facs. 119,
124, 129, etc.). open '/ have not yet
Statistics of the use of '

been collected (except that in Beneventan script onl}- close

p is reported), but it appears that a very open form of p is
a mark of antiquity, as in Introd facs. 142 (from the Anglo- ' '

saxon minuscule of the Moore Bede, written c. 773, with p

often dangerously like Insular r).

84. A form like Greek minuscule Rho characterizes North v

Italian cursive (see '

Introd. '
facs. 114, from the Milan Maximus).
In Visigothic cursive too it is known, but often takes the form
of an perched on the top of the shaft of the letter (see
'
Introd.' facs. 115, line 8).

Cursive scribes threw open the left side of the round

85.
ofp b}' attaching the upper curve to a preceding letter and
beginning the under-shaft of p at the foot of the round- (On
a similar formation of d and q see §§ 20, 87-89). In the Con-
tinental minuscule of our period (in Beneventan till long after
our period) three of these ligatures are current everywhere
except in Spain, viz. sp (see Introd facs. 1 14, line 5), r/, '
'
 40 W. M. Lindsay

xp (facs. 114, line 3). The second is coniusible with sp when

the branch of the r is not (as usual) angular. And sp^ when
narrower than usual, is confusible with that ligature by which
a Continental scribe expressed yi and an Insular, si (see above,
s. V. I ').'
The leaf-shaped ep (see Introd. facs. 114, line 6) ' '

has already been discussed (s. v. E '). Like ap (with su- ' '

prascript '
belongs to the earlier half of our
a), op, tp, it

period; although rop 128, line 3) appears even in a MS.

(facs.

written for Charlemagne's sister, Princess Kis3'la (Munich 4577,

in propter fol. 91'). The ep ligature has other shapes too
' '

v^ (see, e. g-, 'Introd.' facs. 115, line 9).

86. The approach to this cursive Continental p in
onl}'
ò Insular script is a form which might usually (or often) rather
be described as open at the top (PI. I 57). It is a feature
' '

of Insular half-uncial and of some (earl}) specimens of Insular

minuscule; although it turns up occasionally in Continental
script(e. g. fol. 169^ of Verona 37), especiall}* for initial P.
(On an analogous formation of Insular g see below, s. v.).

Q.

87. The minuscule form (like our q) is also the uncial and
the half-uncial. A half-uncial variety (seen in '
Introd. '
facs.
98) with the lower part of the round of q (as of p) not con-
tinued to the shaft does not seem to occur in the minuscule
oi our period. By
open q is meant rather a ^ in which the
'

upper part of the round is not closed. Of open q much ' '

the same ma}- be said as of open / (see above, s. v.). ' '

Statistics have |not yet been collected, except that in Bene-

\ventan script only close q is reported. A very open va- ' '

riety, in which the curve breaks away at once from the shaft
^Pl. I 57) is a feature of such earh' specimens of Insular mi-

nuscule as St John's Gospel in the Book of Mulling (written

by the Irish saint Moling, who died in 696); the marginalia
in the Codex Fuldensis (witten probahly by St. Boniface, who
died in 755) ; the similar marginalia (also b}' Boniface }) in the
Cassel Hegesippus ; the marginalia in the Douce Primasius ;

Petrograd Q. 1 15 (Anglosaxon of Corbie.?),

/ 88. In most of the Insular minuscule and half-uncial of
our period q is often more or less 'open', a common t}pe
having the left side of the round turned back at the top
 The Letters in Early Latin Minuscule 41

(PI. I 59); and thisleft side is sometimes (PI- I 60) ang-ular

^
instead of curved. (On a similar formation of d and / see
^ 25, 85). Equally open is a variety which in Continental
' '

minuscule is confined to the earlier stages. It resembles the

letter it with the second upright prolonged below the line,
and its dangerous resemblance to the ligature iii (with sub- '

script /) has alread}' been mentioned s. v. I "). This variety

'
*

is a feature of the writing of Winithar. the famous scribe of

St Gall in the middle of the 8.*^ century it appears (along ;

with the closed form) in the Tours Eugippius (" beg. 8 cent. "),
etc.. and in Merovingian book-hand often has the shaft not

perpendicular but slanting down (sinuousl\-) to the right.

But close (/ is no evidence of a later date. For example,"^
'
"

in the Luxeuil type of Merovingian script (see Introd. facs. '

"

124) q is open only when connected with a preceding letter.^

* "

89. This connexion of open q with a preceding o (see \

' '

"Introd." facs. 126, line 5) or e (facs. 128. line 3) or .>r. etc., is

an early feature of all Continental and Insular minuscule :..

e. g. eq in the Milan Hegesippus, of c. 700; vq in Wolfenbüttel

Weissenburg. 81 (of 772); sq in the Book of Mulling, Menna

16, Vienna 954 (all of * c. 700 "). It catches the eve in the
" X. E. France minuscule (of Autun 20 and Montpellier
"

Ville 3, etc.) more perhaps than in any other script; for in

such ligatures as aq little remains of the round of the q except
the lowest part.

R.

90. We ma}', with convenience

if not with strict accuracv,^

distinguish 5 r's in the minuscule of our period. Three of

them stand on the line (being of the height of the other shaftless
letters): (1 the "uncial" variety (PI I 60). (2) the 'half-uncial"
(PI. I 62), (3) the "
minuscule "

(^Pl. I 63). Two of them, in

cursive fashion, beginning at the same height above the line,
fall below as far as letters Hke / and q: the "cursive" with
(4) rounded (PI. I 64) or (5) angular shoulder (PI. I 65). The
first three represent three successive steps of departure from
the original type (R) standing on the line; the last two simi-
larly of departure from the same
t}pe placed partly above
and partly below the line. All five (but especially no. 5) ma\- be
sloping (upwards from left to right) instead of perpendicular.
 42 W. M. Lindsay

-^ 91. Of these five minuscule varieties we may rough!}' sa}'

that nos. 4 are Insular, nos. 3, 5 Continental. And in
1, 2,

most Continental minuscule we find no. 5 reserved for its

proper cursive function, to express r when witten in ligature
' '

(^or at least in connexion) with a following letter. One ligature

however is written with r standing on the line, the ligature
ri. For example, Introd. facs. 119 exhibits the 'minuscule""
'

variety everywhere, except that in re (lines 6, 10) and ro

(line 10) the cursive angular variety appears, and in ri
'
"

(lines 3, 7, 9) the same angular form stands on the line (the

i-appendage curving far below the line; cf. § 56). This pre-
vailing usage in the (earlier) script of our period (cf. facs. 114,
127. etc.) became stereot}'ped in the rules of Beneventan script,
viz. that final (*) r has a short stem (i. e. stands on the Hne)
while the stem of every other r (being in this script connected
with the following letter) is long (i. e. goes below the line),
.except the r of the ligature ri. (For details see Loew
Benev. Scr. p. 137; and for an illustration from Beneventan
' '

minuscule a little later than our period, Introd. facs. 120, ' '

taken from a MS. of 874^93, not " 812 ").

) 92. The calligraphers of North Italy (and some other
writing-centres) loved to extend the arm of minuscule r. ' '
' '

Thus in a London MS., Cotton Nero A ii, apparently written

at X'erona, the next letter seems to stand quite at an interval
from the r. Often the\' wrote the next letter (or letters' under
the elevated '
arm ".
In the Cambrai half-uncial of Cambrai 441
we find often a "
high '
form of r with a long '
arm stretched
"

above the letters which follow (e. g. in '

tranquilla '
on fol. 197
the arm high r stretches above afi).
of the ' '

/ 93. We
have seen (s. v. N ') that the resemblance of the '

usual half-uncial r (see Introd. facs. 99, 100) to the letter n

*
"

(in the form of our n) led to the exclusive use in Continental

half-uncial of the capital (and unci<-l) form of n (like our N),
whereas Insular half-uncial included both forms of the letter n.
Insular minuscule scribes were apprehensive of the dangerous

(1) It is only, as a rule, in the earlier part of our period that minuscule
scribes (of any country) so far neglect the division of words as to connect
the final letter of one word with the initial^of another. An example of final r
so maltreated will be found in '
Introd. '
facs. 131. line 1 'cor' from a MS.
of the year 744.
 The Letters in Earl}- Latin Minuscule 43

resemblance of r (no. 2 in cur list) to n.

this few admit A
the no. 3 variet}' along with no. 2 (e. g. the Murbach scribes
of Colmar 38, part ii). A few confine themselves (in the main)
to the no. 1 variet}' (e. g. the Ratisbon scribes of Munich
14080 and 14653). But most found safety in the use of no. 4,
the normal form of r in Insular minuscule. It ma\' be said
that b\' so doing they courted another danger, the confusion
of r with for Insular
JT ; falls below the line (see below, s. v.).
jt

Stillalthough the confusion of these two letters usually points

to an Insular (') minuscule exemplar, the error may be charged
to Continental transcribers who were accustomed to a short-
stemmed r and an (taller than r) which stood on the line
.s-

and therefore failed to notice the difference between the two

unfamiliar forms. In realit}' there is a marked difterence
between Insular r with lull}' developed '
arm and Insular
'
.^

'

with a mere stump in place of an arm see, e. g., '

Introd.
facs. 142), although the hurry or carelessness of a scribe
might occasionalh' almost efface it. •'

94. The distinguishing features were most apt to disappear

in ligatures, and this is true also of Continental minuscule
(see above, s. v. 'P' on the ligatures rp and sp\ and below,
^-
s. v. S '). Anyone unfamiliar with the Corbie ab-tvpe is apt
'

to mistake its r for And \^et it realh' keeps the two letters
.i\

quite distinct. While its /' falls below the line, its (except j-

in ligatures) stands on the line (see Introd. facs. 128). On'' ,

"

the resemblance of Insular r to a very open form of p see

above, s. v. P of Continental ;' to a small form of .y. see
' '
;

"^
below, s. V. S '). A Merovingian type of r (see Introd.
' " '

facs. 126, line 2 necessaria ') is sometimes suggestive of

'
^ or.

95. Of the ligatures (or connexions) of r with a following

letter two have alread}- been mentioned, ri (§ 56) and ro
(§ 81). Not until the close of our period does re (with the
'
angular variety of r disappear from Caroline minuscule.
'

The connexion rg belongs, as a rule, to an earlier stage

(e. g. occasionally in the Maurdramnus Bible of 772-780), but

survives in Beneventan script (e. g. Vat. lat. 3313, a 9.*^ century

(') That the Insular script of German scriptoriums made the confusion
easy is a statement (by Schnetz '
Neue Untersuchungen zu Valerius Maximus ',
p. 6) for which I should like to see the proof.
 44 W. M. Lindsay

MS. of Benevento) : also in Paris 13373 (Corbie, of 817-835).

On rt see s. v. ^
T
on rii, s. v. U
'
: on r/, s. v. '
P '; ' '.

X 96. Of the ligatures of with a preceding letter the most ;'

important is or, in which the uncial form of r is generally

found wanting the stem. In line 7 of Introd. facs. 152 an '
,

'
example will be found with a less usual extension of the leg '

of the r (R) under the line. The ligature (like the rarer iir,pr)
was an inheritance from uncial script (cf facs. 89, line 4),
although in the latter part of our period some scribes seem
to treat the r-portion un the form of our numeral 2 in small
\ size) as if it were a separate letter (cf. Introd. facs. 154, ' '

line 8 for a poor example). And after our period this new
2-form of ;- rtippears after e as well as after o (e. g. in the
Welsh minuscule Corpus Martianus Capella; although the
c>;'-ligature is rather a Continental than an Insular usage).
On tr see below, s, v. '
T '.

S.

97. Halfwa}- between the '

uncial form (our s) and the
'

'
half-uncial '
form (PI. I 83) stands what may be called the
cursive form (PI. I 66). Apart from the union os (see § 82) ' '

and the like, this last form hardly appears, except occasion-
ali}' at the end of a line, in the minuscule of our period. It

. usually stands partly above and partly below the line.

98. In the preceding section we
found a distinction in
early (Continental) usage between '
short-stemmed r (standing '

on the line) and '

long-stemmed '
r falling below the line), the
latter being reserved for ligatures (or connexions) of r with
a following letter. Something of the .same kind applies to s.
(Contrast in Introd. facs. 129 the
'
in the ligature st with
'
.v

the form of the unconnected letter). But many (or rather

most) Continental scribes came to write normally on the .y

line and not allow it to fall (to any perceptible extent) below.
(Statistics of the divergence of practice have not yet been
collected). While therefore in Insular minuscule both r and
s fall below the line and are to a certain degree confusible
(see § 93), in this Continental minuscule both stand on the
line and is usually a higher letter than r. Even a short s in
i"

this Continental minuscule (e. g. Visigothic s) does not resemble

 5

The Letters in Early Latin Minuscule J

r so much as it resembles a form (undotted) of r (see be-

low, s. V.)-

99. We have seen the reasons which impelled Insular

minuscule scribes to use the uncial variet}' of r (i; 93). The ' "

same scribes make great use of the uncial form of j-. pro- ' '

babl}' with the view of avoiding the confusion of s and r, and

sometimes utilize the variet}- to express .tì- (usually in the
order, half-uncial followed b\' uncial). In the rude imitation
of Insular half-uncial which we find in the Gatien Gospel
(written at Tours) I noticed only the uncial form of ' '
.t.

100. '
Subscript '
.s' (the '
half-uncial "
form) is a peculiarity I

of Insular script. In found literally

the Berne Horace it is

subscript, but most scribes write, e. g., 7is with the left extremity
of the attached to the right extremity of the n. When
.Î

'
subscript is attached to
"
subscript / (e. g. m's), the hanging
-f
' '

group spoils the regularit}- of the script in a wa}' that no

Caroline minuscule scribe would tolerate. On split-backed
'
'

5 see above, s. v. F ' '.

^ '

101. The curious expression o( ss seen in line 2 of Introd. '

fase 114 (' potuissent ") deserves mention as a peculiarity (a

cursive usage) of North Italian scribes. It is generali}- con- /
fusible with ns.
102. The ligature of 7i (in the form N) and is almost .c

as common (especially in majuscule script) as that of N and T.

The usuali}' takes the uncial or the cursive form, but
.<r
'
'
' '

is sometimes reduced to a mere curve (the upper part of s)

projecting upwards from the upper right extremity- of the K

(similarly 7is which resembles is similarlv os, etc.). The ' '
;

rare ligature sc with the lower half of c hanging from the

centre of the branch of Insular s) in the Northumbrian minus-
'
"

cule of Cambridge Trin. 216 (written b}- Bede ?) might easiU"

be confused with k. A more prevalent feature of earl\- Insular
minuscule is the union as (especial!}' in the ending bus
' ' ' ",

with '
suprascript '
u). On the '
knob-backed '
s (the '
half-uncial
form, standing on the line or at least barelv projecting below)
which is a characteristic feature of that calligraphic minuscule
practised in the Corbie scriptorium in the abbac}' of Maur-
dramnus (772-780), see above, s. v. *
F '.
 46 W. M. Lindsay •

T.

103. The short cross-beam, often barel}- perceptible (a

mere hair-line) which makes / in Rustic Capitals and early
Uncials often confusible with the letters i and /, is unknown
to Minuscules (and Half-uncials). The Insular minuscule script
of some centres (e. o-. F"ulda) frequently goes to the other
extreme and prolono"s unduly the cross-beam (sometimes ta-
pering and with a slight uj)ward slant; sometimes stretching
over the preceding as over the following letter, so that atii
with open a exhibits five uprights crowned with one horizontal
stroke). And when the maiuscule size of the letter (' high f) '

is admitted into minuscule script whether for the sake oi

varietA' or to save space at or near the end of a line or in
I

the ligature of N with T. e. g. Introd facs. 153) the cross-beam,

' "

^
whether straight or sinuous, is never stinied in length.
**
. 104. The Insular / (PI. I 87) most
is recognized also b\'
Continental scribes, although they often make the left end of
the cross-beam curve downwards. This curve ma}- be con-
tinued until it meets the stem and becomes a (more or less
circular loop. Looped / is the Visigothic form confusible
•
'

'

with a. but not with Visigothic a, which is open see Introd. ' '
;
'

facs. 116). It became the Beneventan form too; but hardly

until after our period, throughout which the curved form ' '

.seems to prevail. The two forms compete in Introd. facs. ' '

taken from a Monte Cassino MS. of 874-893 (not " 812").

120,
They compete
also in Murbach minuscule (see facs. 129, taken
^^from a Murbach MS.), in the Corbie en-type, in Burgundian
minuscule, and so on. In an Echternach MS., Paris 9530 (of
'
" 8-9 cent. " one scribe uses the looped another the Insular ' ", '

variety. Looped t ma}- invade even the Insular script of

' '

some Continental centres; witness its occasional appearance

in Paris 9525 ('' end 8 cent. "), written at Echternach.

\^ 105. Not looped but curved / is the usual competitor

' "
' '

with Insular t in St Gall minuscule. Winithar affects the

'curved' fornì. In St Gall 126 one scribe uses the 'curved'
form, another the Insular But the statement (by Glogger
' '.

Das Leidener Glossar ', part i, p. 5) that this form with a

'

pronounced curve (like a small c) is '' especially common at

St Gall" is a dangerous guide; for, as we have .seen, the
 The Letters in Early Latin Minuscule 47

'
curved form seems commoner than the looped in the
' ' '

earlier stages of Beneventan minuscule and appears even in

an earlv specimen of Visigfothic. Autun 27, not to mention
its frequent occurrences elsewhere (e. g. in the Corbie Leutchar

tvpe). An upward curved right end of the cross-beam is affected

for t at the end of a line bv various Continental scribes which
makes the letter exacth' like the symbol of tnr (and. after
c- 815. we tind in a MS., e. g., amatur {-tus) for
tiis). If

amat, we have some ground for guessing that the word stood
at the end of the exemplar. line in the
106. A g has been mentioned in which
variety ot Insular •

the right half of the cross-beam is wanting, so that g takes

some resemblance to z. We find the same variety of / es pe-
ciallv, but by no means only, in Insular script; e. g. it is a
prominent feature of the Lorsch MS.. Vat. Pal. 195. The
angle sometimes becomes a cun^e in the Anglosaxon minuscule
of a MS. in the same collection. Vat. Pal. 235: and on the
minuscule leaf (fol. 116) of the Book of Durrow we actually
find / written like reversed S (in the word transgressor '). '

Another Insular affectation, which makes t resemble ti (with

'
subscript /) has alread}' been mentioned (§ ^). In Insular
"

script is winged A in which the cross-beam takes high curves

' '

on the right and left, like the wings of a bird preparing to

fly (PI. I 67). It ma\- be the origin of that t b}' which we
detect the hand of a Constance scribe (PI. I 68).
107. Of forms whereb\- space is saved (at or near the end
of a line, to secure correct syllable-division) the high form "
'

has already been mentioned (and the ligature of X and T

along with U and T, e. g. facs. %, line 3 end). \ isigothic
scribes prefer *
usage not unknown to other
suprascript '
/ (a
scribes\ and when / the t stands upon
high form of / is "
'

reproduced. In Msigothic script this denotes it. Insular scribes

prefer subscript /, especialh- in ;//. but also in it (and even
* "

nit, nit, 7iti, 2iti, all with '

subscript "
/).
^
108. Another variety of this Protean letter, but only found
in ligatures, characterizes North Italian (cursive) script. We
may call recumbent /,
it '
'
for the letter, instead of standing
upright, leans on its side, e. g. ct (§ 17). A
mistake like r//w.
for st should suggest an early North Italian exemplar, like
48 W. M. Lindsay

Wolfenbüttel VVeissenburg. 64 or Milan L 99 sup. (both written

at Bobbio) in which this recumbent t is, to all appearance,
' '

reduced to a mere cross-stroke down through the (extended)

arm of the (PI. I 69) and exactly- resembles the common
.9

abbreviation-symbol of rtim. (Contrast the Beneventan st of

'Introd' facs. 120, line 1).

109. A like the North Italian

Visigothic variet\- which,
'
recumbent proper (but not wholly confined) to ligatu-
'
t, is

res we ma}" call circular t (PI. I 70). It resembles Greek

' '

minuscule Alpha and sometimes invites confusion with o. In

other Continental this circular letter denotes c in ligatures, but
in Visigothic t. Thus in Introd. fase 114, line 6 the ce of '
'

'
particeps is so written by a North Italian. A Spanish trans-
'

criber would interpret it as te, although he often writes te, etc.

with an apparent double circle (PI. I 71), the loose ends of ' '

the /\lpha being continued towards the e. A group like tto

' '

presents an extraordinär}' appearance (like four successive

o's) when written in this fashion.
110. The ligatures of this Protean letter (more multiform
than e itself; § 31) are of importance for the palaeographer.
In '
ancient minuscule '
(i. e- cursive pressed into service as
book-hand in scholia, marginal indexes, ete.) / is often repre-
sented by a mere horizontal stroke. Por example, in the mar-
ginalia of the Bologna Lactantius the t of ratio looks like '
'

a mere hyphen between the open a and the longa ' ' '
i
'.

This ancient usage (frequent in North Italian cursive and el-

sewhere) is conserved b}' Insular scribes (mainly in the liga-
ture et). In the Anglosaxon half-uncial of Durham A II 17 we
find it at the ends of lines ; it, at, nt, etc. are there expressed
by z, a, n, etc. followed b}- a long horizontal stroke. (Simi-
larly in the Lindisfarne and IMaihingen Gospels.) Since the
same scribe affects at line-endings a lengthening of the '
ton-
gue '
of ^ (the 'arm' ofr, etc), a transcriber would be prone
to mistake his e so written for et, and errors like faceret for
feuere would arise. In the Boniface Gospels '
dicunt "
(on fol. 9')

is written as dicwi followed by a horizontal stroke- The In-

sular ligature shews this horizontal stroke accompanied be-

et

neath by the lower curve of t (see Introd. facs. 135 136 ' '
; ;

137 column ii, line 1), the horizontal stroke being often sinuous
 The Letters in Early Latin Minuscule 49

(see facs. 142, line 4 expression (') of

: 143, line 15) : and this
et is the old expression found in the ancient minuscule of '
'

fhe Bologna Lactantius and Escurial Augustine marginalia and

the quarter-uncial of the Sessorian Augustine (Rome, Bibl.
•
'

\'itt. Eman., Sess. 55), and even (at the ends of lines) the
uncial of the Oxford Eusebius Chronicles.
111. Continental scribes write the ligature ditterenth- (like
PI. I our etc.
86, as in symbol', and one scribe of Munich
'
"

28118 (a MS. presented by Benedict of Aniane to the abbot

of Trêves and written probably at Treves, at the end of the
8.*'' century) employs both the Insular and the
Continental form
of the ligature. In this Continental et ligature the t-portion ^
often ends in a tag to the left, which prevents confusion (§ 133)
with the ec ligature (' Introd. facs. 120, line 2). ^ "

112. In ligatures like nt the analogous expression of / re-

sembles a sucker rising from the ground (see Introd. facs. '
"

126, line 11 'sunt': facs. 128, line 2 'student'), the ground

being represented by the extended base of the second upright
of the 71- This form of nt is unknown to Insular script. Simi-
lar, but confined to an earlier stage, is the at ligature (seen

in facs. 131. line 10).

113. Visigothic ligatures of ;//. at, etc. make the t confii-

sible with c (PI. I 72), mmtiiis resembling nuncius, natio re-

sembling nacio^ and so on. This c-form of the t-stroke appears
as early as the marginalia of the Escoriai Augustine, e. g.
'
atende (like acende"), along with the h}'phen "-form, e- g.
"
' '

'
gratias (in the subscription
'
deo gratias \ with a hyphen '- ' '

stroke between open a and longa

*
It ma\' be a deve-
'
*
i
".

lopement of the hyphen-form, the h}phen

"
being added "
'

without lifting the pen. If so, the body of the c-mark is cor-
rectl}' described as a mere connecting line and does not pro-
perU" represent the stem of t. -^

114. In the Laon az-script, etc. the '

h\"phen"-mark in nt, it,

etc. becomes oblique instead of horizontal and is attached to

the middle of the n or / (PI I 73). It seems a half-wa\- stage

between the original h^-phen attached to the top of the let- " '

ter (as in the Insular half-uncial examples mentioned above)

(*) Less frequent is the variet)- (PI. I 85) found in the Anglosaxon script
of Durham A II 16 (half-uncial), London Cotton Tib. C ii and Vesp. B vi.
 50 W. M. Lindsay

and the subsequent sucker rising- upwards from the base ' '

of the letter. But until statistics have been collected (espe-

cially from early charters) the history of these ligatures must
remain doubtful. Certainly the Constance scribes reg'arded
the sucker as the equivalent of the stem and not of the
' '

cross-beam of the t. for the}^ crown it with that small half-cir-

cle mark which, as we have seen, the}' substitute for the cross-
beam of unlig-atured t (§ 106).
« 115. Some statistics have been collected for ligatures of ^
with a following letter (b}^ Loew '
Stud. Pal. "), but a fuller
listwould be welcome. The most important of these ligatu-
res is //. We have seen (s. v. I ') that the practice was to '

make the i an appendage of the /. This appendage usually

took the form of a long curve passing below the line. To
append this curve convenienth' to the cross-beam of the /,
the cross-beam was tilted upwards on the left and the whole
letter sloped to the right instead of remaining perpendicular.
The '
Insular form of / with i thus appended became as in
'

Pi. I 74 or (with the under curve of the / continued a little

further) as in PI. I 10 '
unigeni^z", 11 '
graZ/ae '. This is the
natural form of the ti ligature for all scribes who use Insular
/. It is confusible with q (' activam '
with '
ac quam '). Still
more with the Insular abbreviation symbol of qui {q with i ' '

above) when the long curve becomes a perpendicular line and

the / is shifted round more to the right. The curved and ' '

the perpendicular forms appear side b}' side in such MSS.

' "

as Vat. (Irish minuscule of Bobbio).

lat. 491
~~
Ilo. When
'looped' t was used the ligature became like
a reversed Greek minuscule Beta, either sloping to the left
or upright, and either with curved or perpendicular back.
The reversed Beta form is never used by the scribes who
' '

never used looped t, e. Insular scribes., whether in the

' '
i.

ligature ti or in the analogous formations of te, t?i, etc. (see

'
Introd. facs. 124, line 1. 'te', line 2 'tu', line 5 'ti'; 125,
'

>. line 1 ti te tu ").

'
Their rare appearance in the Insular
', ' ', '

script of Continental scriptoriums (e. g. te and ti occasionalh'

in the Anglosaxon minuscule of a Fulda MS., Cassel theol.
O. 6) stamps them as Continental intruders (*).
(') The resemblance of '
Insular "
t in ligatures (other than ti) to a (the
* uncial '
form ; § 1) is probably the reason why ti is the only ligature of this
 ^

The Letters in Earlv Latin Minuscule '

51
•
_ .

11/. The separation of the two loops of the "reversed

Beta an earl\- feature, e.
'
is g. // (as in PI. I 75), /;- in the
Bobbio MSS., Vienna 16 (of. ''
c. 700 '")
and Milan L 99 sup.
(" mid. 8 cent. '").

118. The ligature tri (from 'Insular' t) has in the An-

glosaxon script of Echternach the form seen in PI, I 76 A
peculiari}" Insular form of the //ligature, which with subscript't? '

e.xpresses tio (see PI. 1 14 resurrecZ/cmem ') and with sub- ' *

script i, tis (see PI. I 15

'
men/^V '), will be found in PI I 12 '

*
//muerunt 13 vestimenta It has
",
been anal}'sed above
' ".

54) into / with superimposed Insular cursive

'.
(iî; longa "
i

The formation of the tr ligature öfter this pattern \?\. I 77)

seems to be a malformation on false analogy ot the tis liga-
ture and should properly express tir rather than tr. An- * '

other malformation turns up here and there, the substitution

of the '
Continental '
cursive e (see Ji 28) for / in ligatures,
especially // (PI I 78). Transcribers would be justified in in-
terpreting this as ci. A third is the use (in earh^ Xorth Ita-
lian cursive : also in the Boniface marginalia of the Codex
Fuldensis and the in Paris 13367 and the '
ancient minuscule '

marginalia of the Douce Primasiusi of the ancient Nota o{ tio

(PI. I 79) to express from the commoner ti. Its difterence
form of the merely wants the
ti ligature is \er\' slight. It

projecting stroke (the left half of the cross-beam of the f) above

the loop, and such MSS. as Milan D. 268 inf. (Xorth Italian
cursive shew the ligature sometimes wanting this projection
but with pear-shaped, not circular loop. A fifth (hardh' a li-
gature) is that extraordinary expression of tio by what ought

to express toi (PI. I 80), which is confined to

"
'
ancient mi- '

nuscule (e. g. of Paris 12214-}- Petrograd Q. I 4, of. 6 cent).

"*"

119. The
reversed Beta ligatures are a criterion not onlv
' "

of Continental provenance, but of date. In Beneventan script

indeed the te, tu ligatures as wel as ti see below) survive -,

long after our period (see Loew Ben. -Scr. p. 148 for details), '
"

and also in Visigothic (e. g. tr in Madrid Bibl. Acad. Hist. 25,

form favoured bj' the scribes of our Islands, and why tr was expressed by
another form (^see below). However te (resembling ae\ appears in the Book
of Armagh and is not rare in some Continental centres of Insular minuscule,
e. g. Echternach.
 52 \V. M. Lindsay

of 946 ; Madrid Toi. 15, 12, of 915 to, ti\ ti in iMa-

/r, te in ;

drid Toi. 10, 25, of 902). But outside of Ital}- and Spain they
X^hardl}' occur after Charlemagne's time (e. g. ti, te in a Salz-
burg MS., Munich 210, of 818; ti in a Freising MS.. Munich
6262, of. 854-875, in which however the misuse of the ei-liga-
ture for ti is also in vogue § 118). ;

/^ 120. Another useful criterion is the st ligature (see In- '

trod. '
facs. 143, line 9 '
astra "), a ligature found occasionali}'
even in the half-uncial of the Basilican Hilary. It is a Conti-
nental, not an Insular usage, and its occurrence in Insular
script is proof of provenance from some Continental scripto-
rium where Insular script was practised. A scribe of our Is-
lands would write st as it appears in line 13 of the facsimile
]ust mentioned (' baptista '), with the arm of s never actually
touching the top of t. The distinction is a fine one, and it
stands to reason that the idios\'ncrasies of individual scribes
or the hurry of writing would be likel)' on occasion to cross
the forbidden line. So that the rule is not devoid of excep-
tions. Line 9 of this facsimile furnishes one for the plate co- ;

mes from a MS. written in Mercia. The jBook of Cerne too

uses the ligature. It was the jira\er-book of Bp. Ethehvald
of Lindisfarne (of Lichfield T)- So does London Royal 2 A xx.
So do the Northumbrian MSS-, Durham A II IC (although one
of the scribes avoids it), Durham B II 30, London Cotton Tib.
A xiv and Tib. C ii, Cambridge Trin. 216 (" de manu Baedae ").
On the other hand some foreign specimens of Insular script are
free from it e- g. Metz 76 Prophetae (in the part written bythe first
:

scribe for the second scribe allows it) Munich 6237 (written
; ;

at Freising b\' Peregrinus, c. 780 Colmar 38, part ii (written

;

at Murbach, " 8 cent. "). These three examples are all Anglo-
saxon. In spite of these exceptions we must regard the pre-
sence of the st ligature in Insular script (e. g. in St Gall 51
Gospels, in Irish half-uncial) as fair evidence of Continental
provenance. Often it is difficult to say whether there is actual-
ly a ligature or merel}' close proximity of the two letters with
the stem of the t immediatel)" under the end of the branch
of the s.

-^_ 121. The ct ligature has been described above (§ 17;. It

too is Continental, not Insular (a Continental intrusion in a

rather cursive part of the Anglosaxon minuscule of a Werden
 The Letters in Early Latin Minuscule 53

MS., Berlin th. O. 139, foil 61^-63^). So is the rt ligature

which is usually clearly distinguishable, with its angular shoul-
der (cf. Introd. facs. 154, line 7 participes "). from the st li-
' ' *

gature (cf line 6 abstinentes '), although the confusion of the

'

two b\- transcribers (often late transcribers to whom fhe rt li

...
gature was unfamiliar) is by no means rare.
/

122. The letters ti before a vowel (e. g. in the word na^

tio) had in Late and Medieval Latin an assibilated sound (un-

less s preceded, e. g. quaestio), which they had not in words

'

like noti or natis. For the convenience of readers (of the

lessons in church or of some pious book in the refectory-) iso
lated attempts were made b}' scribes of our period to distin
guish the two sounds b\' the writing. The Laon az-script nor-
mali}' (but not rigorously) reserves for the unassibilated sound
the reversed Beta ligature (for details see Rev. Bibl 24,
' '

22). North Italian scribes often reserve the same ligature for
the assibilated sound, and this became the strict rule in Be-
neventan script (hardl}- in the earliest stages for details see ;

Loew Ben. Scr. p. 303). In some specimens of the Corbie

' '

'

ab-script the assibilated sound is expressed by / with longa '

i

(in the j-form see § 55) but never very consistently (for de-
;

tails see Rev. Bibl. 24, 24), and this became about a hundred

years after our period the Visigothic practice (for details see
Loew Stud. Pal. where it is stated that Prankish scribes
' ',

as a rule preferred to express the assibilated sound b}' ci, e /

g. iiacio, gracia, rado).

U.

123. The danger ot confusion between and (open) a ii

caused a, as we have seen (s. v.), to take certain forms. It

operated no less in the formation of u. The projection of they

right side of ii below the line avoided this danger. That is
'

an early form found, for example, in the ancient minuscule '

of the (contemporary' r) marginalia of the uncial Bologna Lac-

tantius and in the earliest examples of Insular minuscule, e.
g. St John's Gospel in the Book of Mulling before 696), the
Naples Charisius (written at Bobbio c. 700 ") and even la- ''
;

ter Insular scribes often affect it. When St Moling writes

'
quis the q (in open form) differs from the u mainh' in the
',

greater length of the projection below the line. Spanish seri-

 54 \V. M. Lindsay

bes often write with this projection so exaggerated

initial U
that the letter might be mistaken for (open) Q. (Similarl>- in
a LNons MS. of Leidrad's time, L}'ons 608). xA.nd outside of
Spain, especiall}' in uncial and half-uncial script, the ligature
ni (see above, s. v. I ') is often
'
indistinguishable from q (in
that open iorm in which, for example, Winithar, the St Gall
scribe of c. 750, loves to write (/).

r 124. The
was another preventive of confusion
v-form of 7i

with a. It is greatl}- affected b}' Welsh vand Cornish) scribes

(see m\- Wei. vScr. for details), but is by no means rare in
' '

other script (e. g. it is common in the Anglosaxon script of

'

Vat. Pal. 554, even vi with subscript i and the ending orum ,•
'

in the Douai Gospels aormally shews this form). It was in

its turn contusible with an occasional type of}' (see below,

s. Indeed a 3-torm ot u is sometimes found, with the right-

v.).

hand stroke continued far below the line (as in the Veronese
cursive of Verona 33 Augustine de Agone Christiano, in the
ending rum on fol. 41^ etc).
'
When the v-form was em-
'

ploN'ed in the ligature ru see Introd. facs. 131, line 4 la- ^ "
"

"

crimarum "), it somewhat resembled /, and the ending rum '

y might be miscopied rim ' '.

125. The letter u is, we ma}* sa}-, indistinguishable from

i in the ligatures us, wi, ut. etc-, ligatures much used in un-
cial and half-uncial script as well as minuscule), especially (at
least in the older or more calligraphic specimens) at the end
of a line- For an example of the ligature '
ut '
see Introd.
facs. 98, line 4 (from the half-uncial Basilican Hilary-I It was
at the a line that calligraphers welcomed this means
end of
of saving space and enabling the word (or s}-llable to be com-
y
pleted and not left to straggle into the next line.
^ 126. But the commonest device
for econom}' of space was
to write the u above the line, and one can hardl}' turn over
the pages of even an}' uncial or half-uncial MS. without find-
ing 'qui', 'quo', 'quod', etc ^especiall}', but b}' no means
onl}-, at the ends of the lines) written with suprascript «, the,
most frequenti}' suprascript letter in all Latin script. The
'

Hilary facsimile just mentioned (' Introd. facs. 98) offers aput "
'

so written at the end of line 6 quam near the end of line ;

' '

9 qui
;
'of falsiloquium ') and
'
quod in the middle of
' '
'

\ line 10. •
 The Letters in Early Latin Minuscule 55

127. Suprascript îi takes so many forms as to def\- accu-

rate classikcation. We
may ho\ve\er specif}- the followlntî :

(1) a regular v-form, like the sides ot a small isosceles

triangle.
(2 a regular rounded form, like the lower half of a
small circle or oval. This is precisel}' the mark used in gram-
matical writings to distinguish a short from a long vowel. In
such writings there is danger of confusion. Thus one MS. of
Marius Victorinus substitutes /mie, for /nc (i. e. hie with short
/), the scribe having mistaken the mark of shortness for a su-

prascript 21 (see Keil '

Gram. lat.
'
5. 208, 9) ; another substi- /
tutes ruovia for róma (a mispronunciation ot '
Roma "
. Keil
(ibid. 5, 285, 7)has infortunateh' admitted this corruption Ruoma
into his text, from which it has found its way into works on
the Romance languages {see Amer. Journ. Phil. 37, 34).
"^

(3) various irregular v-lbrms, with one or other of the

sides exaggerated in length and curved rather than straight-
The Hilary facsimile (' Introd. facs. 98 offers one of these. *

Spanish scribes attect a form which faces to the left (quite

unlike suprascript a) and which is occasionali}' replaced b\' är-
mere curve like our apostrophe.
(4) various irregular rounded lower half-circle) forms.
Sometimes, the curve is ver}- shallow and broad (as in the
half-uncial of Verona 59, where it occasionalh' approaches to
a mere horizontal line) sometimes the right side shoots up
;

in a hair-line, a favourite torm with Insular scribes. ^

(5) a horizontal stroke beginning in a slight curve, an
upward curve on the left, a downward curve on the right.
This ma}' be a developement of (3) or (4), and the downward
curve ma}' be reali}' nothing but the connexion with a follow-
ing letter. It is exacth' identical with the suprascript sim-
bol for rn used b}' Insular (^and other) scribes (e. g. '
Introd. '
'

facs. 138, line 16 tantuw Thus in the half-uncial Verona

'
"
.

42, on fol. end of a line, quam is written with

148'" at the * '

this as an expression of u (above the a and of m (after the

a . We find this suprascript 2c as earh' as the ancient mi- *

nuscule marginalia of the Bembine Terence and of the Escu-

"

rial Augustine on Baptism. This suprascript m-stroke is often

used as a mere abbreviation-stroke, and in such MSS. as Pe-
trograd F. I 5 Tripertite Psalter (in the Leutchar script of
Corbie) it pla}'s all three parts of u. m, abbreviation-stroke.
 56 W. .M. Lindsay
^
(6) a form resembling (5), but with the dowmvard curve
on the left, the upward on the right possibl}' derived from ;

an irregular v-form like that seen in Introd. facs. 83 (from, ' '

the Herculanean papyrus of a poem on the battle of Actium).

This mark too may play in MSS. all the three parts of ?^,
tn, abbreviation-stroke.
\ (7) a sickle-torm, which is often indistinguishable from
C suprascript a). Sometimes it is so narrowed end carelessl}'
written that it would resemble (5) if (5) were written verti-
cali}' and not horizontally. In the " N. E. France
minuscule "

the '
sickle '
often loses its handle and becomes a mere c in
shape. ^This c denotes w in Burgundian minuscule, etc., and
is by some scribes used as an abbreviation-stroke ; for some
details see Rev. Bibl. 22, 411). The Laon az-type, which sub-
stitutes angles for the curves of a (see above, s. v. A "), turns '

the curve of the sickle into something more like a square (or
' '

oblong), open on the right.

^ 128. That this expression of /^ in regular sickle form or ' '

in irregular form (something like a narrow s with the lower

cur\^e incomplete) was proper to suprascript ti cannot be as-
serted. For in earl}' (cursive) script of various countries (not
of our Islands) denotes u on the line.
it An example (from
a Merovingian book-hand) will be found in Introd. facs. 126, ' '

>^ with the more careful sickle-form in the last line (' sunt '). the
less careful s-form in line 4 ( '-rum '). This cursive 7i is a con-
stant feature of the Merovingian book-hand of London Harl.
5041 ; Berne 611 appears in the cursive mar-
: St Gall 214 ; it

ginalia of Paris 10910 Fredegarius (of the year 678) and of

Paris 13367 Augustine in the Luxeuil type of London Add. ;

11878; in the Spanish cursive passages of Escurial R II 18;

in the Veronese cursive of Verona 33 Augustine de Agone
Christiano and occasionallv in other earl}" MSS., e. g. Berlin
;

Phill. 1743 Concilia from Rheims), and even in the ab-type

of Donaueschingen 18 Orosius.

^
129. Suprascript it is of far more frequent occurrence, but
always confined to the earlier MSS., a cursive feature which
would not be favoured by 9.*^ century scribes (though we find
it, e. g., in Paris 11738, of c. ,840, in 'quo', etc.). A Spanish
example is Verona 89 Breviarium Mozarabicum an Italian. ;

Vat. lat. 5763 Isidore's Et3'mologies (written at Bobbio, " beg.

 .

The Letters in Early Latin Minuscule 57

8 cent. "), It is reg'ularly used in such Merovingian types as

the Luxeuil t^pe (e. g. Paris 9427 the Luxeuil Lectionary),
""
the Laon az-type (see abovej, the '' N. E. France t}pe (see
above) ; not to mention the less cursive Corbie en-t}pe. It
is known to Winithar, the St Gall scribe of c 750 (e. g. St
Gall 907) to Thomas, the scribe of the Treves Gospels to
; :

more than one scribe of Epinal 68 (of the \ear 744), and so
on. In an 8.'''
cent " MS. from Fleury, Paris nouv. acq. 1597
''

Pate ri us, I noticed it in '

qui '
(written qi with this s-mark over
the right side of the That expression of qui so com-
q).
' '

mon in MS.S. of St Gall and Bobbio, etc., the letter g follow-

ed by a long, narrow s-mark which often begins high above
the q and is prolonged down to the line, seems to exhibit this
cursive ?i combined with subscript i. (For a less probable
explanation see Zeitschr. celt. Phil. 9, 305). This expression
o:' ?ci is often indistinguishable from the expression of ?/, but
is rarely found except after the letter q.
^
130. After the letter q however the favourite expression
of saprascript 7i before an\' other vowel, and even sometimes
before i. is in all Merovingian script.
^
an arch, bridging the interval between \q and the
(8)
following letter (see Rev. Bibl. 24, 21). We find que qua ' ', ' ',

'
c^uoetc regularl}- so written in the Luxeuil t3'pe, the Laon
,

az-t3^pe (and its predecessor, the a-t3pe of Petrograd O. I 14

Gregory's Homilies, from Corbie). Similari}- in the earl}' Bob-
bio script of Milan D 268 inf. (e. g. que ') and S 45 sup. e '

g. qua '), in the early Vercelli 183 (e. g.

'
quod fol. 205 ' "

with the arch not connected with the q), in Munich Univ. 4.*"
3 Sulpicius Severus (e. g. qua ') with the arch not connected
'

with the q). An example from the Laon az-t}'pe will be seen
in Introd.
'
facs. 125, line 3 'qua', line 4 and line 5
'
f}uod '
'.

Except (1), and (2), and perhaps (3) and (4), all the û
131.
above-mentioned types of suprascript u are capable of pla\'-
ing the part of a mere abbreviation-stroke (and also of an
expression of m). It remains to specie' a form in which the
identity of the 7i seems entirel}- lost,
p
(9 a horizontal stroke, preciseh' the same as the usual
abbreviation-stroke and a common symbol of m). It seems

unwise to attempt to refer this to an exceptionall}- careless

manner of writing zi. Rather the conventional expression of
 58 W. M. Lindsay

suprascript u had come to be regarded as the same thing as

the abbreviation-stroke and the S}'mbol of w. In Cambridge
Corp. 304 luvencus (of unknown provenance) the abbreviation-
stroke is a horizontal straight line with a barb on the left and
a dot above the centre- The same plays the part of supra-
script 71 (e. g. 'quid' fol. 107") and of the symbol of ;;/. (Fur-
ther details in m\- Not. Lat. p. 358, where it is shewn how
' "

confusion of. e. g., aüm '

avum '
and aüm '
autem '
must have
resulted from the practice).

X.
'^
132. The X of uncial (' Introd. '
facs. 91, line 3) and half-
uncial script (facs. 100, line 3) consists of a thick stroke down
from left and a thin stroke down from right to left
to right
across (The thin stroke is often prolonged below the line).
it.

Departures from this normal type in the minuscule of our

period are usuali}' capricious and do not offer safe clues to
provenance. In Insular script the continuit}- of the thin cross-
stroke is often abandoned for the form seen in facs. 136, line 8 ;

but the same thing is found in some other scripts also, e. g.

Msigothic. The Anglosaxon script of Fulda loves to prolong
the thin cross-stroke away below the line and to end it with
\ a return-stroke (or a loop) half as long as the projection, but
the same trick appears also in, e. g., the Anglosaxon minuscule
of a St Bertin ÂIS. Boulogne 63-64 (For this Insular x with .

'
return-stroke '
from the Le}'den Priscian).
see PI. I 19 '
texunt ',

Two Werden MSS., now at Berlin (th. F 356 and 366, both

in Anglosaxon script) end the thin cross-line below with a large

claw or comma-mark, but this appendage too is found occa-
sionally in Anglosaxon minuscule of other provenance. (PI. 1 19

shews an example from the Irish minuscule of the Le3'den

Priscian). Some Continental minuscule scribes begin the thick
stroke at the top with a curl which might be mistaken for
an 0, hardly evidence enough for calling this freak
but there is

by any less vague name than Merovingian (e. g. sometimes ' '

in the Laon az-type, in the Luxeuil t}pe, in Cassel theol.

Q. 10, etc.). The strange high x which appears now and
'
'

then in our unique specimen of Cornish cursive, Berne 671, is

perhaps worthy of mention (see PI. I 16 lux 17 dixi 18 ' ', ' '

*
sexta ').
 .

The Letters in Early Latin Minuscule 59

133. The ligature ex (facs. 126, line 5). with the thin cross-"^
stroke of the x continued below the line, was hardh" confusible
with the ligatures tV (facs. 127, line 2) or ec (see § 111) bv
transcribers who were familiar with the trio. It was when
one or other of the three (first the ec ligature, then the ex)
went out of use that the real danger of confusion arose. And
when ec went out of fashion and the need of keeping the e^
ligature different from ee ceased, e/ came to be written exacth'
as ee had been written in our period. Indeed there are exam-
ples of the writing of e/ dangerously- like ee before then. /

Y.

134. One method of forming '

in uncial script will be seen
in '
Introd. '
facs. 90, col. ii, line 3. To the thick angular bodv-
stroke is added on the right, opposite the angle, a thin stroke
ending in a knob. In half-uncial and minuscule the bodv-
stroke was usuali}- curved (see facs. 136. line 4) instead of
angular (and often in uncial too), so that the letter bore some
resemblance to half-uncial and minuscule .n-. We may call this
the s t}-pe. Another we ma}- call the Y-t}'pe, for it resembles
our capital letter. When the Y-t}-pe was reduced in size (in
breadth rather than in length) and the v-portion rested on the
line, we get the }--type, resembling our ordinar\- letter } (and
often, like it, ending in an upward curl or dot). When the
lower strokes of the Y-and y-t}-pes are removed or reduced
to a minimum, we get the \'- and v-types. These five t\-pes
are really ten. for each ma}- be dotted or undotted. The dot
between the two tops of the letter was taken by Roman script
from Greek. The minuscule writers of our period seem to
insert or omit it at will. Just as the}- do with the dot accom-
panving an abbreviation-svmbol (e. g. sometimes id.ë and
sometimes idê for id est '). '

135. A peculiarly Insular variet}- seems the result of con- (s

necting the dot with the top (see PI I 20 abyssum ').

left '

For this is a thing Insular scribes are prone to do. The tail '

(or reversed apostrophe) of the h-mark which is the Insular

abbreviation-s}-mbol of autem the similar tail which the}-
' '. ' '

attach to the letter/ to symbolize per the apostrophe which ' ",

ioins the cross-beam of ^

on the right to s}-mbolize igitur ' ",

all these seem the result of attaching the dot to /i and / and £'
 60 W. M. Lindsay

without lifting the pen (see

p. 67). Sometimes my '
Not. Lat. '

(e. two Bobbio MSS. in Irish script. Vat. lat. 491 and
g. in
Florence Ashburnham. 60) the curves of this Insular r become ' '

angles and the letter gets the F-form and might be confused
with f. In Continental script Insular y is a feature of the ' '

Corbie scriptorium, especialh' of the Leutchar-type, but also

of the (earlier?) en-type, etc. The dot is patently joined b\-
a hair-line to the left top in such Corbie MSS. as Petrograd
F. I 5 and 6 (both Leutchar-t3'pe) and F. VI 3.
D 136. A peculiarly V^isigothic variet}' is high \--tvpe
the ' '

(not the Y-type). It is often so shaped b}- Spanish scribes as

if thev meant to suggest ui^ for a mark like suprascript u
(see § 127) seems to cap an i longa And it is difficult to ' '.

resist the suspicion that they thought of their longa and '
i
'

'
high V as related forms.
'
In writing ait Cain A- ' ', ' ', '

chaia \ etc. they use now i longa now high y ', apparently ' ', '

at caprice. Editors are (possibh' with reason) reluctant to

print ayt '
Ca\n Achaya although they print martyr ',
', ' '. ' ', '

etc., where precisely- the same letter appears.

\i 137. Apart from Insular and Visigothic y we get very
' ' ' '

vague clues to the provenance of a MS. of our period from

the way in which y is writter. Some MSS. offer a variet}- of
tvpes, e. g. a Compiègne .VIS,, Paris 17451, has seven more
or less recognizable formations. Perhaps we may say that a
y-t3-pe is likeU' to be found in MSS. of Italy. Switzerland and
Bavaria, and that the s-type. especially in small size, has its
home in France. The great majority of iM.SS. offer one or
other (or both) of this pair, the >'-type and the s-type and
both are as often dotted as undotted. The dot would save
the latter from the danger of confusion with j-, for the
small-sized s-type is ver}' like a small-sized e. g. Visi- .<•,

gothic Perhaps the likeness had something to do with

.y.

the choice of high y by S{)anish scribes. Another way of

' '

escaping the danger was to exaggerate the slope of the 3'^-form

(towards the right). This produced what we may call re- '

cumbent y which seems to lie on its side, resting on the line

'

and only slightl\- projected below. The recumbent form is ' '

known to Insular as well as Continental scribes. A Rheims

MS., Berlin Phill. 1743. offers a 'recumbent' v-form (small-
 The Letters in Earlj- Latin Minuscule 61

sized, with ven- obtuse angle, and dotted. On an occasional

form of 11 resemblinor; see ^ 124.

Z.

138. Insular minuscule (and half-uncial) is wonderfully con-

sistent in projecting the lower angle in a longish spear or horn
(see Introd.
'
facs. 141, lineo).
'
This prevents confusion with
a common type of Insular / and with the z-variet\" of Insular^
(see above, § 45). The upper angle was projected in a longish
horn in the Laon az-script (PI. I 81). Another local variet)-
worth}- of mention appears sometimes in North Italian mi-
nuscule with the ti4igature (see ?i 115), in one or other of its
forms, used as the upper part of the z (as PI. I 82, in Ver-
celli 183 Jeromes de Vlris Illustribus.
**
8 cent. " as PI. I 83. ;

in Verona 42 Gregory's Pastoralis Cura and in Milan Ambr.

C 98 inf. Maximus' Homilies). This is a cursive variet}% found
in earl}- North Italian documents (Loew '
Stud. Pal. '
p. 27).
The Visigothic variety which resembles Insular g has alread}-
been mentioned (§ 38). And it may be worth while to call
attention to unique form (somewhat like Y) found in the
a
Bobbio Missal, Paris 13246 (of unknown provenance), since
Delisle in his account of this MS. mistook it for c in the word
'
Lazarus (fol. %%
'

For the other varieties, large-sized or small-sized, sloping

or upright, with top- and bottom-strokes straight or curved, it
is difficult (in the absence of statistics) to find among them

any satisfactor}- tests of place or time. The high form with

the top deeply curved (like the lower half of an oval) might
easil}- be mistaken for an (with suprascript a\ § 4).
62 P. Liebaert

SOME EARLY SCRIPTS OF THE CORBIE SCRIPTORIUM

BY THE LATE P. LIEBAERT.

In the obituary notice of Abbé Paul Liebaert in the

'
Classical Review '
he was declared likel}- to have become a
second Mabillon. That this declaration was not exagg-erated
I realized when
examined his large collection of photographs
I

(with negatives) of MSS. and his careful notes on a large

number of MSS. in a large number of European libraries.
These photographs and notes are now in the Vatican Librarv.
Of all Liebaert's palaeographical projects the one which
most interested me was a projected Historv of the Corbie
Scriptorium. Many a talk have I had with him on this fa-
scinating subject. So many that, although he left no written
record, can supply from memory, at least a portion of his
I

discoveries, thethree types of Corbie minuscule which pre-

ceded what was called bv Traube " the old script of Corbie ",
but is now called the ab-script of Corbie
'
ED.]. ".

(1) TJie en-type (Plate II).

The appearance of the three might be called

earliest in
'
the em-t\pe because a prominent feature is the form of the
'.

letter e in ligature with such letters as w, r. etc. But since

another prominent feature is its uncial ;/, it seems better to
call it by a combination of these two, the en-tvpe just as ' '
;

we call the type previously mentioned the ab-type because ' ',

one prominent feature in it is a curious form of a and another

is a curious form of h. Indeed since the eve is caught not
only by its c (in ligature) and its n (of uncial, not often mi-
nuscule shape), but also by its a. an open letter with thin
horns, we might describe it best as the ena-type but the
* '
;

mention of three letters is unwieldy, so that the name the '

en-type of Cobrie seems best. '

The plate is taken from fol. 8' of a MS. (lat. 13349) in

the Bibliothèque Nationale, Paris, containing Jerome 'in Ec-
clesiasten '.
The plate is in natural size, like all the plates in
this journal. The eye soon detects the peculiarities of this
early-looking type of script, a type which it is impossible to
mistake or to forget, once one has ac( juainted himself with the
 Some Early Scripts of the Corbie Scriptorium 63

three prominent features, the beetle-headed e in ligatures, the

uncial ;^. the open, horned a.

Another plate from this MS. will be found in Delisle's

'
Cabinet des Manuscrits (pi. xvi. 2) along' with others from
'

its neighbour volumes (with other works of Jerome), lat. 13347

pi. xiii, 2 and lat. 13348 (pi. xiii, 6-7). In Delisle's pi. xiii,
2 a curious form of j' will be noticed (in the word aegvpto in
the second line). The left branch of the r is curved round
at the top towards the right. This is an Insular (especiallv
Irish.^) form of the letter (see Zeitschrift für Celtische Philo-
logie, ix 307), and is a common feature of all types of earlv
Corbie minuscule. Since Corbie was a Luxeuil foundation, it
seems reasonable to find in its use at Corbie a trace of Irish
influence. Less likelv perhaps is the supposition that Insular
scribes and Corbie scribes, quite independently of each other,
hit upon this de\ace of adding the dot between the two branches
of J' without raising the pen (cf. Proceedings of the Rovai Irish
Academ}-, xxxiii 401 and for a similar treatment of the ori-
;

ginal dot in the Insular autem and per abbreviation-symbols,

' '
' "

'
Notae Latinae p. 180). ',

Other MSS. (all. like these three, from Corbie Library)

in the Corbie en-t}'pe are :

Paris 4403' Codex Theodosianus (plate xix, 10 in Cab. '

MSS. '
shews the close, not open, a).

Paris (ff. 1-52, 74 col. ii — end of MS.) Cassiodore on the

Psalms Cab. MSS. pi. xvii, 1).
('
'

Paris 13028 Isidore's Kt3-mologies xvi-xx (often with close à .

Paris 13347 (ff. 1-55) Jerome.

Paris 13348, 13^9 Jerome.
Amiens 220 Paterius (part).
Petrograd (). I 4 Cassian's Collations.

(2) The Leutchar-tvpe (pi. III).

This is the script of Petrograd F. I 6 Ambrose on St Luke,

abbacy of Leutchar (middle of eighth
written at Corbie in the
century) and corrected by a later abbot of Corbie, Paschasius
Ratbert (see Revue des Bibliothèques, xxii 406) also of its :

neighbour, F. I 5 Tripertite Psalter. Traube (Palaeogra-

phische Bemerkungen, p. 1) assigns the two to a Zwischenstufe '

zwischen gallischer Halbunciale und Minuskel a ver\- vague ",

64 P. Liebaert

term under which he includes a variet\- of quite distinct types.

Certainl}- the half-uncial look of the script, with its avoidance
of ligature, is striking, as well as some peculiarities of letter-
forms g. the uncial n with horizontal cross-stroke on the
(e-

line). Another example is at Berlin, Staatsbibliothek theol.

lat. F. 354 Gregor3^'s Moralia (pi. 5 of Arndt's Scrifttafeln), a

volume (presented by Liudger?) from Werden Library. It

ma}- have been written at Corbie in Leutchar's time, some 50
3-ears before Werden monaster}- was founded.
The Insular r, just described, is a prominent feature of
this t}-pe of script (see line 13 of column ii of our plate, allo-
fvlis), a type represented, so far as I know, bv these three
MSS. onlv.
Our plate comes from fol. V of Petrograd F. I 6. The
reduced size of Staerk's plates of this MS, and its neighbour
spoils their value for palaeographical stud}- and gives the
reader an3-thing but a true notion of the appearance of the
Leutchar-t}'pe.

(3) TJie Maurdramnus-tjpe (pll. IV-V).

Maurdramnus, abbot of Corbie ~~2-l&), gave orders for
a Bible to be written in several volumes (now n.°'ö, 7, 9, 11, 12
in the Bibliothèque Communale of Amiens) and by several
scribes. The text is written large (as in our pi. IV, from
fol. 74"" of n.° 6) for the convenience of the lector in the church-
service, but the indexes are in ordinary minuscule size, the size
found in other MSS. our pi. V, from
(as in fol. 81' of Petro-
grad 0- I 16 Jerome's Liber Comitis).
This t} pe is easil}- distinguishable, with its likeness to
careful Caroline minuscule and the prominent knobs at the
back of such letters as /', s. It too, like the other pair of t}-pes
alread}- mentioned, knows the Insular r (with left branch curved
towards the right). Of other MSS. which offer good specimens
of the script may be mentioned :

Petrograd F. I 13 Origen's Honiilies.

Amiens 220 Paterius.
Paris 13373 Orosius, Alcuin, etc.
All three t}pes are older than what Traube called the '

old script of Corbie a type which is described (with plates)

',

in Revue des Bibliothèques, xxii 405-429, and need not be

 Some Early Script of ihe Corbie Scriptorium 65

described here. It is now called '

the Corbie ab-script '.
and
this name ma}' be left imchallen^ed, although we niiofht also
call it '
the Adelhard type ",
since it was in the abbacy of
Adelhard (under Charlemagne) that it fought a losing battle
with Caroline minuscule. And just as the new Caroline mi-
nuscule and the old-tashioned ab-type were written during the
same period in the Corbie scriptorium (e. g. in Bibl. Nat. lat.
11520-11530), so with other Corbie types. Though the en-tvpe
flourished at tlie all, we tind it in an interli-
earliest stage of
near addition on fol. 117' of Amiens 6 (the Maurdramnus
Bible); though the Maurdramnus t3-pe flourished in the ^•ears
772-780, it appears also in a Paris MS. which contains Alcuin
and other writers. Bibl. Nat. lat. 13373, and can be dated
between <S17 and 835 (see pi. xxvi, 2 in Cab. MSS. '). We '

must therefore exercise due (not excessive) caution in using

the Corbie-types as evidence for date. And it should be a
task for palaeographers to trace the earliest approaches to
this or that type as well as the latest developements. Man\-
Corbie MSS. of the eighth centurv' stand on the border-line
between one type and another.
[Indeed we ma\' add to these t\'pes a fourth t\"pe of eighth
centur\' Corbie minuscule, a fairl}- large minuscule (with close
a and without longa ') which ofters so few salient features
'
i

that it is difficult to describe, but which is easily recognized

after one has become familiar with it. It is contemporaneous

with the en-type in Paris 12239, ff. 53-74 col. i. for the en-type
immediately follows it on col. ii of fol. 74. Other MSS. in this
script are Paris 12240-12241 Cassiodore; 12598 Vitae Sanctorum:
13047 Juvencus, etc; 133-17 (ff. 56-end) Jerome. ED.^>
Finali}', a word of caution. How read\- we should be to
infer that the Maurdramnus tv'pe (or the t\"pe just mentioned)
was the immediate precursor of Caroline minuscule at Corbie
scriptorium, had we not proof positive that the curious ab-t}"pe
(a reversion to the cursive stage, with some Anglosaxon fea-
tures, especiall}- the abbreviation-s}-mbol of -fur) had inter-
vened. Let the palaeographer amass details and avoid rash
inference.
[How this brief but. it is hoped, fairl}' accurate reprodu-

ction or Liebaerts views on the (»rogress of minuscule from

the earl}- part of the eighth to the earlv part of the ninth
66 P. Lieb;ierL

century at Corbie can help historical research may be illustrated

from that hu,<?e glossary which played so important a part all
through the Middle Ages, the Liber Glossarum (' Glossar^' of
Ansileubus •). Two MSS. (Paris, Bibl. Nat. lat. 11529-11530:
Cambrai 633). two fragmentar}' INISS. (Paris, Bibl. Ste Gene-
vieve B 1, 19; Carlsruhe, Aug. fr. 140) are in the ab-type and
suggest that no less than four transcripts of this enormous
work (unless one or both fragments belong to either MS.)
were produced in the Corbie scriptoriimi in the opening }'ears
of the ninth centurw Palaeograph}', thanks to Liebaert and
Traube, can prove this. And, using this palaeographical fact,
the historian can infer the great likelihood that the glossary
was a Corbie work, one of the vast undertaking of Charle-
magne's reign. How far it was based on previous Spanish
work is another problem which palaeograph}' can help to
solve. ED.]
 nate i.

12.

18.

30. 31.

'

>.
/40.
^
41.

^ ì;

50. 51.

7 ^
60.
 7 8. 9. 10. 11. 12.

^n^UiJkSGp- fr^yà^ ^TAT^n? imt^.i Zulîji» WtóSf

 V
e

^ 5; N -^

,^ "* •- A^

a-^ ft I y:

57 t|^ " - ^ - g-.

^ 5 -t,
 c«^ ,>
Plate Hi.

'

/ -7^

/.

I cur T^ uTe^aT"^)^^"^^

^i^ T^vim fV) eci e*^

';x-^

j; uT^nxTa veut —
^*
 1»^

Piale IH.

,/'V

EXPosiTvm3r?-i xMiii-ii- i.!v;tT- «rei-e-. qi.itl^-n^ '^uj^/

r
jxLuJr-n ,!l,-u

ïfÏNCïllSTr
ibi'irr

eu^sM e; ein U b 1 »m"

tT\e««Vir>.I.CVlClTrVori:^CCUlTnip.>t^ • /^

ijltr-»-nor-C cjuo «il

IicJniudccu«xicxTn'trucnioc/i^.Mei-i ^^^"^

t aoTni^aictxï-^'uTO dit' »^bur cJlo'ptsltp cJuoT-vitT.rpecie-^'3^"

? 'tTUH«i:iOTae' cniìe^*

UiuoM^CiXAXC Uciuor>-etT>TT>cvi»rtP

1 auCTJTf't 73\y^uTTievnTr>pt-oCivi
 o^
Plate iV.

IWOlt >

^'0 0(1 s

wafilio
TUTTI lîY^

quiiTirrci'

rnii^ulicuTndoTTiiburfairiii

n^oîCi^urn:, \l ubcH-SvnrœTi.

j;urjuclv Irxdxu Zàbulon

''•?*T'I,^-t»t* ì - /<, * »/»<* » *% «
 .J

Plate V.

l.'MMf'-wrm

^|-=v7 Ltrfmf noiuv oo-ntj^erTK^ • fîcitrcfr

(S'lurrrr.
il eJ iru\2v *»i«r finci^rrnxnT' -

rr (fun ^ &Cn^|>H U . l.CC \CTlUl ^

Jivnr:Yirifî^r' uoftcnnrc^uoclfvc
 ST. ANDREWS UNIVERSITY
PUBLICATIONS, X\T.

PÄLAEOGRÄPHIA LATINA

PART II.

EDITED BY
Professor \\\ ^\. LINDSAY.

Published for St. Andrews Università'

'
by
HUMPHREY iMILFORD
OXFORD UNIVERSITY PRESS
London, Edinburgh, Glasgow, Copenhagen, New York,
Toronto, Melbourne, Cape Town, Bombay, Calcutta, Madras,
Shanghai
1923
PRINTED IN ITALY
 CONTENTS OF PART II.

Collectanea Varia by W. M. Lindsay {to be contmtied^ :

I. Explicit and Finit.

II. Correction of MSS.
III. Aids to Readers.

IV. Scribes and their Ways.

V. I Longa.
VI. Transmission of Texts. ^"^
Ein Basler Fragment des Nordfranzösischen az-Typiis von Paul Leh-
mann.
Berne 207 by W. M. Lindsay.
The Lyons Scriptorium by the late S. Tafel {io be continiied') :

I. The Cathedral Library.

IL The Library of He Barbe.

Bibliographie der lateinischen Buchschrift (bis 1050) 1911-1922 von
Wilhelm Weinberger.

Plate I. (Laon az-script) Bale fragment of Augustine Civ.

Dei, fol. S^
Plates II-III. Berne 207 Corpus Grammaticorum, foil. 1 14^ 130'.
 COLLECTANEA VARIA.
BY W. M. LINDSAY.

This journal was desig'ned (cf. Classical Quarterly xi 41)

to collect material for a future comprehensive History of Latin
Palaeograph}' (exoriare aliquis scriptor), just as the xArchiv
. . .

f. lateinische Lexikog^raphie paved the way for the great Latin

Thesaurus. Under this heading I propose to publish some
desultory notes of mine on Latin ÌNISS. of the period with
which this journal concerns itself till c. 1050 A. D.), in the
hope that my readers will send to me for publication supple-
mentary notes of theirs regarding the same or similar fea-
tures of MSS.

I. — Explicit and Finit.

It is known (*) that the rule '

Irish scribes use Finit, other
scribes use Explicit '
holds only for the MSS. in our period
(hardly for the latter part). The dividing line of usage became
blurred later. But even in our period there are exceptions.
(On the form explicit see below, IV A) :

A. Finit in non-Irish MSS.

Finit is known in Spain. shew its occur-
For my notes
rence in Nac, Toi. 2, 1 ori-
the Toledo Bible (Madrid. Bibl. ;

ginally from Seville; saec. viii ex."); the San Millan Bible
''

(Madrid, Bibl. Acad. Hist. 20 " saec. ix ") another San Millan
; ;

(*) At least it ought to be. The other day I heard of a scholar who
rashly proposed to apply this rule to the extant MSS. of Propertius !
6 W. M. Lindsay

theological MS. (ibid. 60; '' saec. ix "). In Albi 29 Synon3ma

Ciceronis, etc. saec ix "), written probably North of the P}'-
(''

renees (at Albi ?), both Finit and Explicit are used. The
Leyden Ausonius (from He Barbe librar}- and probabl}' written
in that region) has Finit the famous Codex Toletanus of Isid.
;

Et3'm. has Explicit but also Finit).

Finit appears sometimes in Ital}' in the famous MS. of :

Lucca Cathedral (n." 490; written probably at Lucca, c. 800 A. D.)

which is being edited for the Vatican Series by Prof. Schia-
parelli; in \'ercelli 183, in the later addition, in minuscule of
the Caroline t\pe, on foil. 105''-106' Hj'mn to Eulalia ('' saec ix ")
finit. Deo gratias (but in the original MS., at the end of
Augustini Retractationes, cxpliciunt... libri diió)\ in the No-
nantola MS. (Rome, Bibl. Vitt. Eman., Sessor. 38) Augusti-
nus, etc. (with dating 825-837).
Finit appears sometimes in other Continental script for
which Insular influence has not been claimed, e. g. Berlin, :

Staatsbibl., Ham. 31 Concilium Aquisgranense written at Albi;

"saec. ix"); Paris. Hibl. Nat. 12097 Canons (from Corbie;
uncial and half-uncial; ''saec. vi" ibid. nouv. acq, 1203 Go-
;

descalc Gospels; Berne 645 Victorius Aquitanus (in a script

designated by Traube " between Gallic half-uncial and mi-
nuscule"); Munich 4577 (one of the Kisyla MSS.).
And often in the Continental script of Insular (even En-
glish) foundations on the Continent, e. g. Bale F III 15" Isi- :

dore (P\ilda ''

saec. viii ex.
: ") Paris. Bibl. Nat. 9530 Jerome
;

(Echternach ; ''saec viii-ix ") ibid. 9561 Gregory (uncial;

;

St Bertin); Munich 14252 Glossar}' (Ratisbon) and so on. ;

Whereas Explicit appears in these MSS. written in Insular

script of (certainl}' or probabh) Continental scriptoriums the :

Maesyck Gospels (explicit liber sancti evangelii secundum

lohannem. Finito libro deposco ut, etc.); the Treves Gospels
(but also Finit at the end of each lesson); the Milan Seve-
rianus (Bibl. Ambr. C 77 sup., foil. 133-end; half-uncial of
Bobbio); Cassel theol. F 24 Primasius in Apocalypsin (from
Fulda; Anglosaxon half-uncial); Paris, Bibl. Nat. 9382 Pro-
phetae (written by Virgilius at Echternach); ibid. 9527 Jerome
(Anglosaxon minuscule of Echternach); ibid. 9538 Augustine
(Anglosaxon minuscule of Echternach); Petrograd F v I 3,
foil. 39-168 Jerome (from Corbie; Anglosaxon half-uncial);
 Collectanea varia 7

Boulogne 63-64 Augustine's Letters (from St Bertin; Anglo-

saxon minuscule of " saec viii"); and so on.
In the Anglosaxon script of MSS. written in England I

found Explicit e. g. Cambridge, Corpus Christi College 69 Ho-

:

miliae in Evangelia ('' saec. viii "); ibid. 173 (part ii) Sedulius
(perhaps from Winchester, '' saec. viii"); ibid. 183 Bede's Life
of St Cuthbert, etc. ('' saec. ix in. "); Oxford, Bodl, Seiden
Acta Apostolorum (uncial perhaps written in Thanet
sup. 30 ;

by two nuns before 752 A. D.); Durham A II 16 Gospels

(uncial); London, Brit. Mus., Cotton Tib. A xiv Bede's Histor}-
C' saec. viii"); ibid. Harl. 2965 Book
of Nunnaminster (pro-
babl}- from Winchester; "saec. viii"); the Worcester Gospel
fragment, and so on. But both Explicit and Finit in these:
London, Brit. Mus., Cotton Tib. C ii Bede's Histor}- (perhaps
from Durham; ''saec. viii ex."); ibid. Cotton Vesp. A i
Augustine's Psalter (from Canterbury; saec. vii-viii " uncial)""
;
;

ibid. Royal 1 B vii Gospels ('' saec viii "), with Finit after the
Canons; ibid Royal 1 E vi Canterbur}' Gospels (''saec. viii
ex. "), with Finit after the Canons; the Codex Amiatinus, with
Finit after the Canons.
In Welsh MSS. I have noted Explicit in the Cambridge
Juvencus (" saec. ix "). But Finit in the Liber Commonei (on
fol. 19'); the St Chad Gospels; the Hereford Gospels.

The Northumbrian MS., Rome, Vat. Pal. 68 (with Irish

and Northumbrian glosses " saec. viii ex. ") has Finit. The
;

Schaffhausen Adamnan, written at Iona in or before 713

by Dorbbene. has (in Greek letters) on p. 103 Finitur se-
cundus liber.

B. Explicit in Irish MSS.

London, Brit. Mus. Add. 40618 Gospels (after St Mark:
but, curiousl}-. Finit after St John, on a leaf b}- a later Anglo-
saxon hand) ; Book of Kells (but also Finit). In the Book of
Durrow and the Book of Armagh both Finit and Explicit.

C. Finit and Explicit in the same MS.

Beside the MSS. alread\- cited, m}- notes offer; the Paris
Avitus papyrus (Bibl. Nat. 8913-4; perhaps written at L3'ons),
which has Finit at the end of Homilies, Explicit at the end
of Letters; Rome, Vat. Bibl., Pal. 829 Orosius (Anglosaxon
 8 W. M. Lindsay

minuscule of Lorsch), finit liber quintus . . . explicit liber sextus ,-

the Maihingen Gospels (written by Laurentius of Echternach) ;

Paris, Bibl. Nat. 16668 Bede, Aldhelm, Lorschetc. (written at

partly in Continental minuscule, partly in Anglosaxon), with
Explicit in one part, Finit in the other; Carlsruhe, Reich. 253
Jerome in a script designated by Traube " between Gallic
half-uncial and minuscule "), explicit coiumentarius. Deo gra-
tias. Avien, finit Autun 4 Gospels (from Flavigny uncial and
,-
;

Burgundian minuscule of " saec. viii"); Epinal 68 Jerome's

Letters (from Murbach; 744 A.D.); Munich 14080 Isaiah and
Jeremiah (from Ratisbon; Anglosaxon and Continental script
of saec. viii": fol. 109"" explicit. Fineni; Colmar 38, part ii
''

(Anglosaxon minuscule of Murbach), fol. \1?>' explicit argu-

7nentum, and at the end finit evangelium secundum loiianneni;
Hague, Mus. Meerm.-Westr. 9 Canons ('' saec. viii ex. "), fol. 17'
finit, fol. 26" explicit; London, Brit. Mus., Egerton 2831 parti

(Continental script) Jerome on Isaiah (Tours; ''saec. viii").

These notes —
so far as they go —
seem to establish one
or two conclusions. First, that home Irish MSS. usually shew
Finit. Second, that most home English MSS. shew Explicit.
Third, that many home Spanish MSS. shew Finit. Fourth,
that (practically) all other Continental MSS. (free from Insular
influence) shew Explicit.
Some exceptions are easily explained. It is natural that
North English centres under Irish influence should shew Finit
as well as Explicit. But it is strange to find this combination
in South English MSS. (of Canterbury. And why should
Continental English MSS. (seem to) admit Finit more readily
than home English.'' Whether these difficulties can ever be
explained remains to be seen. But, clearly, full details are
wanted. And I hope that librarians will supply me with
them.
The following reports have already reached me:
On Dublin MSS-, by the Rev. Dr. Lawlor:
Ricemarch (Finit, once. Never Explicit); Gar-
Psalter of
land of Howth
(no example) Stowe St John (at end, Finit)
;
;

Stowe Missal (fol. 65 Finit ordo commonis fol. 67 finit amen ;

Deo gratia); Cathach (no example); Book of Dimma (Finit);

Usserianus I (Explicit)Book of Mulling (Finit) Book of Kells
; ;

(Explicit, after Gospels and Canons Finit, after Breves Causae

;
 Collectanea varia 9

and Arcrumentum) Book of Durrow Explicit after Canons

; (

and Gospels, except vStJohn where is amen f with abbreviation-

stroke; Finit after Breves Causae and Argumentum; Explicit
after Capitula Marci); Book of Armagh (Explicit, after Ca-
nons; F"init, after Breves Causae. List of Hebrew Names,
Gospels, etc, but Explicit after Rom., Cor., Galat., Ephes.,
1 Tim., 2 Tim., Tit., Philem. fol. 201'' Explicit liber
; primus
Sulpicii de vita sancti Martini, fol. 214'' Explicit secundus liber,
fol. 220'' Finit amen, fol. 221 F"init amen, fol. 222" Explicit aepi-

stula eiusdem Severi ad Auxilium).

On Insular MSS. in the Bodleian, by the Sublibrarian:
Macregol Gospels (Plnit, but Explicit after Luke); Rawl.
G 167 (concluding leaves of both Gospels are missing) ; Douce
Primasius (Explicit throughout ; but fol. 3'" Expliciunt capitula
libri primi, with suprascript gloss Finiunt over the first word) ;

Liber Commonei (fol. 19' Finit amen. Finit opus in domino, etc. ;

the concluding leaf of the Ovid has Finit in a 10*'' centur}-

English hand); Seiden Acts, written probably at Canterbur)^
(Explicit); Digby 63, from Winchester (fol 71 Finit, fol. 87''
Finit); St Augustine Gospels, from Canterbur}' scriptorium
(Explicit, after Matt., Mark, Luke, but leaf lost at end of John);
Hatton 48, from Canterbury (last leaves lost) no example in ;

Bodl. 426, Bodl. 819, Hatton 93.

On Anglosaxon MSS. of Corpus Christi College, Cam-
bridge, by the Librarian: 69 (all the homilies are ended by
Explicit, except the last which is imperfect); 173 (Explicit, in
both the larger and smaller hand) 183 (Explicit, in the hands
;

of both scribes); 12 (Nothing); 23 (Explicit after two poems;

Finit after one); 41 (Nothing); 44 (Nothing); 57 (Explicit throug-
hout); 111 (Nothing); 140 CFinit after Matthew before the
subscription no Latin after the other Gospels) 146 (Nothing)
; ; ;

162 (Anglosaxon endings ^

Amen) 178 (Anglosaxon endings.
;

But the Reg. Bened. has Incipit Capitula. Explicit Prologus) ;

1(88 (Nothing); 190 (Expliciunt Capitula; Explicit Prima Epi-

stola); 191 (Nothing); 196 (Nothing); 198 (Nothing) ; 201 (Nei-
ther Explicit nor Finit); 214 (Explicit ends each book); 265
(Nothing); 285 (Explicit); 302 (Nothing); 303 (Explicit Prolo-
gus. p. 172); 307 (Explicit Prologus, fol. 3 b; Finiunt Capitula,
fol. 6 a); 321 (Nothing): 322 (The Explicits are much later);

326 (Nothing); 367 (Nothing); 383 (Nothing); 389 (Explicit);

10 \V. M. Lindsay

419 (Nothing); 421 (Nothinor); 422 (Nothina); 449 (Finiunt,

p. 85).

II. — Correction of MSS.

A
whole volume might be written on this topic. I will
select one or two things which seem worth}- of mention and
avoid, as far as possible, repetition of Wattenbach's examples
in '
Scriftwesen '
IV i.

A. Collation with the Exemplar.

The technical word is conferve ,• and the signature conUili

or contidimus, usuali}' in cursive or cr}ptic script, is well known.
But I am doubtful how far we are entitled to contrast conferve
on the one hand with contra legere, relegere, legere, recogno-
scere, etc, on the other, and claim the use of the exemplar
for conferre onl}'. Even to insist that emendare, corrigere
imply independence from the exemplar seems rash, e. g. in
the Basilican Hilar}-, where at the end of book I, etc, we find
contiiU, but at the end of book X emendavi. And certainly (*)
contra legere seems necessaril}- to mean to read the transcript '

face to face with the exemplar but the rest I would be slow *
;

to limit to this or that denotation by any hard and fast rule

(cf. Adamnan's subscription, cited b}- Wattenbach, conférant

et emendent ad exemplar also Jerome's admonition, ibid.,

. . . ;

lit conferrent diligenliiis et emendar ent. ego enÌ7n relegere non . . .

po(ni). Paulus (Diaconus.^) certainl}- uses rclegere in the sense

of '
collate '
in his letter to Adelhard (accompan}ing a Selection
of Gregory's Letters): tua itaque fraternitas, si se facultas de-
derit, reliquas ad emendatiorem relegere studeat
epistolas
codicem. I take it that legi legi legi means I have revised '

this MS. thrice a contrast to the entry in Paris 9533 emen-

',

dabi semel. Deo gratias-

(*) In the new O.xford edition of Livy (vol. II, p. xxiii) this phrase (in

Sidonius ApoUinaris Letters v 15) is actually used to recall the bogey of

'
Dictation of MSS. '
! A sad indication how ignorant an editor of a Latin
text can be of Latin MSS.
 Collectanea varia 11

B. Record of Variants.

In iny small Introduction to Latin Textual Emendation

(London, Macmillan, 1896; French translation by Waltzing- in
the Klincksieck Collection, XXIV) I mention (V i) some actual
attestations in a Würzburg- MS. that another MS. than the
exemplar was consulted. Others are Oxford, Bodl. Libr., Laud, :

misc. 139 Augustinus in lohannem (Würzburg: saec. ix "), ''

fol. 50'' (marg.) sicut alius codex habet; ibid., Laud. 135 Augu-
stinus Würzburg; 842-855), foil. 114' and \l(f velut in alio
and 58"" in alio codice ita. But no one who
codice, foil. 57''
knows the tricks and manners of correctors (and scribes) can
have any doubt that the formulas or s3'mbols) alibi 2iX\à. alias
and aliter are )ften (or generali}-) used to cloak mere scribal
i

errors- Luckih- we have clear proof of this untruthful or at

anv rate misleading In London, Brit. Mus., Add. 11852
use.
(written at St Hartmut's time), on a blank page
Gall in

(fol. 168') another hand reproduces the whole of fol. 169\ pre-

sumabl}' as a writing-exercise, and miswrites rectani of 169"^ as

certain^ adding a marginal note alibi rectam. When alibi was
reduced to three letters alb (with abbreviation-stroke) there
was danger of confusion with albus. (Editors please note).

C. Correction-symbols.

The ideal corrector would collate transcript with exemplar,

word for The usual corrector would glance through the
word.
transcript, marking an3'thing which puzzled him, and afterwards
consult the exemplar in these passages only. What marks did
he use .-

ci) Z (i. e. Çi']T8i or 'Çt|ti]t£ov). Paulus (Diaconus used .^^)

this symbol in the Selection of Gregory's Letters (54 in all)

which he made for Adelhard of Corbie. We seem to have
his autograph in the corrections in Petrograd F v. I 7. His
preface is (fol. 1'') : quia mihi eas ante relegere praeoccupatione
totas non licuit, triginta quattuor ex eis scito relectas et prout
potui emendatas esse, praeter pauca loca... quibus in locis et
forinsecus ad aurem {leg. oram !) zetam, quod est vitii signum,
apposui. An example will be seen in the plate (in Hartmann's
edition) of fol. 3', where Greg^ory wrote casto mag. ml (with
abbreviation-stroke) '
Casto magistro militum ',
but the scribe
 12 W. M. Lindsay

of Paulus' exemplar casto magni I (with abbreviation-stroke).

Tliis Paulus alters to casto magnifico.
M}' notes suggest that this symbol is affected by Insular
scribes, e. Book of Armagh; Bale F III 15'' Consentius
g.:
(from Fulda); Rome, Bibl. Vat., Pal. 1547 Seneca (from Lorsch);
Cassel, theol. F 22 Jerome ^Anglosaxon minuscule of Fulda) ;

Vienna 16 (Bobbio), e. g. toi 9^

h) R (i. e. require or requirendum). The St Gall eleventh
century corrector, Ekkehard IV, used this symbol in his cor-
rection of a ninth centur}- Augustine (St Gall 174) init. : Liber
optimus, nimis autem vitiose scriptus hunc ego quidem cor- ;

rigere per me, exemplar aliud non habens, si poteram temptavi ;

ergo ubi minus potui r literam apposui). An e.xample occurs

on p. 39 where the scribe's ad te petis iit scvibererii is altered
by Ekkehard to ad te petis id scribam. An eleventh century
transcript (St Gall 139) of this MS. shews passage ad
in this
te petis ut scribam, a useful indication how an\' abnormal diction
Or spelling might be wholly effaced in transcription.
The capital form R is the usual s}'mbol, though RO or
rq (with abbreviation-stroke) is common too. An example of
the two-lettered symbol in Brussels 9403 Gregory of Tours
(' saec. VIII-IX ') is instructive. On fol. 79"' the scribe wrote
sigilatus ; the corrector (in the margin) vel ligatus, rq; a third
hand (after consultation of the exemplar? has written / above
the second syllable of sigilatus. So the exemplar 'had sigil-
latus, while ligatus was the corrector's guecs.
The simple verb, quaere, quaerendwii, is also common (ab-
breviated to q or qr, with abbreviation-stroke but the pre- .

dominance of the compound is perhaps indicated b}' the use

of reqidrere to correct a MS '.
'
Thus in Rome, V.t. Bibl,
474 Augustini Sermones (written at Ferneres in the time of
Lupus) there is a marginal entry on fol. 95'; hue usque ab
abbate et praeceptore Lupo requisitum et distinctum e.st. (Yet
on fol. 78' the scribe wrote ibi ?ios periclitantur and set q [not
rq, in the margin Lupus corrects to ubi nos periclitamur).
;

Editors should know the errors produced by these symbols.

The 0-S3'mbol suggested quae (if the stroke was above), qui
(if the stroke cut the shaft), etc. the R-symbol respondit, res,
;

or the s}'llables -runt, -rum. The Corbie (.^) compilers of the

Liber Glossarum used R (with a cross-stroke through the leg)
 Collectanea varia 13

throug-hout. Here are some miswritings: Abolitarum (for

Abolita R); Virgilius respondit (for Virgilius R); Rex (for Re-
quire exempkimr).
c) The Cryphia (cf. Isid. Et\-m. I xxi 10 cryphia, circuii
pars inferior cum puncto, ponitur ubi quaestio dura
in his locis

et obscura aperiri vel solvi non poluit). A Tours (?) corrector

of the Liber Glossarum used this mark (as the Corbie comp-
ilers had used the R-s}'mbol) and it has been preserved in

two MSS. derived from this (lost) corrected MS., viz. the ninth
century Tours MS. and the eleventh centur}- Vendôme MS.
For example, at IN 1721 where the compiler had written In-
tempestivus, which had been corrupted to Interprestius, these
two MSS. offer Interrestrius but the presence of the Cryphia
;

shews that this was a wild guess of the Tours corrector.

The symbol has usually (e. g. in the MSS. just mentioned)
the form of c with dot inside, and this c hange of form ma}'
be due to the fact that c with dot inside was the symbol of
cor (suggestive of corrige ').
^
Sometimes the c is revers ed
(like the antisigma cum puncto of Isid. Etym. I xxi 12),
' '

but that corrige was meant may, I think, be inferred from, the
appearance occasionally of this word written in full; just as
requireis occasionalh' written in full (e. g. Paris, Bibl. Nat.
1732 Ambrose, fol. 31^ fol ZV Oxford, Bodl. Libr., Hatton
•

98 Sidonius. fol. 118', fol. 118'), or quaere (e. g. ibid.. Laud,

misc. 134 Augustine, fol. 31'' quere in aliis codicibus). Châtelain
(Not. Tiron. p. 113) cites a shorthand note in Verona 59, hie
requirendum (cf. Servius ad Aen. 6, 177: Probus tamen et Do-
natus de hoc loco requirendum adhuc esse dixerunt).
It is quite possible that investigation may find in these
corre ction-S3-mbols a clue to the provenance of a MS. (or ar-
chetype). M}^ notes shew that the R-symbol appears in these
Corbie MSS. of Adelhard's time Paris, Bibl. : Nat 11681 Beda
in Lucam; ibid. 13354 -[- Petrograd O v. I 19 (for '
Adalhardus
monachus jussit fieri '
seems to refer to Adelhards degradation).
Also in an earlier MS. from Corbie, ibid. 12205 Augustini
Opuscula. But also in Murbach MSS.: Epinal 68 Jerome's
Letters (written in 744); Geneva 21 Beda in Acta x\postolorum
Ç saec. viii-ix "). The Milan Josephus papyrus has sometimes
R, sometimes r. From a Lorsch MS. in the Vatican Library
(Pal. 966) I noted the r-.symbol. And so on. The Cr3phia
14 W. M. Lindsay

(in correct form) I happened to note only in the Florence

Orosius. fol. oò (marg.) ?iûn est sensus in hoc loco Verona 68
:

Rhabaniis Maurus ("' saec. ix "'), where it marks omissions.

D. Correctors' influence on Texts.

A which was not revised {I'clectus) with the

transcript
exemplar was a Complete absence of cor-
sloveni}' thing.
rections is a bad sign. Such a MS. would rightl\- find scant
favour in scriptoriums and be debarred from the role of exem-
plar. When the transcriber is an ignoraiiius the text is often
unintelligible e. g. the text of Gregory's Moralia (Munich
:

Staatsbibl. 6297) written by the Englishman (.') Peregrinus at

Freising about the year 780. But a modern editor prizes what
a medieval teacher shunned, a text in which the spelling had
not been altered to Carolingian requirements. The codex
optimus of Nonius Marcellus is a ninth centur\' Tours trans-
cript (i.e\'den, Voss. F 73), in which, although a subsequent
corrector has altered in the usual Carolingian style adfero^ to
afferò, conloco to colloco^ etc., the transcriber's spelling can
easily be distinguished from these corrections. A Carolingian
corrector who revised the spelling {orlhographiam praestare)
would never pedantically foist an archaic form on a text (like
the Renaissance copyists' qiicis where the exemplar had qids
Dat. Abi. Flur.) The apparent instances in Chatelain's plate
of the Tours Livy (Rome, Vat. Bibl., Reg. 762) I have shewn
to be illusory (Class. Rev. x 233). And in medieval times
there was the same perverted taste as now for conjectural
emendation. A codex ab abbate emendatus would be the
' "

chosen exemplar in a scriptorium. That is wh}' a modern

editor of (let us sa\) Juvenal finds the family ol inferior MSS.
so much larger than the better family.
Corrections h\ such an abbot as Lupus of Ferrières are
of a different tvpe; and I have published in Class. Phil, iv 113
a photograph of a page which shews his handwriting in order
that editors of texts may
recognize his intervention. Lupus,
knowing the weakness of mere conjectural emendation, bestirred
himself to find and borrow ancient MSS. (See Traubes Kleine
Schriften, pp. 3 sqq., for Lupus' work on the text of \'alerius
Maximus). The Florence MS. Nonius Marcellus of (part of)
is transcribed from the LeN'den MS., but was subsequently
 Collectanea varia IS

corrected with the help of an excellent text, now lost. (Its

transcript, the Harleianus, preserves most of these good cor-
rections but, of course, o-ivesno indication that they come
from an alien source). A modern editor must be capable of
distinguishinc{ from mere Carolingian conjectural emendation
these corrections (in a corrected MS.) or readings (in a trans-
cript of a corrected exemplar) which come from such a source.
For example, the corrector of the Florence MS. substitutes
rightly muliinumtmis for muäimi mumnus, iniserinwii (cf. Ital.
poverino) for uiiserrimiDii in lines of Varrò and Lucilius. Such
substitutions could never be mere conjectural emendations by
a Carolingian corrector.
Traube's Textgeschichte der Regula S, Benedicti (.Munich
1898 =
Abhandl. bayer. Akad. Ill CI. XXI Bd.) provides the
star to which every twentieth century editor of a Latin text
must hitch his waggon. The punctilious, detailed, neat or rather
ornamental record of variants in St Gall 914, the cop\' of the
MS. prepared at Charlemagne's orders, shewed that this MS.
had a history; but it was Traube who waved his magic wand
over it and made it tell its whole tale. What histor}'. I wonder,
is attached to the Milan Orosius (Ambr. Bibl., D 23 sup. ;

half-uncial) ? Zangemeister finds none. It too shews a wonder-

full\- careful record of variants, sometimes inserting them

(between s\'mbols like the Insular abbreviation of est) in the
text, a practice which must have puzzled transcribers e. g. :

fol. 23'' 2it quisque ut quis non. The record of so insignificant

a variant as (fol. ó"") aetJiiopìa and ciethyopia suggests that it

was no common MS. which the corrector brought to his help.

III. — Aids to Readers.

The huge letters, the narrow column (with onl\' some

10 letters to the line), the absence of all division of word from
word in such MSS. as the Vatican Cicero de Republica make
the reading very difficult and slow. One wonders whether
these unciales litterae were not rather meant for show than
' '

for use, like (let us say) a Kelmscott Press edition. Even those
ordinary uncial MSS. (of the fifth or sixth centuries), from which
we may suppose most of our extant Carolingian MSS. of
ancient authors to have been transcribed, \ve cannot now read
16 W. M. Lindsay

with ease and speed. How they taxed the intelligence ot Caro-
lingian transcribers is shewn in detail by Shiple}' in his useful
stud}' of the Tours transcript of the Codex Puteanus of Liv\'
(Certain Sources of Corruption in Latin Manuscripts, New York,
1904, pp. 15 sqq.). And though Traube pointed out (in vol. II
p. 133 of his Einleitung in die lateinische Philologie des Mit-
telalters) that the most fertile ground for corruption of text
was the transcription of MSS. in Insular or other unfamiliar
minuscule, with their puzzling forms of abbreviations and liga-
tures, there are still editors whose sole panacea seems to be:

Rewrite the passage in majuscule script, with no division

between the words, and see what it looks like. Wattenbach
(in chap. IV section of his Schriftwesen im Mittelalter and
i
' '

on p. 86 of his Anleitung 1886) has indicated the chief stages

' ',

in tlie long process of making MSS. easy to read a process

completed b\- Charlemagne and his successors. I will add a few
details.
A. Separation of Words by Dots.
Words are often separated bv dots in these MSS. : the
Bobbio Missal (Paris. Bibl. Nat. 13246), e. g. fol. 15"" sit. aelutio.
scelerum. sit; Laon 423 Isidore (Laon az-script of saec. viii);
the Paris Oribasius (Bibl. Nat-, nouv. acq. 1619; from Fleur)^;
in a script designated by Traube '' between Gallic half-uncial
and minuscule ") Cambrai 836 Isidore (uncial) the Leyden
; ;

Gregor}' of Tours (Univ. Bibl., Voss. O 63; Merovingian); the

Leyden Symphosius (ibid. Voss. Q 106; " saec. ix "), e. g.
Mechus, eram. regis, sedlignea., etc.; Carlsruhe, Reich. 222
Apocalypse ('' saec. viii-ix ") Verona 62 Cresconii Canones
;

(in Verona cursive); St Gall 213 Gregory ('' saec. viii "), e. g.

p. 172 Ibi. cadere (end of line) superbiam. ubi. nasci., etc ;

Einsiedeln 18 Commentary on Psalms (" saec. viii-ix ").

In some MSS. it is only (or main!}') small words that are
so treated, e. g. the Maurdramnus Bible (Amiens 12), fol. 70""
:

Adiurovos. o. filiae hierusalem Oxford, Bodl. Libr., Laud, ;

lat. 92 (Anglosaxon minuscule of W^ürzburg, 832-842), fol. IP

Et. ut. magis. The last MS. (like some others) so treats
unusual or foreign words (see below), e. g. ruben. gad. and ;

this practice recalls the treatment of abbreviation-s}'mbols (see

m}- Notae Latinae I p. 73), e. g. oblita. e. (with abbreviation-
stroke above) oblita est clearl}- distinct from oblitae.
' ', Simil-
 .

Collectanea varia 17

arl\' with Numerals. The symbol v "

five "
is distinguished
from the letterz\ the symbol x ten from the letter .v, etc, '
'

bv havincr a dot before and after (and often by having an

apex above, just as we set a dot above).

B. The Apex.
The use of the apex (in the form of the French acute
accent) to indicate long vowels on inscriptions of the Empire
survived in our manuscript-period to a very limited extent.
îtwas retained mosti}- to preserve some monosyllables from
being annexed by their longer neighbours. The monosyllable
most frequentlv so preserved is perhaps the Interjection o.
Also the Pronouns w^, tii, te, se, etc- Also the Preposition a,-
for the combination of Preposition with Noun, ajiiictu, was
almost as normal as its composition with a \^erb, aßuo. And
so on.
But in Late Latin vowel-qualit\- had supplanted vowel-
quantity.So the practice was extended to monosyllables
whose vowel was short, a departure from the ancient use of the
apex, e. g. a7i, ac, at, es.
Unluckily the original form of the apex was not strictly
conserved. Some of its phases were harmless but when it :

was written horizontalh', instead of obliqueh.it became indistin-

guishable from the abbreviation-stroke a (with apex) the In- ;

terjection and a (with abbreviation-stroke) aut e (with apex) * '

;

the Preposition and e (with abbreviation-stroke) •

est ': os (with
apex) '
mouth '
and os (with abhreviation-stroke) '
omnes '
; an
(with apex) the Conjunction and an (with abbreviation-stroke)
'ante': sie (with apex) the Conjunction and sie (with abbre-
viation-stroke) '
sicut '.
Indeed the abbreviation-stroke itself
sometimes took an oblique direction.
The sensible practice of ancient times is well stated by

Isidore (Et\m. I xxvii 29) : in dubiis quoque verbis consuetude

veterum erat ut. cum eadeni littera alium intellectum correpta,
alium producta haberet. longae svllabae apicem adponebant ;

utputa populus arborem signihcaret an hominum multitudi-

' "

nem apice distinguebatur. In the Codex Mediceus of Virgil

??izsere' they sent" is thus distinguished from misere wretchedly ' '.

/ez'issmooth from /ez'is

' '
*
light '

; although the apex is put to

sone less traditional uses now and then, and in the great ma-
2
18 W. M. Lindsay

jority of its occurrences stands over long monosyllables, e. g.

^, Jiac, sus (cf. Isid.Etym. I xviii 6). M}' notes on various MSS.
suggest that it was Insular scribes who utilized this indication
of -is Dat. Abi. Plur.. e. g. the Lichfield Gospels the Cambr-
: ;

idge Juvencus; the Book of Mulling (also -os Ace. PI, -as of
ignoras, etc.) Laon 26, extis Paris, Bibl. Nat. 9527 (Anglosaxon
; ;

of Echternach), servis; Hereford Gospels, vobis, nostris the ;

Carlsruhe Augustine, nobis (also -us of virtus ; similarly the

Carlsruhe Bede, -us Ace. Plur.) the Gatien Gospels. Certainl}' ;

this mark over -is had the disadvantage that, when apex and
abbreviation stroke coincided in form, servis was confusible
with servi sunt^ as in Cassel th. F 22 Jerome (Anglosaxon
minuscule of Fulda). A ninth centur}^ Breton MS. (Paris, Bibl.
Nat. 12021) shews degeneration, e- g. fol. 73'' quanto magi's a
'

vivi's, which ought to mean as much from living magicians

'

but which was intended for as much more from the living ' '.

C. Foreign Words.

In the beautiful uncial of the Bologna Lactantius a Greek

word hardh' distinguishable from a Latin, unless it contains
is

a characteristic letter like Xi or Phi or Omega. Where Greek

was unknown (i. e. in most scriptoriums) Greek words are, as
a rule, omitted by scribes, a space being left for the corrector
to supply them. Often the space was never filled and that ;

is wh\" many extant MSS. of the Abavus Glossary have dis-

carded the Greek words in those items which the compiler
borrowed from a bilingual glossary. (Luckil}' the archetype
of a cognate glossary, Abavus Major, had preserved them).
The corrector, if he filled the blank, would replace the Greek
letters b\" Latin and in an\- MS. where a Greek word shows
;

itself we may expect to find at least one attempt (often more)

b}' some monastery-teacher to write above it (or in the adjacent
margin) a Latin transliteration). Often a Greek word survives
in a h^'brid form, some of the letters being (or tr3'ing to be)
Greek while the rest are Latin. We can sA'mpathize with the
Fulda corrector of a MS. of Consentius (Bale F III 15'), who
suggests (in the margin) salpo for Calpe (of the text, fol. Il')-
He did not know whether the first letter was meant for Greek
Sigma or Latin c.
 Collectanea varia 19

lust as we distinguish foreign words by italics or b\- in-

verted commas, so medieval scribes drew a horizontal stroke
above. At an\- rate this is the normal practice with Greek
and Hebrew words. But not with Celtic words. These usually
have an apex (or, as in the Schatthausen Adamnan. a short
horizontal stroke) over each S3llable or each vowel), e. g. :

the Cambridge Juvencus the Corpus Martianus Capella; the

;

Book of Armagh a Northumbrian Psalter in the Vatican Li-

;

brar}- (Pal. 68). In the last (and in other MSS.) Teutonic words
are similarly treated.

D. Quotation-signs.

We use inverted commas (double, as well as single) also

to warn the reader when a quotation comes. The ancient
sign was the Diple (like a horizontal V. with opening to the
left). Isidore says of it (Etym. I xxi 13): hanc scriptores no-
stri adponunt in libris ecclesiasticorum virorum ad separanda
vel demonstranda testimonia sanctarum Scripturarum. Of
course, in ecclesiastical writings the quotations would normally
come from the Scriptures. But I find no reason to believe
that in the practice of scribes (in Spain or elsewhere) it was
only Scriptural quotations which were so treated. For example,
in the Codex Toletanus of Isidore's Etymologies (Madrid, Bibl.
Nac., Toi. 15, 8; " saec. viii ex. ") the quotation-sign stands on
fol. 3^ opposite a quotation from Persius and opposite a quo-

tation from the Bible.

Dulcitius, the corrector (568-590) of the \'ienna Hilary, uses
the Diple in its proper shape. Most subsequent scribes write
what might be called a cursive form of this horizontal W a
form resembling our letter s (see Thompson Tntrod. \ facs. 178),
and this s-mark is often doubled (ibid., facs. 87). Curiously
enough, a mere ninth century MS., Colmar 82 Cassiani Collat-
iones Murbach " saec. ix in. ") shews along with it a mark
;

which must have been the half-way stage, a horizontal V be-

ginning with a small curve and with the angle rounded. It
shews also a rare quotation-sign, like our Y (with a dot between
the branches). This Y-mark (sometimes without the dot) I
have found (at least, noted^i only in London, Brit. Mus.. Add.
1187(S Gregorii Moralia (Luxeuil t3-pe); ibid., Add. 31031 Gre-

gorii Moralia (Laon az-script); Rome, Vat. Bibl, Pal. 177 Jerome
20 W.U. Lindsay

(Anglosaxon minuscule Lorsch; saec. ix in, ") Cambrai 441

of ''
;

Philip pus' Commentary on Job (half-uncial. Also the s-mark,

with a dot to the left). A more common variet}' (but not
nearh' so common as the s mark) resembles the Arabic nu-
meral 7, and this 7-mark usually has two dots (sometimes more),
one following the other, on the left of the top (see Thompson
'
Introd. ',
facs. 131). It was an English variet}', transmitted

to the Continent. For my notes of it are from these MSS. :

London, Brit. Mus., Ro3^al 1 B

(Anglosaxon half-uncial);
vìi

Rome, Vat. Bibl., Pal. 68 (Northumbria) Cassel, th. F 22 ;

(Anglosaxon of Fulda); ibid., th. F 21 (Anglosaxon of Fulda,;

Paris, Bibl. Nat. 9525 (Anglosaxon of Echternach); ibid. 9527
(Anglosaxon of Echternach) ibid. 12281 Bede (also the s-form);
;

Petn grad O v. I 18 Bede (Anglosaxon minuscule); Munich

14653 (Anglosaxon of Ratisbon), not quite the usual 7-form;
Epinal 68 (Murbach; 744 A. D.); Rome, Vat. Bibl., Pal. 195
(Anglosaxon of Lorsch) Berlin th. F. 356 (Anglosaxon of
;

Werden); Munich 4549 (a Kisvla MS.). The Moore Bede uses

a mere horizontal stroke; and the same, with a dot above and
below (like the Insular est -svmbol), appears in Petrograd F
' '

V. I 2 (from Corbie; uncial and half-uncial and the Douce

Primasius; also in the Corbie ab-script of Paris, Bibl. Nat. 12135
and of Cambridge, Corpus Christi College 193.
The quotation-sign stands in the margin and is repeated
opposite each line of the quotation. (In the Echternach MS.
in Anglosaxon half-uncial, Paris, Bibl. Nat. 9382, Virgilius' me-
trical subscription has the 7-mark opposite each line). But
when the quotation begins in the middle of a line, the sign
often stands at the quotation's beginning (i. e. in the text):
Paris, Bibl. Nat. 653 Pelagius (written at Verona?) e. g. fol. 209";
Autun 24 Cassianus (half-uncial), e. g. fol. 124'.

IV. — Scribes and their Ways.

(A supplement to chap, V of Wattenbach's Schriftwesen).

A. Illiteracy of Scribes.

This is abundantly illustrated by their subscriptions, e. g.

of Martin the priest in Avranches 229 ('' saec. x ") Martinus :

presbiter.., Fratres qui legit ora pro illuci ; of Joseph of St Gall

 Collectanea varia 21

(St Gall 189 Eucherius) iosepuss srüpsit. : How could Joseph

be trusted to transmit faithfully to us the spelline of his exem-
plar? And by probationes pennae
'
e. g. in the Essen Gospels
',

thrice on fol. V omnia vincit amor et nos rt'/w6>i- amori in a

: ;

Fulda MS. in Anglosaxon script Cassel theol. F 22) on fol. lo":

birillns topatsius cyj'ieleig'soii. Vet South Italian scribes who
carried out the rules of Beneventan script must have unders-
tood what they were copying (see Loew's Beneventan Script).
Excerptum becomes in scribal Latin excarpsiim or scarpsnm
or the like, e- g- London. Brit. Mus.. Add. 29972 (Luxeuil
:

type) excarpsum (in title); Petrograd O v. I 20 (Corbie;

" saec. ix ") incipiunt scarpsa testimonia sancti Hieronymi
;

St Gall 125 ('' saec. viii-ix "), p. 158 incipit sententias excarapsas
de humilias sancti Gregorii papae. Explicit^ a suspension of
explicitus, is often said tohave produced the perfect explicnit,
e. g. Turin, Arch. I b vi 28 Lactantius (Bobbio; uncial and

half-uncial) fol. 8L Ouintus luHus Hilarianus explicuit; though

the Le\den Priscian shews (fol. 163^ fol. 189^) explicitus and ;

explicuit in a neuter or passive sense is defensible (cf. Jerome

ap. Anecd. Helv. Less defensible is expliciunt (on
179, 19).
the model of incipiunt). Perhaps the bad Latin in scribes'
subscriptions should really be taken for a relapse (Jrom the
literary Latin of the text he transcribed) into the "voilgar Latin
(halt\va\' to Romance) which was current in talk.

B. Fidelit\- of Transcription.

How far can we rely on a transcript as a faithful repro

duction of the exemplar.- When, by an error, a page or a
passage has been copied t\vice (b}- two scribes), we always
hnd divergence in spelling and abbreviation, if not also in ac-
curacv, in the two transcriptions. Here is an example: Lyons
471 (401) Bede on Ezra (" saec. ix in. " fol. 75' copied again :

by the scribe of the next quaternion on 76^) completo and con-

picto, iJierosolimam and kierosoliniani. est abbreviated and writ-
ten in full, aretem and ariaetem^ esse written in full and ab-
breviated, fickle and fidèles, vet written in full and abbreviated,
temptari and temtari, ostes and Jiostes (The scribe of 75' gets
almost two lines more into his page). Even when only one
scribe is concerned, his second transcription is likely to differ
in some point or other from his first. For example : in a mere
22 W. M. Lindsay

passage of eight lines in a Bodleian MS. of the year 818 Bodl.

849 Bede), temptis abbreviated and written in full in a single ;

sentence of a Lorsch MS. (Rome, Vat. Bibl, Pal. 1547 saec. ''
;

viii-ix ") the am of quicquam is abbreviated and written out,

the er of here is written separately and in ligature.

C. Rate of Transcription (to supplement Wattenbach

Schriftwesen ^ p. 289).

When in our period (i. e- before the time of paid scribes)

a scribe mentions in a subscription the time he has taken over
his task, we may usually infer that he was proud of a feat of
rapid writing. Or rather, rapid and neat writing. Bishop
Baturicus of Ratisbon (814-847) records in a subscription (cited
b\- Wattenbach) in Munich, Staatsbibl. 14437 Augustinus in

Epist. S. lohannis: Librum hunc pro remedio animae meae

i

ego in Dei nomine Baturicus episcopus ad Pranchonofurt scri-

bere praecepi scriptus est autem diebus septem et in octavo
;

correctus in loco eodem, anno septimo regiminis episcopatus

mei et octingentesimo vicesimo tertio Dominicae incarnationis ;

scriptus autem per Ellenhardum et Dignum, Hildoino ortho-

grafiam praestante orate pro nobis. This codex, in neat and
;

careful script (see the plate in Pal Soc I pi. 23), contains 109
leaves of 10 X 8 inches, with 20 lines to the page, and (Prof-
P. Lehmann tells me) no leaves have been lost. That means
about thirty pages a da}' by tlie two writers or fifteen by one
(about a (juaternion a day).
To Prof. Lawlor's paper on the Cathach of St Columba
(Royal Irish Academy IVoceedings, XXXIII C 11 [1916J) I
have added an appendix, which discusses the famous subscript-
ion in the Book of Durrow. I declare it to be a ver}- close
copy of Columba's actual subscription in the (lost) exemplar,
with the alteration of the single-letter abbreviation of nostri
(which at the time of the transcription would suggest non and
spoil the sense) to the symbol. In the last line
three-letter
Columba had been pressed for space, and he abbreviated (as
well as the necessar}- abbreviation dni Domini ') gratia, no-
'

stri and siibscripsi. using for the last the symbol current in
his time (but obsolete later) s. s. Columba wrote Rogo bea- :

titudinem tuam sancte praesbiter Patrici ut quicumque hunc

libellum manu tenuerit meminerit Columbae scriptoris, qui hoc
 Collectanea varia 23

scripsi mihimet evang"elium per XII dierum spatium gratia Do-

mini nostri. Subscripsi. Book of Durrow follows (at a
In the
by the same writer (probably
respectful interval of seven lines)
the Abbot of Durrow) an invocation of St Columba ora pro :

me, frater mi: Dominus tecum sit. Columba's subscription,

with record ot his writing^-feat and containing a phrase not found
in subscriptions I wrote for my own use
'
tallies exactl}' with
',

the well-known stor}' of his surreptitious transcription of the

Gospels brought by St Finnian from Rome (see Lawlor's full

account). The quarrel between the two saints led to Columba's

self banishment to Iona. We cannot doubt that the immediate
exemplar of the Durrow Gospels was Columba's hurried trans-
cript of Finnian's Gospels (the Wilgate text, unknown in Ireland
before Finnian's visit to Rome). The script of the Cathach'
i. e. the Psalter associated b\' tradition with Columba. may

quite well be Columba's own handwriting, a rough, flowing

half-uncial t\-pe which would lend itself to rapid writing (see
Lawlor".s plate How many pages of this script would be
i.

needed for the Gospels cannot be determined with certaint}'.

Between 20 and 30 would have to be written each day. I fancy,
to transcribe all four Gospels in 12 days. On the other hand,
the entry in the Book of Armagh is no evidence of rate of
transcription. It mereh" records the curious coincidence that

the scribe Ferdomnach had hnished St Matthew's Gospel on

St Matthew's Day (explicit... scriptum. Atque hnitum in feria
Mattel), without mention of when he began. How to interpret
the entr\' at the end of the Frobae Canto, a poem which oc-
cupies foil. 129''-140'' (line 7) of a quarto MS. in the Corbie
ab-script (Paris, Bibl. Nat. 7701) is doubtful Idibus Octobribus :

(onh- last syllable certain) sic percraxatus (first s}'llable un-

certain) fui horis written // with transecting stroke) tribus (but
\{ iioris trihits was meant for *
at three o'clock ",
it should have
been hora tcriid).

And when a long time is mentioned,we cannot determine

the rate of dail\- transcription (since we cannot be sure that
each da\' was a working-da}') nor may we infer rapid writing.
The Conservateur en chef tells me that the Brussels Paschasius
(Bibl.Roy. 8216-8) of the \ear 819 (mentioned by Wattenbach),
written b\' one hand in 15 weeks, originali}- consisted of 304
leaves.
24 W. M. Lindsay

In Milan, Ambr. Bibl. 301 inf., we find on fol. 100' the

pious entry (in the top-margin) : in nomine Jesu et sanctae
Brigdae. The same entry reappears in the top-margin of
fol. 108". May we infer that foil. 100^-108' were the task of
a day, i. e. slightly more than the amount of a quaternion?

D. Conversational Jottings.
This was a bad habit of Irish scribes. All scribes were
forbidden to talk in the scriptorium, but the irrepressible
Irishman scribbled his remark on the margin of the page he
was transcribing and shewed it to his neighbour. I have given
examples in my Earl}' Irish Minuscule Script (p. 42), one of
which, since it bears on the last section, ma}' be cited here, the

Irishsentence meaning this page has not been written very '

slowly It was written by Coirbbre in the St Gall Priscian.

'.

Clearly his teacher had been rebuking him for slow transcription,
for he refers in another jotting to his teacher's displeasure.
Other examples are given by Kuno Meyer in Zeitschr. Celt.
Phil. VIII 175 from a ninth century MS. of Cassiodore on the
Psalms (Laon 26) e. g. It is cold to day. Naturally. Winter
:
' ',

'
The lamp gives a bad light It is time for us to begin to do ', '

some work Well, this vellum is hairv

', '
Well, I call this vellum ' '

thin I feel quite dull to-dax.

', '
I do not know what is wrong

with me '.
These cited here are in Irish. And all are written
in the top-margins of the pages as clearlv and carefullv as the
text The head of the scriptorium, if he understood
itself!

Irish, must have been ver}' indulgent. But I do not see any
reason for believing this MS. and the other Laon MSS. in
Irish script to have been written in Ireland. One of them
(Laon 55) has on a fl}'-leaf a part of a Latin elegy on Cathasach
'
prudensque magister Atque pius iuvenis castus custosque de-
corus', clearh' (though the script looks earlier than the end of
the ninth centur}') tlie Abbot of Armagh, the '
religiosus iu-
venis '
of the Annals of Ulster, who died in 896. Meyer shewed
this in Beri. But Meyer went on to make
Akad. Sitzber. 1914.
one of these rash theories which hinder rather than help
learning, in this age of minute but unintelligent bibliography
when every theory good, bad or indifferent —
is recorded —
with the same respect. He inferred that the scribe, (a fair
enough scribe, but nothing out of the common) was that
 Collectanea varia 2&

'
scriba optimus '
of Armagh, Mochta, who was a contemporary
of this Cathasach. \Vh\- ? Was Mochta the onl\- scribe at
Armagh? And were there not Irish scribes (any of whom
may have come from Armagh) at Laon, where we know there
was an colony? Irish
The
onU" example of conversational jottings by a Conti-
nental scribe which I have noted is an entr}' b\' the scribe in
the top-margin of fol. 32' of Treves 122 Ratpodo archiepisco- '

pus enables us to date the MS., for Radbodus was elected

'.
It

Archbishop of Treves in 883. But the entry ma}^ not have

been made for conversational purposes. Still less another
dateable entry, b\- the Englishman (?) Peregrinus at Freising
(in Munich, Staatsbibl. 6297 Gregorii Moralia). In the margin
opposite a passage describing Job's undeserved calamities Pe-
regrinus writes (fol. 67'') Arbeo episcopus; Miserere Deus',
'

a clear reference to the persecution by Duke Tassilo of the

Bishop of PVeising. Aribo or Arbeo.

E. Marginal PraN'ers.
was the habit of beginning the da\''s task of
Irish too
transcription by a prayer recorded in the top-margin. The
favourite Irish formula was xb ^with abbreviation-stroke through
the second letter), probably to be interpreted Christe, be- '

nedic although Stern (Zeitschr. celt. Phil. VI 546) mentions

'
;

that both this symbol and the fuller phrases xQ^i ßoTjoiiaov,
adiìiva nos Christe appear in St Paul 25 D 26. (The entry
in the uncial leaves at the end of the Utrecht Psalter, (tyia
Maoia tw yoa^pavri seems rather a subscriptio, written
|3oiii)iiGov

at the end of transcription). In the St Gall Priscian xf" Christe

fave" appears along with xb on p. 171, etc. fave Brìg{ittd)\ ;

on p. VèV fave P{atrici)e\ on p. V/S P{atrici)e b{enedic). The

loquacious Irish scribe (already mentioned) of Laon 26 writes
(fol. 12') : benedic. Domine, hanc operationem ut plana fiat ;

also (in Irish; fol. hands to-day

18') He also
'
God bless my ".

puts xb in the left top-margin of each folio verso. So does

one of the scribes of a Winchester MS. of c 850 (Oxford,
Bodl. Libr., Digby 63).
For the pious practice spread to EngHsh scribes too. The
Mart3Tolog3' of Willibrord shews at the beginning (fol. 2^ top-
margin) Christe fave votis and at the foot of a Lorsch MS. in;
26 W. M. Lindsay

Anglosaxon script (Paris, Bibl. Kat. 16668; fol. 5Ì\ the first

page of a new quire) there seems to be x b.

The Boniface Gospels (Fulda, Bonif. 3) are written under

Celtic influence. Their pious entr}- o Emamiel (i. e. God with '

us appears at the head of the pages ot St Luke, etc. The

')

same entry is found in the Bobbio MS. (Milan, Ambr. Libr.

C 301 inf.), e. g. fol. 46% fol 136%
In the Vatican Librar}- (Pal. 237 " saec. ix ") is a MS. of ;

Prosper, part of which (foil. 46-58) is in Anglosaxon script, part

in Continental. The Anglosaxon pages shew a cross in the
left-hand top-corner, but not the other pages. The Chi-Rho
monogram appears in the Lindisfarne Gospels in this part of
the page, at the beginning of each Gospel.
In all countries however it was a common practice to write
an invocation at the beginning of a MS. ;
'
In nomine Domini
incipit ", '
In Christi nomine incipit ", Hv ovo|.ia{)i] Kvgi incipiunt
(Paris, Bibl. Nat. 13026; on fol. 12P). Indeed the Irish scribe
of Laon 26 has three such entries in his opening pages (fol. 2'') ;

Dei in nomine incipio; (fol. 4') In n«mine Dei summi; (fol. 5')
In nomine Trinitatis while on a page, written presumably on
;

New Year's Da}-, he sets (in Irish) Guard me, m}- Lord, from '

the dangers of the year '.

F. Distribution of Transcription.

Hver}-one knows that the distribution among several

scribes of the task of copying an exemplar led to blanks or
spaced writing or crowded writing (with much use of abbre-
viation-s}-mbols) at the end of a quire (where one scribe fi-
nished the allotted portion). There is no need to multiply
examples, though this practice should be remembered in cal-
culations of the pagination of an archetype. When a poetical
exemplar was transcribed, these awkward results could be
avoided b}- selecting for the transcript vellum of exactly the
size of the exemplar (or double the size which would enable ;

two pages of the original to fill one of the copy). Indeed

this laudable practice of making the transcript correspond,
page page, with the exemplar was not confined to poe-
for
The Douce Primasius shews, at the ends of pages,
tical texts.

sometimes a small blank space, sometimes a word or two of

'
overflow ', features which prove that this practice was followed.
 Collectanea varia 27

It is a MS. in Insular half-uncial: and one imag-ines thas its

exemplar must have been a MS. of note.

When a MS. is parcelled out between a few scribes, I
have usualh- found that the first part shews better writing
and more accurate transcription than the other parts, and
that there is often a great difference between the value of
part and part as a witness to the ipsa verba of the exem- ' '

plar. An editor of a Latin text notices these things and frames

his edition accordingly. But even if he is careful to state them
in his Preface, those saecli incommoda "
reckless emendators, ',

take Httle heed of them. It will be one of the problems for

twentieth-century editing to detect and determine the inter-
vention of a stupid and careless scribe in an archetype.
The the work of expert scribes in the same
similaritA' of
scriptorium makes us grateful for their not uncommon habit
of signing their names in the lower margin at the end (or
beginning) of their portion. (Where these signatures are all
b\' the same hand and at the beginning-page of a quire, we
ma\' refer them to the head of the scriptorium). The signa-
tures in the ninth century transcript 'in the \*atican Library
(Reg- 762) of the uncial Codex Puteanus
of Livy enabled
Traube to prove that the scribes were Tours scribes. (For
the latest treatment {') see Rand and Howe The Vatican '

Liv\- and the Script of Tours ',

in the IMemoirs of American
Academy in Rome, vol. I. 1917). The tradition that a Petro-
grad MS. (F V. I 11) of the Historia Tripertita was written by
the monks of Xoirmoutier in honour of their guest, the exiled
Adelard of Corbie, is confirmed bv the fact that the whole
writing-stafi of the scriptorium seems to have been employed
on it. For almost each quire has (or rather had, before a
modern binders interference) the signature of a difterent
scribe. In an Arras MS. too Regino (n.° 675; " saec. xi ") of
de Disciplina Ecclesiastica there seems to have been special
reason for the emplo\ment of a large number of scribes. For
the MS. has verses recording this:

Hoc non una manus offert, pastor, tibi munus. etc.

(*) But the signature g-ys/ar, with suspension-stroke through the arm
of r, is intended for Gyslarius (not Gyslarus nor yet Gyslanim nor, as an;

editor '
alienus a palaeographia '
made it, Gyslarx !).
28 W. M. Lindsay

The names stand at the top of the quires: Albertus '

scripsit ', '

Albertus finem fecit ', Richuinus scripsit \ Ri- ' '

chuinus hie Albertus reincepit ',

desinit Albertus hic
', . . .
' *

dimisit ', etc. But there is nothing to shew that the Cologne
nuns who wrote, in Hildebald's time, Cologne 63 and 67
had any special reason for signing so prominently (in the
lower margin), e. g. (63) fol. 86'' Girbalda scripsit ', fol. 263"
:
'

'
Agleberta (67) fol- 105''
scripsit Vera scripsit '
at end ;
' ',

'
Agnes Probably binders' have cut off most signa-
scripsit '.

tures of this kind in medieval MSS., so that we think the

practice (though examples are not rare) to have been rarer
than it reali}' was.
Of course, the usual reason for the employment of a great
number of scribes was that the transcript had to be made in
a great hurry. Since some twenty scribes' hands appear in
the 74 leaves of Alcuin miscellanea now at Cologne (n.° 106)»
some leaves in Anglosaxon script but most in Tours minuscule^
it is clear that this is the identical MS. which was prepared
at Tours in a hurr}' b}' Alcuin in 802 for Bishop Arno of
Salzburg (as W. Meyer shewed).

G. Dictation.

The few eccentric persons who still believe that our MSS.
were written to dictation cannot have read Wattenbach's re-
marks (in his Schriftwesen) on this delusion. Silence was
enforced in the monastery-scriptorium. Die tare (Germ, dichten)
means to compose in such sentences as (Engelberg 18 Augu-
' '
:

stini Confessiones " saec xii ") Hie Augustini liber est opus
;

ac PVoewini, Alter dictavit. alter scribendo notavit; (Vienna

743 Origen) Explicit liber quem dictavit Origenis. It means
'
to command
in the subscription of the Book of Armagh
"

(seeGwynn's edition) Ferdomnach hunc librum dictante '

. . .

Torbach berede Patricii scripsit Contra legere means to '. '

collate in such a sentence as that mentioned above, at the

"

'

beginning of section II. Legere means the same or to revise '

in the famous Juvenal entry Legente Aepicarpio, scrinbentis :

Exuperantio servo. Lector means, I think, the monastery-

teacher (in Irish fer leigind ') in the crj-ptograph in Munich,
'

Staatsbibl. 4115 ("saec. viii-ix ") Lex Salica: Ut potui sub

omni giletate {leg. agilitate) exaravi Ne lectoris animum tedio
 Collectanea varia 29

oppremerem (unless it refers to his future readers). In the

first picture in the Egino-Codex Egino can hardly be dictating
to the scribe; for Egino holds in his hands a book open at
the page In '
principio erat verbum while the scribe's page',

shews Audistis, "

fratres karissimi '.

Those who reh' on the confusions of words of similar

sound forget that in all transcription there is such a thing as
'
subconscious dictation '.

II. Probationes Pennae.

The purpose is avowed in e. g. Oxford, Bodl. 340

(•* saec. X "): probatio penne si bona sit. probatio incauxti [leg.

encausti) si bonum sit ; ibid. Laud. misc. 159 ('' saec. x "')
pro-
batio scriptoris.
To fill pages (to no profit) with examples would be easv.
I select a few worthy of mention. The use of Catonis Di-
sticha in the monaster\-schools explains the frequenc}- of the
couplet:

Si Deus est animus nobis, ut carmina dicunt, etc.,

e. g. Paris, Bibl. Nat. 8093 Anthology ("saec. ix'' Visigothic

minuscule of Lyons), fol So'; ibid. nouv. acq. 1750 the L}'ons
Pentateuch (uncial), fol. 226^; Amiens 88 Theodorus -Mopsue-
stensis (Corbie), fol 182'; Autun 107 Augustine (Visigothic
half-uncial of Burgundy), fol. 65", fol. 103\ fol. 120\ etc.; London,
Brit. Mus., Harl. 1772 Epistles and Apocalvpse ('* saec. ix "),

fol. 38\
In the last example the first line of Cato is followed by
the first line of Virgil's Eclogues (ignorantly written) :

Titure ti patule recubens sue tecmine fagis.

Mrgil's use in monastery-schools is attested by the same

'
probatio pennae in other MSS., e. g. the Fulda Agrimen-
'

sores "'saec. ix "), at end: T}tire tu patulae tenui musam. . . .

The conclusion of the Eclogues appears in the Essen Gospels,

fol. omnia vincit amos et nos cetmos amori (see above,
l^:
paragraph xA); in another Essen MS. Düsseldorf B 3; ab-type),
at end a b c d omnia vincit amor.
:

Special interest is attached to a probatio pennae in the ' '

bottom margin of fol l(f of Autun 27: Rure morans quid

30 \V. M. Lindsay

agam. This is the openino' of a charming poem, descriptive

of the of a countr^'-gentleman and scholar, which appears
life

in the best lamily of Martial MSS. (and in Latin Anthologies)

and which I would Mar-
fain attribute to the ancient editor of
tial (see mv
Ancient Editions of Martial p. 9). Its appear-
' ',

ance in a MS. of Burgundy may have some significance.

Indeed the list above suggests that a clue to provenance may
be furnished in these probationes '
'.

On the fly-leaf at the end of Durham B IV 24 ('' saec xi ") :

In his duobus versibus continentur omnes litterae abecedarii,

Equore cum gelido zephirus fert xenia kymbis

Y quod Habens signât mihi lex profitens katecizat.

Similary St Gall 913, Vocabularius S Galli, has on p. 89

two hexameters containing all the letters of the alphabet:

Te canit adcelebratque polus rex gazifer hymnus

Trans zephyrique globum scandunt tua fata per axem.

Those who like thrills mav read Dieterich's brilliant art-

icle '
ABC-Denkmaler (Rhein.' Mus. LVI [1901], p. 79) on the
'

magical significance of SATOR AREPO TENET OPERA

ROTAS, that forward-backward m3Ster3' of manuscripts' fly-
leaves-

V. — I Longa.

To the summary in Palae. Lat. I, pp. 26-29 and to the

fullertreatment in Loew's Studia Palaeographica (where the
' '

refusal to recognize I longa in Insular script is a hard saying)

I add the details which I happened to note in a number of
the MSS. used for my Xotae Latinae: (Where the script is
not specified, it is Continental minuscule. The words selected
are typical).

Albi 29 Synonyma Ciceronis, etc (" saec ix " ; Visigothic mi-

nuscule of Albi b}- more than one scribe
.'
; foil. 78) ;

In, lure, prolnde, elus, Ignoro, etc. Ille (by one scribe ;

only).
Amiens 6. 7, 8, 9 and 1 1 Maurdramnus Bible (Corbie ; by more
than one scribe) None.
18 Corbie Psalter ('' saec. viii ex. "')
None.
 Collectanea varia 31

88 Theodorus of Mopsuestia on the Pauline Epistles (Cor-

bie '' saec. ix None.
;

Autun 3 Gospels (Burgundy; 754 A. D.) In the minuscule pas-

sages In, luxta, hulus (sometimes). But not in i//e.
20 Gregorii Dialogi. Augustini Enchiridion (•* saec. viii-ix " ;

in the so-called N. E. France minuscule by more than :

one scribe; foil. 203) In, iuris, Ipse. But not in

eins, nie.
20'^ Cassiodori Expositio Psalmcrum (Burgundian minu-
scule; "saec. viii-ix") None. Other parts of this MS.
are Tro\-es 657 (it too has none), Lyons 402.
21 Gregorii Moralia (" saec viii " b}' more than one scibe
; ;

foil. 174) None.

23 Isidori Sententiae (" saec. viii-ix " : by more than one
scribe; foil. 1-13) None.
24 Cassianus de Capitalibus Vitus (Visigothic half uncial)
I longior, In. lam, lelunus.

107 Augustini Sermones super Psalmos

(half-uncial) None.
Bale F III de Natura Rerum et de Proprietate Ser-
15' Isidorus
monum (Anglosaxon minuscule of Fulda; saec. viii "; "*

by more than one scribe) None (except sometimes In).

F III 15 Isidorus de Natura Rerum \'itae Sanctorum ;

(Anglosaxon minuscule of Fulda; *' saec. viii"; by

more than one scribe foil. 45) None (except Ipse, not
;

however m, b\- one scribe).

F III 15'' Isidorus de Synon3-mis (Anglosaxon minuscule

of Fulda; "saec. viii"; by more than one scribe:

foil. 62) None, except (by one scribe) delude, elus.

F III 15^' Isidorus de Summo Incommutabili Deo. etc (from

Fulda; "saec. viii ex."; b\' more than one scribe;

foil. 66). I longior sometimes, In. Iudex.

F III 15" Isidorus de Natura Rerum Beda de Computo, ;

foli. 1-59 (from Fulda; "saec. ix ") None.

F III 15' Isidori Liber Difterentiarum (from Fulda; Insular

minuscule saec. viii by more than one scribe foli. 20)

; ; ;

None.
Bamberg E III 19 Scriptores Historiae Augustae (Anglosaxon

minuscule; "saec- ix ") In, lam (sometimes). Ita i^so-

metimes). But not eins, Ule.
M \ 12. part ii Jerome on Book of Proverbs (" saec. viii
32 W. M. Lindsay

ex. " ; None. This MS. belonged to Me-

foil. 101)
ginfrit. Charlemagne's chamberlain.
Berlin Hamilton 31 Concilium Aquisgranense (from Albi :

saec. ix) None-

Hamilton 553 Salaberga Psalter (from Laon; Anglosaxon
half-uncial: foil. 64) Sometimes In. lam, Ignis. But
not eins.
lat. theol. F
356 Gregorii Homiliae (written for Hildegrim
and presented by him to his brother Liudger's new
foundation at Werden; Anglosaxon minuscule; saec.
viii ex. foil. 96) In.
; But not mm, cìiÌìis, ignis.
lat. theol. F. 366 Pauline Epistles (written b}- or for Liudger ;

Anglosaxon minuscule; saec. viii ex.; foil. 94) In, lam

(sometimes). Ita (sometimes).
lat. theol. O 139 Gospels (from Werden Anglosaxon
; mi-
nuscule; "saec. ix"; foil. 145) In. lam. Ita (some-
times).
Phillipps 1667 Liber Sacramentorum ("saec ix in."; b}'
many scribes foil. 220
; In. But not Jniiiis.
Phillipps 1676 the Egino Codex (written at Verona saec. ;

viii ex.; foil. 309 In, ludeus, malor (sometimes. Ita

(sometimes).
Phillipps 1743 Concilia (from Rheims; "saec. viii"; by
more than one scribe; foil. 301) None except; that one
scribe prolongs / below the line in /;/, index, etc.
Phillipps 1825 Commodiani Instructiones (the '
codex An-
degavensis apparently rather Veronese
.'
'
;
" saec ;

viii-ix
" by man}' scribes foil. 46) In, lam, cuius. Ignis ;
;

(sometimes), Ille (sometimes).

Phillipps 1831 Beda de Temporibus (Verona minuscule;
saec ix in. bv more than one scribe foil. 143) In,
; ;

lam (sometimes), malor (sometimes). But not ipse.

Berne 167 \'irgil, with Servius' Commentar}- (with Breton
glosses; "saec. ix-x bv more than one scribe;'
;

foil. 214) None.

363 Horace, Virgil, etc. (Irish minuscule, probably written

in North Italv bv one of the Sedulius circle saec. ix ex. ; ;

foil. 197) In^

Boulogne 11 (10), foil. 1-92 Gospels (from Arras ; large Anglo-
saxon minuscule " saec. viii-ix ") In. ;
 Collectanea varia 33

4^ (44) Augustini Retractationes (written at St Bertin in

the abbacy of Nantharius, 804-820; foil. 77) None.

52 (48) Augustini Opuscula (written at St Bertin before
823) None.
63-64 (58) Augustini Epistolae (from St Bertin ; Anglosaxon
minuscule b\' more than one scribe) None.
;

66 (60) Augustinus c. Cresconium from St Bertin saec ;

''

ix in."; foil. 151) None.

Brussels 8780-93 Canons (perhaps from Stavelot; ''saec. viii ";
by many scribes) In, lam, malor (sometimes).
9403 Gregory of Tours (*' saec. viii-ix " b}' man}- scribes) ;

In, lohannes.
Cambrai 282 Augustinus de Trinitate (from Cambrai;
(300)
foil. 197) In. lam. conlungo (sometimes).
"saec. viii":
441 (470) Philippus" Commentar}- on Job (half-uncial, with
some Insular features by more than one scribe ; ;

foil. 205) In, lustus, elus, Ita, Ille (sometimes), etc.

619 (679) Canones Hibernenses (written at Cambrai, 763-
790, from an Irish script exemplar; foil. 75) None;
except that / is sometimes prolonged below the line
in in. 7 its fits.

624 (684) Gregory of Tours (uncial and half-uncial: b}'

more than one scribe; foil. 184) None.
836 (937) Isidori Liber Prooemiorum. de Natura Rerum, etc.

(uncial; foil. 70) In, luxta, etc.

Cambridge (Univ.) Ff W 42 Juvencus (Welsh minuscule :

" saec. ix " ; foil. 55) In, lustus, conlunx. Ipse.

Kk \' 16 the Moore Bede (from Le Mans; Anglosaxon
minuscule, c. 737; foil 128) In, lam. Ille (sometimes).
LI I 10 the Book of Cerne (Anglosaxon large minuscule;
foil. 99) In, delnceps, lustus. adiutor (sometimes). Ita
(sometimes).
(Corp. Chr. Coll.) 69 Homiliae in Evangelia (Anglosaxon
large minuscule; foil. 83) In, ludaei. Igitur (rarely).
But not. 77iaior.
144 the Corpus Glossary (from Canterbur}-; Anglosaxon
large minuscule; foil 65). The / is prolonged above
and below in in. coniimxit, ipse.
153 Martianus Capella (Welsh minuscule ;
" saec. ix post. ''
;

b}- many scribes; foil 86) Sometimes / longior in in.

3
34 W. M. Lindsay

173, part ii Sediilius (perhaps from Winchester; Anglo-

saxon minuscule : foil. 27) In, lugum (sometimes),
Ignis.
183 Bede's Life of St Cuthbert, etc. (Anglosaxon minu-
nuscule ; '' saec. ix in."; foil. 96) In. lussus, etc.
193 Ambrosii Hexaemeron (Corbie ab-script; b}' more
than one scribe ; foil. 170) In, ludico (sometimes), elus,
Ignis (sometimes).
320, part ii Canones Poenitentiales (probably written in

Brittany ; saec. ix "

54) None.
''
; foil.

334 Origenes super Lucam (Laon az-script b}' man}' scri- ;

bes; foil. Ill) In, loseph, cuius. Ipse (sometimes).

399 Juliani Toletani Prognostica (" saec. ix ant. " by more ;

than one scribe) None.

(Trin. Coll.) 216 Pauline Epistles (" de manu Baedae " An- ;

glosaxon minuscule; foil. 67) In (sometimes), Ita (so-

metimes.
Carlsruhe, Reichenau 57 Isidori Et\mologiae XIII-XX (written
in N. Italy, possibly Verona, in a unique type of earl}^
minuscule " saec. viii " foil. 90) In, lam (sometimes),
; ;

hulus (sometimes). Ignis (sometimes).

Reichenau 222 Apocalypsis et Epistolae Canonicae,
(1)

(2) PrimasiusApocalypsin ('' saec. viii ex. " and

in
''
ix in, by more than one scribe foil. 224) In, lam
"'
; ;

(sometimes), hulus (sometimes), Ille (sometimes).

Cassel, theol. F 49, foil. 1-87 Origen's Homilies on the Books
of Kings (from Fulda " saec. ix '') In, lustus (some- ;

times).
theol. F. 54 Liber Sapientiae, etc. (from F"ulda ; Anglo-
Saxon minuscule saec. ;
''
ix " ; by more than one
scribe ; foil. 76) In, ludico.
theol. F 65 Hegesippus (from Fulda library half-uncial ;

foil. 156) In, ludaei, hulus. Ipse (sometimes).

theol. O 5 Apocah-psis ; Homiliae in Apocal3'psin (from
"
Fulda librar}-; '*
saec. viii ; by more than one scribe;
foil. 82. None (unless i longior in m).
theol. O 2 Bedes History (from Fulda ; Insular minuscule ;

'•
saec. viii " ; foil. 50) None (unless i longior in in,

iuxta).
theol. O6 luliani Toletani Prognostica (from Fulda Anglo- ;
 Collectanea varia 35

saxon minuscule :
'
saec. ix " ; foil. 51) In. But not iatn,

Jmiiis. One
however writes sometimes lam, Ita.
scribe
theol. O 10 Augustinus in Librum Proverbiorum Homi- ;

liae (from Fulda " saec viii by more than one ;

'"
;

scribe foil. 142. Sometimes In, Iudex, etc.

;

Cheltenham 8400 Homiliary of Ottobeuren (Swabia saec. viii ") ;

''

In. ludei, hulus. Ipse, etc. But not ille.

12261 Augustini varia ; Commodiani carni, apolog. (pro-

bably the first part of Berlin Phill. 1825) In, lam,
cuius, delnde, But not ille^ iste, etc-

17849, part i Concilia (between half-uncial and minuscule;

" saec. viii " foil 62) In, hulus. etc. But not ibi, etc.
;

'"
part ii Concilia (minuscule of " saec. viii-ix foil. 32) None. ;

Colmar 38, part ijGospels (from Murbach ; Continental minus-

cule ' saec. " more than one scribe
of viii ; b}' ;

foil. 172) None.

part ii Murbach Anglosaxon minuscule of
Epistles (from :

"
''
saec.by more than one scribe foil. 66) None.
viii ; ;

39 Isidorus de Ortu et Obitu Patrum. de Mystica Nume-

rorum Sign, (from Murbach saec. viii " by more ;
'*
:

than one scribe foil. 180) In, lam. But not eins, ille.
:

40 Gregorii Cura Pastoralis (from Murbach " saec. ix " :

;

by many None. scribes)

82 Cassiani Collationes probably from Murbach saec. ix :
''

in " b}- more

; than one scribe foil. 75) None. :

Cologne (and, most of them, written at Cologne)

41 lohannes Constantinus in Ep. ad Hebraeos (time of
Hildebald by more than one scribe foil. 175) None.
: :

43 Hieronymi Praefationes in libros V. T. (between half-

uncial and minuscule " saec. viii " foil. 167) In lu- : :

deus, Ignis, etc. But not eins, ille.

51 Hieronymus in Ezechielem (time of Hildebald by many :

scribes ; foil. 208) None (but by one scribe In).

54 Hieron\-mi varia (time of Hildebald foil. 163) None ;

55 Hieronymus in Michaeam, etc. (time of Hildebald by ;

many scribes : foil. 144) None (?).

63 Augustinus in Psalmos (time of Hildebald : by three

nuns foil. 264) None.
:

67 Augustinus in Psalmos (time of Hildebald ; by many

nuns foil. 183) None.
;
36 \V. M. Lindsay
"'
76 Augustini opuscula {'" saec. vili ; bv more than one
scribe; foil. 222) None.
83" Isidori varia; Scholia in Aratea, etc. (written for Hil-
debald in 805 by many scribes ; ; foil. 219) None.
91 Canones saec
('' viii " ; by more than one scribe ;

foil. 112) None.

92 Grei^orii Epistolae (time of Hildebald by more than ;

one scribe foil. 180) None. ;

98 Isidori Ouaestiones {'" saec. viii med. " b}' more than ;

one scribe foil. 166) The / is prolonged above and

;

below in in, iiistus, cuius, ita (sometimes) etc.

106 Alcuini varia, foil. 1-47, 61-77 (Tours minuscule time ;

of Alcuin by ver}- man\- scribes) None (except by

;

one scribe In, lustus).

foil. 48-60 (Anglosaxon minuscule of Tours tiwe of Al- ;

cuin b}- more than one scribe) In, lustus. But not
;

/milts.

108 Homiliae Patrum (time of Hildebald ; b}' more than

one scribe) None-
165 Vitae Patrum (half-uncial) None.
210 Collectio Canonum Hibernensis ('' saec. viii
" by many
;

scribes ; foil. 151) In, lam, cuius. Ideo (sometimes).

212 Collectio Canonum (half-uncial ; foil. 170) None.
213 Collectio Canonum (Insular half-uncial ; by more than
one scribe foil. 143) None (.')•
;

Douai 12 Gospels (from Marchiennes Abbe\', near Douai ;

saec. viii-ix "

''
foil. 113) In, luxta. elus (rarel}'). But
;

not ipse, ille.

Dublin, Trin. Coll. s. n. Book of Armagh (written at Armagh,
808; Irish minuscule; foil. 222) In, Igitur.
Durham A II 16 Gospels (uncial, both of Continental and of
Insular type, and Insular half-uncial) In. luxta, Ite-
runi, etc. But not ille. One scribe has none.
A II 17, foil. probably
1-102 Gospels (Insular half-uncial ;

by more than one None. scribe)

B II 20 Cassiodorus in Psalmos (between Insular half'un-
cial and minuscule) The i is prolonged above and be-
low in in, iustitia. But not id, etc. One scribe has
none.
Einsiedeln 18 Commentarius in Psalmos ('' saec. viii-ix " ; by
 Collectanea varia 37

more than one scribe ; pp. 333) In, lustus, hulus. But
not ille.

21 Ascetica, foil. 1-24 ('" saec. viii ") None (except In so-
metimes).
157 Gregorius in Ezechielem (" saec viii " pp. 291) None. ;

281, pp. 1-178 -f 199, pp. 431-526 Ascetica C' saec. viii
"
med. ; bv more than one scribe) None.
347 Rufini Historia Ecclesiastica {"' saec. viii " pp. 458) None. ;

Epinal 68 Hieronymi Epistolae (from Murbach 744 A. D. ;

;

b}' mam' scribes foil. 209 In, lustns (sometimes), eins

;

(sometimes). But not ilk (tough one scribe prolongs

the Ì below).
Essen Gospels ('' saec. ix in
"'
; by many scribes ; foil. 170)
the i is prolonged above and below sometimes in in^
iam^ ems.
Florence Am. 1 Codex Amiatinus (witten at jarrow, 690-716 ;

uncial; foil. 1029) None.

Fulda D 1 Codex Theodosianus ; Formulae Andecavenses
(from Constance viii " b}' more than one
''
; saec. ;

scribe Sometimes i longior in in, iam.

; foil. 185)
Bonif. 2 the Ragyndrudis Codex (Luxeuil type by more ;

than one scribe) In, lam, malor (sometimes), Iste (so-

metimes). But not ille.

Glasgow Hunter T IV 13 Medica varia ('' saec. viii-ix " ; by

more than one scribe) In, etc.
Gotha I 85 Canones Murbacenses (" saec. viii-ix ''
; by more
than one scribe) None.
Hague '
Fhilippus Commentary on Job (probabh- from Metz
'
;

"saec. viii med- "; by ver}- man}- scribes; foil. 199)

In. Iudex, elus (sometimes), Ita (sometimes).
Hereford P I 2 Hereford Gospels (Insular large minuscule ;

foil. 135) None.

Ivrea 42 Bede de Temp., etc. ; Canons (" saec. ix in. " ; foil. Ill)
Sometimes In.
99 Gospels (North Italian minuscule " saec. viii-ix ") ; In,
lacob. But not eius.
Laon 26 (flyleaves) (Irish minuscule of '' saec viii-ix " In, lu-
stus, etc.
81 Joh. Scottus on St John's Gospel (Caroline minuscule ;

saec ix med. ; by more than one scribe) None.

38 W. M. Lindsay

96 dementis Expositiones (Laon minuscule ;

'"
saec. ix " ;

68) None.
foil.

137 Orosius (Laon az-script ; foil. 130) In, Joseph, etc. But
not Cialis, tôt, etc.
201 Glossary Canons (written at Cambrai, 831-863, by
;

many foil. 112) None.

scribes;
319 Taionis Sententiae (Laon minuscule " saec. ix in. ;
"'
;

foil. 180) In. lam (sometimes). But not Indus, ipse.

328*"' Cassiani Institutiones (" saec. ix " by more than ;

one scribe foil. 146) None. ;

423 Isidorus de Natura Rerum (Laon az-script h\ more ;

than one scribe; foil. 79) In, prolude. But not iove,
mills, ipse. etc.
Leyden 67 Prisciani Periegesis et Grammatica (from Egmont
Abbe}' ; written in 838 in Irish minuscule by many
scribes ; foil. 218) In, hulus (sometimes).
67 F Glossary (" saec. viii-ix " ; by many scribes) In, lure
(sometimes), elus (sometimes), etc.
114 Codex Theodosianus (from Rheims ''saec. ix in " ; ;

by many scribes) In, Iudex, Ipse (sometimes).

Scaliger 28 Bede (from Flavigny in the diocese of Autun,
with Insular abbreviation saec ix in. b\- more than ; ;

one scribe) In (sometimes).

Voss. F 26 Glossaries (from Ghent ;
" saec. viii-ix " foil. 48)
None.
Voss. F 73 Nonius Marcellus (Tours minuscule of saec ''
ix
ant. bv many scribes) None.
''
;

Voss. O 5 Ydacius (apparent!)- from St Gall saec. ;

''
ix
in. ") In, lustus.

Voss. Q 60 Liber Pontificalis (from Rheims ;

" saec viii-

ix " ; foil. 122).

Voss. Q
63 Gregory of Tours (" saec. viii by many "

scribes) In, lubeo. Ignoro (sometimes). But one scribe

has none.
Voss. O 110 Eusebii Chronica (written at Micy, c. 850;
foil. 166) Sometimes In, losephus, Trolanus, Ita.

Lichfield Gospels (Welsh half-uncial; foil. 110) None.

Liege 306 Jerome's Ouaest. Hebr. Bede on Books of Kings ;

(from St Trond 834 A. Ü. by more than one scribe

; ;
;

foil. 107) In (by one scribe).

 Collectanea varia 39

London Add. 5463 Gospels (S. Vincenzo al Volturno ; l^f^-l&ò ;

uncial) None.
Add. 11878 Gregorii Moralia (Luxeuil t\pe ; foli. 78) In,
But not tile.
ludicio, elus. Igitur, alt.
Add. 118S0 Vitae Sanctorum (perhaps from Bavaria;
*'
saec. ix " ; foil. 240) In (sometimes).
Add. 18332 Theologica varia (Carinthia r " saec. ix post " ;
;

b}' many scribes foil. 178) Sometimes In. ;

Add. 24143 Gregorii Moralia (Merovingian minuscule :

"'
'*
saec. viii post. ; foil. 59) In, ludaeus.
Add. 29972, 26-38 Augustini Sermones (Luxeuil type)
foil.

In, cuius (sometimes). But not il/e.

Add. 30S52 Orationale (^X. Spain Visigothic minuscule : ;

" saec- ix ex. "; by more than one scribe; foil. 115)

In, ludicium. elus, Ille.

Add. 31031 Gregorii Moralia (Laon az-script by more :

than one scribe foil. 145) In, lam. Ignoro. But not
:

ille. cuius.
Cotton Cal. A xv, fol. 1-117 Jerome; Computus, etc.
France 743 A. D. b\- more than one scribe) In,
; :

delnde, ludaea elus. :

Cotton Xero A ii, toll. '14-45 (\'erona 767 A. D. In. .-

: .-)

But not ins tus, etc.

Cotton Tib. A xiv Bede's Histor}' (Anglosaxon minus-
cule :
" saec. viii
foil.
'"
201) In, lubeo (sometimes).
;

Cotton Vesp. B vi, foil. 104-109 Miscellanea (Anglosaxon

minuscule of Mercia 811-814) In. lam, etc ;

Egerton 609 Gospels (from Marmoutier, near Tours ;

" saec ix in. by more than one scribe foil. 102)

'"
; ;

In (sometimes).
Egerton 1934 Fragment of Isidori Pacensis Chronicon (Vi-
sigothic minuscule :
" saec. ix in. " foil. 2) In, Iudex, ;

hulus. Ignis.
Egerton 2831 Jerome on Isaiah (Continental and Insular
minuscule of Tours saec viii " foil. 143) In. luxta, ;
'"
:

elus (sometimes).
Harle\- 2965 Book of Xunnaminster (Anglosaxon half-un-
cial or large minuscule; foil. 40) In.
Harley 3034 Isidore extracts : Augustin's Enchiridion, etc.
(Middle Prankish region "
.-
:
"'
saec. viii ; fol. 96) Xone.
40 AV. M. Lindsay

Harley 5041, part i Theological Tracts (Merovingian mi-

nuscule ;
" saec. vii "'
foli. 8) Sometimes lam, malor.
;

But not (1). m

Harley 7653 Litany (Anglosaxon half-uncial or large mi-
nuscule ; foil. 7) In.
Royal 2 A xx Lectionary (Anglosaxon half-uncial or large
minuscule ; by many scribes ; foil. 52) In, Iuris, Ipse.
Royal 1 B
Gospels (Anglosaxon half-uncial and mi-
vii

nuscule " saec. viii "

; probablv by more than one ;

scribe foil. 155) None.

:

Royal 1 E vi Canterbury Gospels (Anglosaxon half-uncial

or large minuscule by many scribes foil. 77) In, luro,
; :

malor, alt, Itaque (sometimes). But not ilk.

Lons-le-Saulnier Bede (St Claude. Jura; 804-815; foil. 232)
None.
Lucca 490 Eusebii Chronica, etc. (written at Lucca, c. 800 ;

uncial and minuscule of various types) In. delnceps,

luxta. Ita (sometimes). But not eins (b}' one scribe).
Sometimes malor (by one scribe).
Luxemburg 44 Gregorii Dialogi (from Echternach " saec. i.x ;

•'
in. foil. 129) None.
;

68 Augustinus in Psalmos (from Echternach " saec. viii- ;

ix " probably by more than one scribe foil. 142)

; ;

None.
Lyons 378 449) Bede on Books of Kings (written at Lyons,
840-852) None.
523"" (607) Augustini Civitas Dei (the ancient minuscule
of the marginalia) In, lam, hulus, always after r.
524 (608) Augustinus de Natura et Gratia, etc. (written
at Lyons saec. viii-ix foil. 168) None.
; :

526 (610) Augustinus contra Faustum (written at Lyons ;

saec. viii-ix foil. 155 None. ;

Maeseyck Gospels (Abbey of Alden-E}ck Anglosaxon half- ;

lam (sometimes). Ignis (sometimes).

uncial) In,
Maihingen Gospels (Anglosaxon half-uncial of Echternach ;

foil. 157) In (sometimes). Ignis (sometimes).

Manchester 15 Cypriani Epistolae (Murbach :
''
saec. viii " ;

by many scribes foil. 203) None. :

194 Origenes in Epist. ad Rom. (Beauvais :

" saec. ix " ;

foil. 163) In, ludaeus.

 Collectanea varia 41

Metz 7 Bible (Metz saec. viii-ix " foli. 359) None.

;
**
:

76 Prophetae (Anglosaxon minuscule " saec. ix " by ; ;

more than one scribe) In. Iudex (sometimes). But not

eins.
134 Theolo^ica varia (Metz ''
saec. vili
" ; by more than
;

one scribe) Sometimes / longior in in.

Milan (Ambr.) C 301 inf, Theodorus" Commentar}' on Psalms
(Irish minuscule of Bobbio " saec. viii ") In. ;

H 150 inf Victorius Aquitanus (from Bobbio c. 810 by ; ;

more than one scribe) In, exinde, lam. But not cuius.
(Trivulz.) 68<S Juliani (X. Ital. minuscule Epitome
before ;

c. 800) But not eins.

In, delnde, lus.
Modena O I 11 Isidori Chronica; Medica varia (N. Ital. mi-
nusc 800 A. D. Sometimes In, delnde, lubeo. But
;

not huiics.

O I 17 Isidore's Etymologies (N. Ital. cursive :

''
saec. viii

med. ") In, lus, conlunctus. Ipse, etc.

Montpellier 55 Passiones Sanctorum (Burgundian minuscule ;

" saec. viii-ix " ; by many scribes) None but by some :

scribes In, lam.

61 Gregorii Homiliae (from Troyes ;
""
saec. ix ") None.
"
141, foil. 1 80, 95-135 Alcuin. etc. (" saec. ix in. : b}- more
than one scribe) None.
409 Charlemagne's Psalter (from Auxerre saec. viii ex) ;

None.
Munich (Staatsbibl.) 3514 Passionale (from Augsburg uncial, ;

with parts in large minuscule by more than one ;

scribe pp. 307) In. But not eins, ita.

;

3731 Gregorii Homiliae (from Augsburg Anglosaxon mi- ;

nuscule " saec. viii

I by more than one scribe "

; ;

foil. 91) In, ludea. But not eins.

4542 Gregorii Homiliae (Kisyla group by more than one :

scribe foil. 256) None.

;

4547 Sermones (Kisyla group foil. 250) None. ;

4549 Cassiani Collationes (Kisyla group by more than ;

one scribe foil. 141) In. But not. inre. huiits.

;

4554 Passionale (Kis\la group b\' more than one scribe :

;

foil. 164) None.

4564 Alani Farfensis Homiliarium (Kisyla group b}' more ;

than one scribe foil. 244) None. ;

42 W. M. Lindsay

4577 S. Pauli Epistolae, etc. (Kisyla group foil. 95) None. ;

6224 Freising Gospels (uncials of Italy ? foil. 251} None. ,

6237 Gregory on Ezechiel .written at Freising by Fere-

grinus Anglosaxon minuscule and Bavaria minuscule
; ;

saec. viii post foil. 186) In, lam, adiungo, Jgitur (so-
;

metimes).
6239 Tobias etc. (F"reising ''
saec. "
job, ; viii ; foil. 103)
None.
6262 Hrabanus !*tlaurus in Paralipomena (written at Freis
ing, 854-875 ; foil. 147) None.
6273 Ambrosius in Lucam (written at Freising, 812-834 •

foil. 224) None. (In sometimes.?)

6278 Gregorii Moralia (from Freising uncial and large ;

minuscule saec. viii "

; by more than one scribe
''
:
;

foil. 131) Sometimes lon^ior in in, iam. But not ^zW.

•/

6298 Augustini Homiliae (from Freising Anglosaxon large ;

minuscule " saec. viii

'"
; by more than one scribe ; :

foil. 114) In.

6382, foil. 44-172 Gregorii Moralia (Freising; saec ix ant.)

None.
6402 Juvencus (from Freising " saec. viii by more than . ;

one scribe foil. 63) Apparenti}- / longior in m, iam^

;

conuix (sometimes).
14080 Isaiah and Jeremiah (from Ratisbon Anglosaxon ;

large minuscule or half-uncial, with passages in Conti-

nental minuscule "saec viii'" foil. 112) In. Butnot^/V^j-.
; ;

14210 Rabani Mauri Inst. Cleric. (Anglosaxon and Conti-

nental minuscule of Ratisbon; saec. ix "; b}' many ''

scribes; foil. 127 In, perlnde, lelunium. Item (some-

times). But not ciiis.
14437 Augustinus in Epist. I S. Johannis (written b}' two
Ratisbon scribes in 823 foil. 109) In. :

14468 Theological Tracts (written at Ratisbon in 821 ;

foil. 112)None.
14653 Augustinus in Johannem (from Ratisbon Anglosa- ;

xon large minuscule saec. viii ") In, delude, lam

;
''

(sometimes). But not eins, ipse.

19101 Gospels (from Tegernsee late uncial) None. ;

28118 Regulae Patrum (written at Treves, or else Aniane;

saec. viii ex. by more than one scribe foil. 215) None.
; ;
 Collectanea varia 43

(Univ.) 4-to 3 Sulpicius Severus by many (" saec. viii-ix " ;

scribes Sometimes / long'ior in m.

; foil. 124).
8-vo 132 Leges Baiwariorum (Bavarian minuscule ;

saec. ix in "
^'
foil. 87.) Sometimes / longior in in.
;

Namur 11 Bede's Histor}' Gregory of Tours (from St Hubert ;

in the Ardennes " saec. ix " b}' man}' scribes) Kone;

; ;

but b}' one scribe sometimes In, lus.

Orléans 193 (221) Canons (Breton by more than one scribe) ;

None.
255 (302), pp. 1-82 Sedulius (Breton ; in Insular half-uncial)

None.
Oxford Auct. D. II 19 Macregol (or Rushworth) Gospels (Irish
half-uncial of about 800 ; by more than one scribe)
SoMietimes i longior in in.

Auct. F IV 32, foil. 19-36 Liber Commonei (Welsh minus-

cule of probabl}' 817) In. lustus.
Canon Patr. lat. 112 (written at Corbie; ''
saec. ix in. ";
by more than one scribe) None.
Digb}' 63 Tracts on the Paschal C\'cle (from Winchester ;

written about the middle of the ninth century in a

curious minuscule of Insular type) In, lure. But not
cuius, ipse.
Douce 176 Evangeliary ('' saec. ix ") None.
Junius 25 Ethici Cosmographia Glossaries, etc. (from Mur- ;

bach " saec. viii "

; by a great number of scribes, ;

possibly not all contemporar}) None except some- ;

times / longior in in.

Lat. theol. d .3 Commentary on the Pentateuch ('' saec viii-

ix " ; by many scribes) In, loseph, Ignis, Ipse, Ille (so-

metimes). But not cuius.

Laud- lat. 22 Maccabees (with German glosses " saec. ix ") ;

None.
Laud. lat. 92 Deuteronomy, etc. (written at Würzburg, 832-
842, in Anglosaxon minuscule) In, lam, elus.
Paris 528 Theologica Varia (from Limoges " saec. ix ; in. "
;

by man}- scribes) In, luro sometimes). But not huius.

653 Pelagius in S. Pauli Epistolas (written in North Italy ;

''
saec. viii ex. by more than one scribe) In, lohan-
'"
;

neS: Ipse, Ille.

1012 Gregorii Opuscula (from Limoges; "saec. viii-ix";

44 W. M. Lindsay

by more than one scribe) In, delnde. But not eius^

iustus.
1451 Canons (from St Maur-les-Fossés ; written in 796 b}^
more than one scribe) None b}' the first scribe. The
second uses i longior in in.

1603 Canons (fi-om St Amand ;

" saec viii ex. " ,• by many
scribes) In.

1771 Augustini Opuscula (" saec. ix in.

" foil. 1-30 in An-
;

glosaxon minuscule) In, Iudex, elus (sometimes), Ipse

(sometimes).
1820 Jerome on Isaiah (from Micy; '' saec viii post " ; by
many scribes) In, luda, elus.
1853 Hieronymus in S. Pauli Epistolas ('' saec. viii " by ;

more than one scribe) None.

2109 Eugippius (written at St Amand under Lotharius
scriptor, saec. viii-ix, by more than one scribe) None.
2110 Eugippius (written in a script between half-uncial
and minuscule of "" N. E. France " by more than one
scribe probably " saec. vii-viii ") None.
;

2123 Liber Pontificalis; Marculh Formulae, etc (of 795-

816; b}' many scribes) None.
2341 Computus; Liber Comitis, etc. (from Le Fuy in Haute
Loire.; written before 843 by many scribes) None.
2706 Augustine on Genesis (half-uncial of N. E. France " ; ''

by man}' scribes None.

2739 Theologica varia (Merovingiam minuscule of " saec-
vii-viii ") In, laceo. Ideo, elus (sometimes).
2796 Gregory's Homilies; Computus; Canons, etc. (written
in 813 by many scribes) None.
2824 Isidori Prooemia, etc. (written in an early form of
the Corbie ab-script by more than one scribe; saec ''

viii ") In, lubeo, elus, delnde.

3837 Canones Apostolorum (apparenti}' written at Angers
before 829 by more than one scribe) I longior in in
iiotctus, suòiaceo.
4403 Codex Theodosianus (''saec. viii"; by more than
one scribe) In, Iudex. But not huius.
4403* Codex Theodosianus (Corbie en-t}-pe by more than ;

one scribe) In, lam, malestas, Igitur (sometimes).

 Collectanea varia 45

4403" Codex Theodosianus C saec. viii ex. ") Sometimes

elus (projecting below also).
4404 Breviarium Alarici (probably b}' more than one
scribe; 803-814) None.
4413 Breviarius Alarici written at Ba}eux in 833) None.
4568 Juliani Constitutiones (written apparently in Italv;
" saec. viii-ix ""
In, lustus, malor (sometimes), Ipse

(often).
5543 Dionvsius Exiguus, etc. (written, perhaps at Fleur}-,
in the middle of the ninth century hv more than
one scribe) None.
6400^. foil. 112-193 Isidorus de mundo, etc. (uncial) None.
6413 Isidorus de natura rerum, etc. (uncial) None.
9380 Theodulfus Bible (from Orleans, 788-821) In. lohan-
nes. elus. Ipse.
9382 Prophetae (written b\' Xergilius of Echternach in an
Anglosaxon script between minuscule and half-uncial)
None.
9389 Codex Epternacensis (Insular half-uncial and large
minuscule) In, lam. But not eÌ2és. ipse.
9427 Luxeuil Lectionar}- (Luxeuil t3-pe) In, lulianus. elus
(sometimes), Ille (sometimes).
9451 Liber Comitis (in silver and gold uncial and large
minuscule apparenti}' from the same scriptiorum as
;

n.° 653) In (sometimes).

9517 dementis Recognitiones (from Beauvais; not after
8-iO; by more than one scribe probably) None.
9527 Jerome on Isaiah (from Echternach Anglosaxon mi- ;

nuscule of " saec. viii med. " by more than one

;

scribe) In, Iudex (sometimes». Iste (sometimes). But

not ei?(s.
9530 Jerome on Matthew (from Echternach; saec. viii- *'

ix ''; by many scribes) In, ludaea (sometimes). Ita

(sometimes). But not cidus. One scribe does not
use i longa at all
9565 Taius Samuel (from Echternach Anglosaxon rude :

minuscule of saec. viii "'; by more than one scribe)

*'

Sometimes In. lam. Ipse. But not ciims^ ille.

9575 Claudius Taurinensis (apparently the authors own
cop}-; written at Poitiers in 811) None.
46 \V. M. Lindsay

10588 Canons (" saec. viii "')

None.
11504-5 Bible (written probably at St Rlquier in 822 by
more than one scribe) None.
11533 Bible (from Corbie; written in 850 by more than
one scribe) None.
11631 Jerome's Letters ('" saec. ix in. "; probably by more
than one scribe) None.
11738 Eusebii Historia Ecclesiastica (written at St Maur-
les-Fossés, c. 840, by more than one scribe) I longior
in in.
12048 Sacramentar}- of Celione (written at Rebais, in the
diocese of Meaux, about 750, in a script between
French half-uncial and minuscule) In (sometimes), Ita
(sometimes). But not mstiis, eins.
12050 Corbie Sacramentary (shortly after 853) None.
12097 Canones (from Corbie; saec. vi half-uncial and un- ;

cial) In the half-uncial portion In, luxta, elus (some-

times).
12168 Augustini Quaestiones in Heptateuchum (from Cor-
bie: Laon az-script; b}' more than one .scribe) In,
lohannes, delndc. But not (or seldom) ipse.
12240 Cassiodorus in Psalmos (written at Corbie, " saec
viii ", by more than one scribe) In, malor.

12254 Gregorii Homiliae (from Corbie librar\-, but written

in minuscule of Visigothic type, presumabl}' in France,

by more than one scribe ''

saec. ix '') In, elus, Igitur.
;

12281 Beda in Lucam (''saec. ix in. "; with Insular ab-

breviations) None.
12598 Vitae Sanctorum (written at Corbie; ''saec. viii";
by many scribes) None. But in the Merovingian '
'

minuscule of foil. 47-53 In, lelunus.

13026 Grammatica Varia; Prudentius, etc. (" saec. ix in. ";
b}' many scribes) None. But (b}^ one scribe) In.
13028 Isidori Et)'mologiae (Corbie en-t^pe) I longior in
zam.
13029 Smaragdi Grammatica (from Corbie library, but
with Breton glosses; " saec. ix "; b}' more than one
scribe) None.
13047 Juvencus ; Patristica Varia (written at Corbie ;
" saec.

viii " ; b}' more than one scribe) None (but rl, tl).
 Collectanea varia 47

13159 Charlemagne's Psalter (small uncial of 795-800) In,

lustus.
13246 '
Bobbio '
Sacramentary (rude uncial half-uncial,
large minuscule) In, lam, elus sometimes). Ita (so-
metimes).
13373 Orosius Augustine Alcuin Bede (written at Cor-
; ; ;

bie,between 817 and 835, by many scribes) None-

13386 Peregrinus (" saec. viii " b}' more than one scribe)
;

In, ludico, delude.

13729 Liber Pontihcalis (written in 824-827) I longior in

in (sometimes).
";
16668 Bede: Aldhelm, etc- (from Lorsch: "saec. ix
partly in Anglosaxon minuscule, partly in Caroline)
In the Anglosaxon part. In, lustus, malus, Ita.
17227 Gospels (written by Adalbald of Tours before 834)
None.
17371, foil. 1-153 Jerome on Jeremiah (written at St Denis,
793-806, by more than one scribe) None.
17416 Fulgentii Opera (from Compiègne library; written
before 827 bv manv scribes) None by the first scribe.
B_v the second In, luro (sometimes), Ignoro (sometimes).
17451 Beda in Lucam (from Compiègne :
•'
saec viii ex. ")
None (except in the leaves written in the Corbie ab-
script).
17655 Gregory of Tours "saec. vii";
(from Corbie:
Merovingian minuscule of cursive type) Prolonged
below and above. In, ludico. But not ipse.
Nouv. acq. 445 Augustini opuscula (written at Tours b\' Adal-
bald) None.
Nouv. acq. 1203 Godescalc Gospels (uncial of 781) None.
Nouv. acq. 1575 Eugippius' Extracts from St Augustine
(from St Martin's, Tours: Merovingian minuscule saec ;
'*

viii in. by many scribes) None by one scribe. B\'

'"
;

another, i longior in JniÌ7is. etc.

Nouv. acq. 1587 Gatien Gospels (from St Gatien's Tours ;

rude imitation of Insular half-uncial) None.

Nouv. acq. 1597 Paterius (from Fleury " saec. viii partly ;
""
;

in a script between French half-uncial and minuscule,

parti}- in minuscule of various t}'pes) None.

Nouv. acq. 1619 Oribasius Medicus (from Fleury saec. ;
*'
48 W. M. Lindsay

vii-viii " ; between French half-uncial and minuscule).

Projected below and above, In, lobet, elus (some-
times).
Nouv. acq. 1740 Pentateuch (from Lyons; None. uncial)
Rome (Vat.) 583 Gregor3''s Moralia (written apparently in
German}- " saec. ix in. ") None.
;

1480 Priscian (with Breton glosses; ''saec. ix ") None.

1512 Tib. CI. Donatu's Commentar}- on Virgil (" saec ix
in ") None.
3321 Glossary, etc. (uncial; Central Italy apparently;
''
saec. viii None.
ant. "")

5763 Isidori Etymologiae (written probably at Bobbio, in

N. Italian cursiv^e of " saec. viii in. ") In, conlunctus.
5764 Et}'mologiae (written apparently at Verona;
Isidori
"saec- by more than one scribe) I longior,
ix in.";
In, luxta, cuius (sometimes).

5775 Claudius Taurinensis in Epp. S. Pauli (written for Bp.

Teuduifus ol Tortona, near Alessandria, in 862) None.
Barb. 570 Wigbald Gospels (Anglosaxon half-uncial; by
many scribes) In. lus (sometimes). Ignis (sometimes).
Pal. 68 Commentary on Psalter, with Irish and Northum-
brian glosses (written in some Northumbrian mona-
stery in Insular minuscule of ''
saec. viii ex. ") In.
Pal 161 Lactantius (written at St Amand
under Lotharius
scriptor, saec. viii-ix, by man}- scribes) None.
Pal. 169 Ambrose on Pauline Epistles (from Lorsch; " saec.
ix in."; by more than one scribe) None (except by
one of the scribes).
Pal. 172 Jerome on Isaiah (from Lorsch; ''saec. ix"; by
more than one scribe) None.
Pal. 177 Jerome on Matthew 'from Lorsch; Anglosaxon
minuscule of " saec. ix in. "; by more than one scribe)
In, lure, cuius sometimes), Ita (sometimes).
i

Pal. 195 Augustine (from Lorsch, part in Anglosaxon mi-

nuscule, the rest in Continental; ''saec. ix ") In the
Anglosaxon part, In (sometimes), lohannes (someti-
mes) but not eÌ7ts. In the other, i longior in in, iam.
;

Pal. 201 Augustinus c Faustum Manichaeum (from Lorsch ;

" saec. ix " by more than one scribe) In, lam, prolnde.
;

But not ehis.

 Collectanea varia 49

Pal. 202 Augustinus de Trinitate (Anglosaxon minuscule,

perhaps of Lorsch " saec. viii-ix " by more than one
; ;

scribe) In, lam, delnde. Ita (sometimes).

Pal. 220 Augustini Sermones, etc. (from Lorsch; Anglo-
Saxon minuscule of saec ix in. '') None.
**

Pal. 237 Prosper (foil. 46-5S in Anglosaxon minuscule, the

rest in German of Maintz apparenti}-; ''saec. ix in.")
In the Anglosaxon part In ludaei. Ita (sometimes).
In the rest. In.
Pal. 238 Prosper (from Lorsch " saec. viii-ix ") In. lam
;

(sometimes; rather / longior). But not cuius.

Pal. 249 Gregorii Moralia (from Lorsch; "saec. ix"; by
many scribes) None by one scribe. B\- another. In,
lugum, Igitur.
Pal. 259 Gregorii Homiliae (apparently Anglosaxon large
minuscule of " saec. viii in. by many scribes) In.
""
;

Pal. 266 Gregory (from Lorsch saec ix '') In, Iudex.

;
*'

But not eins.

Pal. 554, foil. 5-12 Theologica varia (Anglosaxon minus-
cole of '' saec. viii '") In, lus. But not mains.
Pal. 574 Canones (from Lorsch; "saec. viii post.") In,
lungo. But not eins.
Pal. 577 Canones (from Maintz Anglosaxon minuscule of
;

"saec viii"; probabh- b\' more than one scribe) In

(sometimes), leluno (sometimes). But not huiiis.
Pal. 822 Rufinus (from Lorsch: '"saec ix ant. "; by many
scribes) In, lustus (sometimes, hulus (sometimes). Ita
(sometimes).
Pal. 834 Beda de Temporibus, etc. (from Lorsch; written,
probably in 836, b}- more than one scribe) In, delnde.
But not ignis.
Pal. 845 Sulpicius Severus. the part in Anglosaxon script
(from Maintz; "saec ix ") In, lam. Ignoro (someti-
mes). But not Cîmis.
Pal 1547 Seneca de Benefìciis (from Lorsch librar}', but
perhaps written in Italy. " saec. viii-ix by many ',,

scribes) In, Iudex, malor (sometimes). Ita (sometimes).

Pal. 1588 Rhetorica varia (apparenti}* written at Lorsch,
" saec. ix " by man}- scribes) None (but some scribes
;

appear to use / longior in in^ index).

50 W. M. Lindsay

Pal 1753 Grammatica varia (apparently written at Lorsch ;

" saec. ix " by man}- scribes) In, lungo, hulus (so-

;

metimes), Iterum (sometimes).

Reg. 124 Rhabanus Maurus (written atFulda before 847
b}' many scribes) None b}' the second scribe. B}'
the first, (sometimes) but not iustus^ eins-
In, Ita ;

Reg. 296 Orosius, with Breton glosses ('' saec. ix; by more
than one scribe) None.
Reg. 316 Sacramentarium Gelasianum (late uncial; b}'
man}' scribes) None-
Reg. 762 the Tours transcript of the Codex Puteanus of
Livy. None (except possibly / longior in in^ iani).
Reg. 846 Juristica varia (from Orléans written before 814, ;

probably by more than one scribe) None.

Reg. 1209 Alcuin (Anglosaxon minuscule of '' saec. ix ";
by two scribes) In, lus. Ita (sometimes). But not eius.
St Gall 70 (written b}- Winithar, 760-761 In, Iudex. But not
ilk.

110, pp. 275-510 Breviarium Apostolorum (written pro-

bably at Verona, " saec. ix ") In, lus, elus.
193 Caesarius (" saec. viii ex. " by more than one scribe)
;

None, except / longior in in (sometimes).

194 Caesarius (" saec. viii ") None, except /longior min.
213 Gregorii Dialogi; Augustinus (''saec. viii") In. But
not ins, eins, ita.
214 Gregorii Dialogi (Merovingian minuscule by two scri- ;

bes) In, elus (sometimes), Iste (sometimes). But notille.

227 Isidori Sententiae (North Italy; saec. viii "; by more
''

than one scribe) In. lob, Ipse (/ longior). But not

huius, ille.

348, pp. 32-376 Sacramentarium Gelasianum (Chur; c. 800)

None.
555 Adamnani Vita S. Columbae (851-872 by more than ;

one scribe) Apparently i longior in in.

722 Breviarium Alaricianum (Chur; 800-820; by more than
one scribe) None.
731 Leges Visigothorum, Salica et Alamannorum (written
— where? — in 793 or 794) None.
761 Hippocrates (Anglosaxon minuscule of saec. ix ") ''

In, hulus. Ipse.

 Collectanea varia 51

904 Priscian (written in Ireland in Irish minuscule between

844 and 869 by. practically, two scribes) Jn, lungo,
conlung-o.
St Omer 15 Hieronymi Breviarium in Psalterium (from St Ber-
tin :
•'
saec. ix in. "; by many scribes) In (sometimes),
ludaei (sometimes). At least b\' one of the scribes.
279, foil. 1-2 (from St Bertin : Insular half-uncial or large
minuscule " saec- viii ") In.
St Petersburg F I 3, foil. 39-168 Hieronymus in Isaiam (from
Corbie; Anglosaxon half-uncial of " saec. vii-viii ": bv
many scribes) In. ludaea, Idola.
F I 5 Tripertite Psalter (Leutchar-script of Corbie by more
;

than one scribe) In, lustus.

F I 6 Ambrosius in Lucam (Leutchar-script of Corbie) In.
F I 7, part Selected Letters of Gregory (sent by Paulus
i

Diaconus in X. Italv to Adelhard of Corbie) In, cuius.

Ita.

O I 14 Gregorii Homiliae (from Corbie ; Merovingian a-script,

a predecessor of the Laon az-script; "saec. vii": In,
lam. But not ille.

Q 1 18Bedae Historia (Anglosaxon minuscule saec. ;

'*
viii
"':

by more than one scribe) In, lam, adiuncto. But

not ipse.

O I 19 Rufini Expositiones (written at Corbie during Adel-

hards degradation) I longior, In (sometimes). Iudex
(sometimes).
Q XIV 1 Paulinus Xolanus (from Corbie ; Anglosaxon half-
uncial and large minuscule) In, lam. But not huins^
ipse.

O 4 Cassiani Collationes (Corbie en-type) None-

1

Schaffhausen »Minist.) 7.S Cassiodorus ("saec. viii-ix "') None.

Stuttgart H. B. VII 39 Bede on Proverbs (writtens at Con-
stance, 811-839) None.
H. B. XIV 15 Vitae Sanctorum ('* saec. ix in. ") None.
Treves (Stadt.) 118, foil. 124-183 Disputatio Fidei (" saec. ix ")
None.
122 Ambrosius (written at Treves in 883) In, lunctus (so-
metimes), Ihs the symbol (sometimes). But not hiiins.
(Dom.) 134 Gospels (partly Insular half-uncial, partly un-
cial) In. ludaeus. But not imius.
52 \V. M. Lindsay

Tro\es 657 Cassiodorus (written at Flavigny, "

saec viii ex.",
b}' mam- scribes) None, (see above, Autun 20^).
Vercelli 148 Gregorii Homiliae ("' saec. ix " by more than ;

one scribe) In. lam, Igitiir. But not ems, ilk.

W'olfenbuettel August. O 67.5 Annales Guelferb3-tani (written
in 813) In. But not iòi.
Helmsted. 496* Augustini Opuscula (Anglosaxon minus-
cule: ''saec. ix ") In (sometimes). But not ia?n^ ez2is,
ignis.
Helmstedt. 513 Lex Alamannorum (" saec viii ") In, lam,
elus, Ille.

Weissenburg 64 Isidori Et3mologiae (N. Italian cursive

of Bobbio: ''
more than one scribe)
saec. viii in. "; b\-
In, But not ille.
lam, hulus. Igitur.
Weissenburg. 81 Mart}Tologium (written in 772), In, leluno.
Weissenburg. 86 Pompeius Grammaticus, etc. (" saec. viii '';
by more than one scribe) In, lam, cuius, delnde. But
nqt ipse.
Weissenburg. 97 Lex Salica; Breviarium Alaricianum
(" saec. viii ''
; by more than one scribe) In. But not ille.
Weissenburg. 99 Augustini Homiliae (Merovingian minus-
cule, of saec. vii-viii ". by more than one scribe) In,
''

lam, delnde, Ille (often).

Wuerzburg, theol. F" 12 S. Pauli Epistolae (Irish minuscule of
'•
saec. viii in. ") In, lam. Ita.

theol. F 27 Origenis Homiliae (partly Insular half-uncial,

partlv minuscule; by many scribes) None.
theol. F 67 Gospels (uncial or large minuscule; ''saec.
viii ") None.
theol. F 69 Epistolae S. Pauli (Anglosaxon large minus-
cule; "saec. viii in."; by more than one scribe' In.
theol. O 1 Caesarius part in minuscule with no Insular
trait, part in Anglosaxon minuscule) None (except
sometimes / longior in in).

Zürich, C 12 Psalter (from St Gall; ''saec. ix in.") None.

140 Hieron3'mi Epistolae, etc. (from Rheinau; " saec. viii ")
In, alt, But not ipse.
elus.
In this list I have omitted MSS. in the Corbie ab-type
and (usuali}-) those in Visigothic and Beneventan minuscule.
.On their practice see Loew 1. c).
 Collectanea varia 53

VI. — Transmission of Texts.

Those who occup\- themselves with what

classical scholars
is called feet '
on the hob
emendation have a poor opinion
'

of the extant MSS. They sit by the tire with Virgil in one
hand and a pencil in the other and jot down in the margin
an\- alteration of a word or line which caprice suggests. When
this marginal litter has accumulated they send it, under the
misleading title Emendations
'
to an indulgent magazine-edi-
'.

tor. If any one thinks it worth while to censure them, they

justifv their action by some argument like this The trans- :
'

mission of Latin texts was whollx' capricious and wholly igno-

rant one can have no confidence that the traditional form of
:

a passage was what the author wrote the form in which we ;

ourselves re-write the passage is just as likeh' to have been

the author's form '.

No the transmission of texts was not so capricious as

:

Mr. Feet-on-the-hob's emendations nor so ignorant as his

' '

iustihcation of them. It follow^ed (or tried to follow) the great

Law of Eugenics, to breed from the best types, to transcribe

the standard exemplar- Sometimes pedigree-record is pre-
served CODEX ARCHETYPUS AD CUIUS EXEMPLARIA SUNT RELIQUI
:

CORRIGENDI (the Bamberg Cassiodore) peremexdavi ut potui ;

SECUNDUM codicem DE BiBLioTHECA EUGIPPI (the Codcx Epter-

nacensis); iste codex est scriptus de illo authextico quem
DOMXUS ADRIAXUS APOSTOLICUS DEDIT GLORIOSISSIMO CAROLO REGI
FRAXCORUM et LAXGOBARDORUM AC PATRICIO ROMAXORUM QUAXTDO
FUIT R0MAE (the W^ürzburg Canons); ex autextico libro bi-
bliotecae cubiculi scriptum (in an Oxford and a Frankfort
MS. of the Gregorian Sacramentary). Though none of our
MSS. of Paulus Diaconus' De Significatione Verborum have
(or retain) any such entry on a fl\-leaf, those which shew (as
most of the earlier MSS. do) the same form (square quarto,
with two columns to the page etc.) may, I think, be called
direct descendants of the author's copy (whose form they,
imitate).
A
pedigree MS. of (say) the eleventh century can hold
its own
with a papyrus (say) a thousand }ears older, because
the papyrus page, picked up from the refuse-heap in an
Egyptian village, is a page of a carelessl}- written, cheap copy,
54 W. M. Lindsay

whereas the vellum text can boast a descent, stage by stage,

from some standard edition of still earlier date than the pa-
p}Tus. The ancient uncial MSS. of the classical Latin writers
were brought Charlemagne's order, from the shelves
out, b}*
of some great library (Fleur}*, L3-ons, etc.) where the)^ had
lain safely during the Dark Ages, and were multiplied in beau-
tiful minuscule transcripts and the Carolingian transcripts were
;

again rescued from neglect by the Italian Humanists of the four-

teenth and fifteenth centuries. These uncial texts themselves
would be standard texts, directh' transcribed (or carefuU}' cor-
rected) from some standard ancient edition. Once we can
reproduce the uncial text which was discovered in Charle-
magnes time (and subsequently lost) we find little to correct.
Of course, even printed texts cannot wholl}' avoid error, and
written texts lacked the regularity of printing. Emendation
cannot be dispensed with. But the difhculty of emending a
Carolingian MS. is slight;it is the Renaissance texts, altered

by the caprice of emendators. which, in the absence of earlier

MS. evidence, offer the hardest problem. It is the fashion
to abuse medieval scribes but what of the iniquity of Re-
;

naissance editors.' Ot later editors.- Take Scaliger, the prince

of them all. Virgil in his studentda}'s had often talked with
his fellow-student Tucca (afterwards Mrgil's literary exe-
cutor) of a face caught his fanc}'. The lady, after a
that
long absence, returns to her house and Virgil writes to Tucca
(Catalepta 1).

De qua saepe tibi venit, sed, Tucca videre

Non licet : occulitur limine clausa viri.

De qua saepe tibi non venit adhuc mihi, namque

Si occulitur, longe est tangere quod nequeas, etc.

Surely the phrase de qua saepe tibi (seil locutus sum) might
be left in peace. But nothing is safe from the emendator.
Scaliger (and therefore everv other editor) emends to Delia ' '

saepe tibi (And what does tibi mean.^).

When Burns" Poems are published nowada3-s the editor
does not take at random some badl\' printed cop}', a pirated
or a cheap edition perhaps of Burns' own time- He follows
the standard edition. So in ancient and medieval times. Ci-
cero's Speeches ma}' have been hawked in the streets of Rome,
 Collectanea varia 55

in Cicero's own age, in a very inaccurate form of text. But

that was not the text which would be transmitted to subse-
quent ages. It would be some standard edition (say) by Tiro.
That edition would be available, in a public librar}- at Rome,
for future transcripts. From the public libraries of Pagan
Rome these standard texts would pass into the great Church
Libraries- Thus the good stock was transmitted the bad :

went to limbo. If Beer's theory be right (and it has been

praised b\' bepraised men) Cassiodore's wonderful librar\' was
acquired by Columban for Bobbio and has provided these
palimpsests which Mai recovered for us. Their fly-leaves, which
might contain their pedigree-record, have been lost.
And a new task devolves on the twentieth-centur\- editor,
to investigate the ancient editions of Latin authors. A nine-
teenth-century editor was satisfied with the classification cf MSS.
in families : we must henceforth see whether these families re-
flect ancient editions. Sometimes clear record of this remains ;

e. g. the Italian family of MSS. of Martial reflects an edition

made by Torquatus Gennadius in the year 401 (See m}- '
An-
cient Editions of Martial "
and '
The Year's Work in Classical
Studies "
1906. section XVII). Good fortune has kept apart
this thread and, I think, the other two threads of the Martial-
There has not been an}-
tradition. intertwining. And so the
modern emendators (and there are plent}- of them) who ven-
ture to alter a reading found in ail three families of MSS. are

defying the consensus of no less than three ancient editions

of Martial. The}- cannot (or will not) see that the condition of
these poems differs from the condition of poems which sur-
vived the Dark Ages in a single MS. onlv.
The best treatment of the Transmission of Ancient Texts
to modern times will be found in Traube's Einleitung in die
lateinische Philologie des Mittelalters (chap. Ill üeberlieferungs-
geschichte).

( To be conUimed)
56 Paul Lehmann

EIN BASLER FRAGMENT

DES NORDFRAXZOESISCHEX AZ-TYPUS.
vox PAUL LEH>L'\NN', MUEN'CHEN.

Wenn seit 1907 Faläographen und Kunsthistoriker zu mir

kamen, um L. Traubes für einen einzelnen Gelehrten sehr
grosse und erlesene Sammlung von Handschriftenphotogra-
phien zu benutzen, dann verlangten sie fast immer mit be-
sonderer Lebhaftigkeit die Mappe « TN'pen ». Das ist bezeichn-
end und erfreulich. Denn Proben vor- und frühka-
in ihr sind
rolingischen Minuskel man noch vor wenigen
vereinigt, die
Jahrzehnten als « merovingisch » kurz und mit einer Art von
Verachtung abtun zu dürfen meinte. Erst L. Delisle und dann
namentlich L. Traube, der nach seinen Aufzeichnungen schon
1900, wahrsjheinlich sogar früher < T\pen » unterschieden und
benannt hat, schliesslich des grossen Münchner Forschers
Spuren folgend andere haben begonnen in der « merovingi-
schen > Schrift eine wegen ihrer Bedeutung für die karolin-
gische Paläographie und Kunstgeschichte dringend erforschens-
werte Schriftfülle näher ins Auge zu fassen und zu gruppieren.
Wir stehen bei der Bewältigung der vorkarolingischen Minuskel
immer noch ani Anfang. Aber man sieht doch bereits dass
eine sehr kräftige Entwicklungslinie von der Luxeuilschule
über Corbie führt. Als eigentliche Corbieschrift des 8-9 Jahr-
hunderts ist uns jener Minuskeltype bekannt, der charakteri-
siert wird durch die Formen von a und ò, der ab-Typ von
Corbie (vgl. Lindsav, Revue des bibliothèques, xxii [1912]
pp. 405429).
Mit Corbie hängt fernerhin irgendwie der az-T3-p zus-
ammen. Ueber ihn hat 1914 in der Revue des bibliothèques (')
Lindsa}- gehandelt. Den Buchschmuck dieser Handschriften-
gruppe wie des ab-Typs hat, ohne Lindsays noch vor dem
Kriege erschienenen Aufsatz zu kennen, 1916 E. H. Zimmer-

ei) XXIV 15 27 < The Laon az Type >, mit 2 Tafeln nach Laon 137
und 423.
 Ein Basler Fragment des Nordfranzoesischen az-typus 57

mann (') vorgeführt. Ihre Untersuchungen zu stützen und zu

ergänzen dient meine Miscelle.

Beide Korscher kannten —
wie übrigens auch schon Traube
— 5 Handschriften des az-Typus in Cambridge, Laon, London
und Paris. Nun kommt als no. 6 ein Stück der Universitäts-
bibliothek Basel hinzu, das ich im April 1922 an Ort und
Stelle sah, später dank dem Entegegenkommen von Basel
und München im Sommer dieses Jahres in München selbst
prüfen durfte- Es galt bisher für angelsächsisch, da Konrad
Escher es 1917 öffentlich so genannt hat (« Die Miniaturen in
den Basler Biblioteken, Museen und Archiven », Basel, 1917,
S. 29). Aber diese Bezeichnung ist durchaus nicht zutreffend,
wie die ganze Beschreibung des Schweizer Kunsthistorikers
an Irrtümern und Flüchtigkeiten reich ist. Könnte man viell-
eicht, und nur sehr vielleicht, in den Initialen einen schwachen
ist völlig französisch. Für
insularen Einfluss spüren, die Schrift
den Kenner wird ein Blick auf die Tafel (') genügen, um jeden
Gedanken an Ursprung des Codex von England-Irland oder
einer insularen Kolonie des Festlands zu verscheuchen.
Zur Kennzeichnung der Schrift hat Lindsay wohl alles
Erforderliche gesagt. Es handelt sich um einen kalligraphi-
schen Versuch, bei dem noch viele Ueberreste der Kursive
in Buchstaben und in Ligaturen geblieben sind.
einzelnen
Auffällig ist erstens das aus zwei nach rechts oftenen Winkeln
gebildete a und das hochragende z (s. Palae. Lat. Taf. I, i

no. 8.1). Da z nur selten in lateinischen Werken vorkommt,

erscheint es auch auf den Basler Blättern nich oft. wenn ich
recht zugesehen nur dreimal (f ll^'^, 16^^, 18^-^), und zwar in
zwei Formen. Die früher veröffentlichten Proben des Typs
gaben kein Beispiel der z-Form. Man vermisst auf ihnen ferner
eine bizarre V^ariante des a, die den von Lindsay besprochenen
Handschriften freilich nicht fehlt, auf den Basler Bruchstücken
jedoch ziemlich häufig, wiewohl selterier als die t3'pische Form,
allein und in Ligaturen auftritt (s. Taf. col. ii 1. i
'
tempenzn-
tiae '), eine Variante, die Lindsay (p. 21) nicht übel mit einer
dicken von einem schmalen Ast herabhängenden Birne verg-

(1) Vorkarolingische Miniaturen. Berlin, 1916, S. 85 ff., 222, ff., Tafel

144-149.
(2) Plate I bietet fol SR in geringfügiger Verkleinerung : 22 X 28 statt
25,4 X 31 cm.
58 Paul Lehmann

leicht. Wir verschweigen nicht, dass dieses oder ein ähnliches

a (das « Merovingische unciale ^», Palae. Lat. i, Taf. i 23)
auch sonst in vorkarolingischer französischer Minuskel bege-
gnet ebenso wie die übrigen etwas ungewöhnlichen Buch-
staben: das steile d mit dem unter die Grundlinie gehenden
und dort ein wenig eingeknickten Schaft unbeholfene offene e
;
\

die zahlreichen Ligaturen im Anschluss an c, e, g\ r, / u. a.

Typisch, obwohl nicht auf die az-Gruppe beschränkt, ist das
hakenförmige, übergeschriebene ?/, das nach q steht und sich
mit den folgenden Buchstaben verbindet, die Unterscheidung
zweier Ligaturen ti vor Vokalen oder vor Konsonanten bez. am
Schkiss. Auch in den Abkürzungen trifft man nichts von
Lindsa3's Liste Abweichendes. Die einzige, nach Lindsay dem
T3'p eigenartige Abkürzung in (/ longa mit Querstrich) fand
' '

ich zweimal (f. 13^^-^ u.^^), von '

vel dadurch unterschieden,
'

dass der Querstrich wagerecht, bei vel schräg gestellt ist.

' '

Nicht selten ist zur Bezeichnung von Textlücken und Ergän-

zungen h (mit Querstrich) am Rande oder zwischen den Zeilen
bei der Lücke sowie vor und nach der Ergänzung gebraucht.
Alles in allem passen die Basler Fragmente trefflich in
Lindsa3s Gruppe, und zwar stehen sie am allernächsten Lind-
says D. Das gilt auch für den Buchschmuck. Man trifft kleine,
meistens umtüpfelte Initialen, die gern an den Enden Tier-
köpfe zeigen und mit den Proben aus D bei Bastard (pl. 29
rechts) fast ganz übereinstimmen, ferner grosse Initialen, in
verschiedenen Farben aus hundeähnlichen Tieren, Hirschen und
Schlangen gebildet. Besonders zu verweisen ist auf den das /
vorstellenden sprinijenden Hirsch (f. V"'^); auf das Q (f. 2^'^'^),
bei dem ein nach links schreitender Hund als Schweif des Q
die O-Rundung auf dem Kopfe trägt (wie bei Bastard pl. 29!);
auf den stehenden nach rechts gerichteten F-Hund (f. 7^^), der
mit dem E bei Bastard grosse Aehnlichkeit hat; auf das M
(t. 14^ '^), das aus zwei aufrecht stehenden, einander zugewandten
und sich die Vorderpfoten gebenden schlanken Hunden ge-
formt ist.
Die Basler Handschrift ist, nach dem Graphischen und
dem Bildlichen zu urteilen, wenn nicht von denselben Schrei-
bern und Malern so zum mindesten am selben Ort ziemlich
zur gleichen Zeit wie D geschaffen, d. h. wohl unì 770 in
Nordfrankreich.
 —

Ein Basler Fragment des Nordfranzoesischen az-typus 59

Das Band III der Fragmentensammlung in der

zur Zeit in
Univ.-Bibl. Basel aufbewahrte Stück besteht aus 23 Pergam.-
Bll. verschiedener Grösse (f. 1,2 23,7 X
30,9 cm.; 3: 24,4 X31.6
:

cm.; 4,5: 23,4X30,9 cm. 6 23,9X317 cm.; 7: 24,9X31,2

; :

cm.; 8: 25,4X31; 9— 23: 26 X

38,8 oder 38,7 oder 38 cm.). Die
vollständig unversehrten Bll. werden ungefähr 26 39 cm. X
gross gewesen sein. Die letzte Beschädigung erfolgte wohl,
als die Blätter, die heute erhalten sind um 1600 beim Einband
mehrerer Bände mit Disputationen u. a. Werken des Jenenser
Juristen Leopold Hackelmann (7 1619) gebraucht wurden. Da-
mals schrieb man rücksichtslos auf die Seiten die Jahreszahlen
1586 15%. Im Jahre 1880 gelangten die Blätter als Geschenk
des Basler Staatsarchivs in die Bibliothek. Die Schrift ist auf
2 Kolumnen von zumeist 37 Zeilen —
nicht, wie Escher
behauptet, auf 1 Kol. mit 47 Zeilen —
verteilt (Schriftspiegel
der einzelnen Kolumne etwa 9,5 X
32 cm.). Textlich bieten
die Blätter x-\ugustinus de civitate Dei lib- II c- 28— III 1 III ;

e 13-14, 19-31 \' e. 2-5, 19-22; VI e. 10— VII e. 1 IX e 13-X

; ;

c- 7; X e. 11-32; XI e. 21-23. Häufig sind durch Beschneiden

vonseiten des Buchbinders mehrere Textzeilen verloren ge-

gangen. Die von mir angefertigte genaue Inhaltsangabe gehört
wohl nicht in diese paläographische Zeitschrift. Ich möchte
nur bemerken, dass f. 23 der jetzigen Folüerung mit Hb. VI
c. 10 VII 1 zwischen f. 8 und 9 liegen müsste, und dass f. 21
und 22 offenbar ursprünglich (mit Hb. X
c. 32 u. XI c. 21-23)

das erste und das letzte Blatt der f.

22'''
unten bezeichneten
Lage XIIII gewesen sind.
Durch genaue Kollation mit der kritischen Ausgabe E.
Hoffmanns habe ich festgestellt, dass die Handschrift, die einst-
mals ein Prachtexemplar gewesen sein muss, höchstwahrschein-
lich die unmittelbare Kopie des für die Textkritik von Au-
gustinus de civ. Dei sehr bedeutsamen Parisinus lat. 12214,
einem Halbuncialcodex des 7 Jahrhunderts ist. Das ist keine
belanglose Ermittelung. Denn einmal ist es immer lehrreich,
\'orlage und Abschrift zu beobachten, zweitens erhält corbie
eine verstärkte Wichtigkeit für die geschichtliche Einordnung
des az-Typs. Stammt doch Vorlage unserer Basler Bruchs-
die
stücke. Paris lat. 12214, aus Corbie ('), ebenfalls dorther Paris
(') Von Cambridge C.C.C. 333 behauptet M. R. James, A descr. cata-
logue of the MSS. in the library of Corpus Christi College, Cambridge, II
60 Paul Lehmann

lat. 12168, welcher Codex uns aber mit der Sigi D nächster
als
Verwandter der Basler Blätter begeg^net ist. Man beachte auch
dass die beiden sich so ähnhchen Handschriften des az-T3-ps.
Basel und Paris lat. 12168 Augustinwerke enthalten (De civi-
tate Dei u. Quaestiones in heptateuchum). Ob der Typus in
Corbie zu Hause war oder etwa in Laon (vgl. Laon MS. 137
u. 423), wo vielleicht eine bestimmte Schreibergruppe Codices

Corbeienses kopierte, lässt sich noch nicht ausmachen.

(1911) p. 163 keineswegs, wie Zimmermann, Vorkarol. Miniaturen S. 226

glauben macht, bestimmt Herkunft aus der Corbier Bibliothek, sondern er sagt
sehr richtig und vorsichtig: « Very problably this book came from the same
place as no. 193 (? Corbie). A note in the catalogue gives Dr. Traube's opi-
nion that it is French, probably Northern >.
 Berne 207 61

BERNE 207.

BY \V. M. LINDSAY.

An article by Prof. E. K. Rand (in the Philological Quar-

terly I p. 258), '
A Vade Mecum of Liberal Culture in a MS.
of Fleury ', gives one something of the
which Traube's thrill

fascinating articles used to give, and prompts me to publish

some notes on this MS. which I wrote recenti}- at Berne.
(Plates II and III shew its strange script). Rand associates it
with the Leyden MS., Voss. Q 86. and a MS. mentioned in
the twelfth centur}- catalogue of Cluny. The three, he thinks
lead us to a chrestomathy and compendium of liberal cul-
'

ture used by students at Fleury, and constructed on lines sug-

'

gested by extracts from Isidore's Etymologiae.

The contents of the Clun}' MS. (which Peiper identified
with the Le3'den MS.) were luvencus, Sedulius, Arator, Prosper,
:

quoddam metrum Tertuliani, Cato, Avianus, quedam diverse

collectiones versuum diversorum actorum {leg. auct.), libri Ar-
chini {leg. Alcimi) episcopi, ars Isïdori de grammatica et de
disciplinis aliarum artium. Rand thus describes the Leyden
MS.; (159 pages lost at the beginning, enough to hold Ju-
vencus and Sedulius' Paschale Carmen), Arator, Prosper, the
little epics on Jonah and Sodom, the two h3'mns of Sedulius,

the Disticha Catonis, Avianus, selections from the /anthology,

selections from Martial, Avitus, part of Isidore's chapter on
grammar from his Et3'mologies the end of the MS. is lost. ;

From Hagen's account (in Anecdota Helvetica) of the contents

of Berne 207, Rand finds that its Isidore extracts correspond
with the Isidore extracts (so far as these go) in the Le}den MS.
What is the histor}' of the Corpus Grammaticorum in the
Berne MS. ? My notes cannot solve the problem, but I pu-
blish them in the hope that the}' ma}- lighten the labour of
any future researcher. Curiousl}- enough, the}- ended with a
use of Rand's actual phrase Vade Mecum Here is the last
' '.

sentence The Donatus pages are much worn and were glossed
:

at Fleur}-, where the}- seem to have been a class-book ; the

Petrus pages too (with the schoolgrammar of Peter of Pisa)
62 W. M. Lindsay

have many later marginalia ; that the codex was a '

vade
mecum '
for the monastery-school where it was written is

snewn bx' the miscellaneous Isidore extracts.

What took me Berne MS. (formerly in
recentlv to the
Fleur\' Librar}-) was
Toletanus portion.
its The
Julianus
forthcoming- edition of the Liber Glossarum required the pu-
blication of Julian's account of rhetorical faults and figures.
That publication {= St Andrews Universit}- Publications,
n°. XV) shews the MS, used by the compiler of the Liber
Glossarum and the three (known) extant MSS. (at Rome, at
Erfurt, at Berne) to have a common archet}'pe (with vcrmibiis
for Vergilius at VI 21). And that this archet3-pe was a MS.
in was proved by the numerous and varied
Visigothic script
'
Spanish symptoms in all four. Goetz had contended that
'

the Liber Glo.ssarum was compiled in Spain and had used

its Julian items in support of this contention. But Goetz had
not noticed that a great source of the Liber Glossarum, the
Abstrusa Glossary (in its full, original form), was a glossary
based on Adamnans Virgil Scholia nor yet that the earliest ;

MSS. of the Liber Glossarum are all in the Corbie ab-script

(which suggests that it was compiled in Corbie). The Julian
portion of the Berne Corpus Grammaticorum we may refer
to a MS. in Visigothic script (and this script was in vogue
not merely in Spain but in parts of France), the same MS.
from which a transcript was made at Lorsch (now in the Va-
tican Library) and another (perhaps) at Erfurt or in the Er-
furt segion.
Since the contents of Berne 207 have been detailed by
Hagen (first in Anecdota Helvetica then in the Berne Cata-
logue), I limit m}self to the clues to its histor}'. (Notice Mo-
relli's attribution to '
Maximus Victorinus '
= Keil VI 229, not
to Servius, of 1L-12\ And foil \Ti\ middle to 129^ are
foil.

a later addition in Fleury minuscule). The title of the volume,

in coloured capitals (fol. L), I read as rude dactylic hexa-
meters :

Auctorum curavi adligere nomina horuni

Huic libello quem nimio sudore conpegi :

Donati, Asperi, Prisciani, Sergii, Probi,

Isidori sophi, Bedae, Petrique, Martis Eruli.
 Berne 207 63

I am inclined to think that the closing blunder, martis

ERULi for MARii sERUTi, should be charged to the compiler who
misread the majuscule heading in a MS. ot this grammarian
(cf. fol. 77' Incipit Martis Eruli grammatica)- But I do not
insist; for a versifier who ended a hexameter with S(rgii
Probi was perhaps capable ot finding a dactyl and spondee
in Marii Servii. The order of the grammarians' names in
this title does not correspond with their order in the Corpus
itself, a transposition due to metrical necessity. The Bede
postion is missing.
This has been misplaced by a modern binder. Be-
title

fore this leaf we must place the two leaves which now follow
it. (Or did they stand elsewhere, in a missing part.'). They
contain characteristic Insular lore : a variety of alphabets (along
with Bede's account of finger-speech).These alphabets (Greek,
Hebrew, Runic, etc,) might repa}' a more careful study than
I had time for. N'. 4 is Aethicus Ister's alphabet: alamoi,
becab, etc., the letterforms not very difterent from the mere
fantastic script affected by Insular calligraphists in colophons, etc.
a script used in this Corpusfol. 81') to embellish title-
(e. g.

headings. (Indeed the ver}- pains taken throughout with the

embellishment of the script of mere grammatical extracts, the
red colouring, the dog-head initials, etc., suggest Celtic work;
since the veneration of Latin Grammars was a Celtic failing).
N". 5 is the Ogam alphabet: but since the essential feature
of Ogam (the distinction of letters (1) above the line (2) below
the line (3) transected by the
has been lost, the scribe
line)
clearl}- copied an exemplar wnich he did not understand (where
as the corrector of the St Gall Priscian writes the equivalent
of contuli
'
on p. 196 in good Ogam). This leads us to Ire-
'

land. And there are Irish glosses on foil. 146'' and 148' (if

elsewhere, I overlooked them) in the scribe's handwriting

(I think), and (perhaps) therefore from the exemplar (but when
when added to the exemplar, who can say.-) e. g. amre won- '

derful '
(over papae). These pages belong to the portion with
Servius' (Sergius') adaptation of Donatus.
The same portion (e. g. 143^", 144', 145'j shews omission-
symbols which (see Revue Bibliothèques XXIV p. 18) point
to the Continent rather than Ireland and have their chief
home in Italy; h ci (in text) and /// (in margin). In the head-
 64 W. M. Lindsay

line of the next portion (fol. 148''), the Petrus portion, DR

'
dicitur '
DN
Anecd. Helv. p. xxx).
has been miscopied (see
The MS. has, I think, been written by two scribes (exem-
plified in the two plates appended to this article), both of
whom use the same t3'pe of script. And what shall we call
it? have wavered between Continental-Irish, Welsh, Breton,
I

but finally acquiesced (with no complete confidence) in Con-

tinental-Irish. The abbreviations, of which a list will he found
on p. 64 of my Early Irish Minuscule Script, suit the Irish
t\'pe (e. g. the absence of a symbol of saeculwn), but cannot
be said to demand it.
When I published these abbreviations (in 1910), I had the
idea that this MS. preserved some Irish abbreviations which
earlier Irish MSS. had lost especially the monogram-signs of
nihil and nisi. But subsequently I found at Rome a Lorsch
grammatical MS. of the beginning of the ninth century, a
transcript of an uncial exemplar, with the unfamiliar syllabic
suspensions faithfull}' reproduced b\' the transcriber (see my
Notae Latinae p. xiv but delete the remark on potius). And
;

the use of these Ancient Notae (e. g. the ;////// and nisi mo-
nograms) in Grammars (as well as in Law-books) is confirmed
by other relics of them in grammatical transcripts (see my
article Ancient Notae and Latin Texts in the Classical Quar-
' '

terl}- XI [1917] 38). Therefore these unusual symbols in Berne

207 may be s}'mbols of the exemplar, rather than current
symbols of the transcriber.
If so, this Corpus Grammaticorum was made with the help

of ancient MSS. (or one MS. collection) of the older Gram-

marians here exhibited, e. at some great library (Bobbio ì
i.

Fleur}' }). The Isidore extracts share these Ancient Notae

(e. g. the 7iisi monogram frequently; the nihil monogram on

fol. 187'). So does the Julian portion (e. g. the nihil and nisi
monograms frequently), but not Peter of Pisa (the hr horum ' ',

7i)n nomen in his pages mav be mere technical curtailments).

' '

As to the date of the Berne MS. or an exemplar, I doubt the

wisdom of an appeal to the red smudge in the margin at the
year 792 of the Paschal Cycle. It looks like an impress from
the paint on the next page. Though the scribe abbreviates
'
nostri etc. b}' the three letters nri, his exemplar had the
',
 Berne 207 65

older two-letter s\'mbol as we

see from relapses like (157'' pa-
radigm of '
noster '
in the Petrus portion) nr ni no nrin.
should be mentioned that Greek letters are
Lastl}', it

transcribed with ver\- fair accurac}' (another Irish feature ?).

And that the suprascript letters here and there seem to me
to be aids to the student e. g. 134'" [ieimio, oi'o, vigilo) the n
:

over the first and third verb

I interpret neutrum ', the a over '

the second I interpret activum 138'' {ei, que, at, atqne. ac,
' '
;

est) the p over the last four I interpret as prima ', i. e. first '

word in sentence, and think the s over et should stand over

que, secunda ', i. e. second word in sentence.
'

Recently (Palae. Lat. I 65) I published the warning: Let

the palaeographer amass details and avoid rash inference.
But, after all. it is well that one who has examined a MS.
should state (tentatively and with acknowledgement of the
fallibility of hast}' impressions) what impression the MS. has

made on his mind. For there is a vast difference between

seeing Berne 207 for one self and merely reading Hagen's
account of it. (The Berne Catalogue omits to state that on
fol. 138\ there is a later entry in the bottom margin: hie
liber sancti Benedicti Floriacensis. It is this entr}', and this
alone, which tells us that the MS. once belonged to Fleurv
Library). The ornamental nature of this Corpus Grammati-
corum gave me the impression that was the compiler's own it

copy, written under his direction b\' two skilled penmen. The
compiler, whose Martis Eruli bids us write him down with
' '

Dogberr}', makes display of his Greek learning with his Isi- '

dori sophi quite in the style of the ninth centur}- Greek

',

teacher (such as Martin the Irishman who taught Greek at

Laon). If stress were to be laid on his claim multo sudore '

compegi we might believe him to have hunted up gramma-

',

tical MSS. from all quarters as diligenti}- as Lupus of Ferneres,

but it is more likel}' that he got all of them in Fleur\- Li-
brary; unless indeed the phrase is a mere tag to complete
the line. The collection was a text-book for his lectures, and
does him credit. It (or part of it) was subsequenth" used as
a class-book at Fleurv. But whether it was made originali}-
at and for Fleury is a question which can hardh" be settled
until some other specimen of this peculiar script turns up.
\\'ill those who visit libraries please look out for it.
66 S. Tafel

THE LYONS SCRIPTORIUM.

BY THE LATE S. TAFEL.

one of Traiibe's pupils, who fell in the war,

[Dr. Tafel,
was engaged on a histor}- of the Lyons Scriptorium. His
large collection of papers and photographs has been placed
in my hands. To print even a part of them as they stand
would be a mistake. The best wa}- to rescue from the wreck
what will help the progress of Palaeograph}- is to write in
m}^ own words whait I conceive to be Dr. Taf el's arguments
and conclusions. After a visit to Lyons I hope to publish
more meanwhile I offer this small selection to my rea-
:

ders. ED.].

I. — The Cathedral Library.

The larger (earh) medieval libraries preserved for us MSS.

of the ancient world. Pagan and Christian. And they pro-
vided material for students within their own walls or in the
surrounding region. So we must ask two questions: What
ancient MSS. were preserved at Lyons What learned re-
.-^

search was carried on in medieval times there and in that

region ?
Look at the map and \'ou can guess what Lyons' work
in the transmission of culture would be. At Lyons the Rhone
and the Saône join, and the two make a long channel of
communication from .South to North, from the Mediterranean
to the heart of the Continent. Lyons, at the middle of this
water-way, was by its very destined to be the
position
transmitter of the learning of Spain and Italy to France and
Germany.
The Roman colony Lugdunum, the centre of that culture
of Southern Gaul so strongly influenced b}- the culture of
Greece and the East, was in the early Empire the chief town
of Gallia. P^rom the fourth century it had to yield to Treves
and Aries, but yet retained much of its old preeminence.
The first important Christian comniunit}^ of Gaul was sheltered
 The Lyons Scriptorium 67

within walls; and the lame of its mart\TS spread tar and
its

Although its Archbishop had authority over a limited

^vide.
region only, it seems to have increased its power in the
seventh century, for not a few Merovingian Vitae Sanctorum
call it caput Galliae
'
In the fifth and sixth centuries, from
".

about 470 to 534, it was the capital of Burgundy and the re-
sidence of the Burgundian kings even later, in the vicissitudes ;

of stirring times, it may have regained this position more

I'^.an once. But the eighth century was a time of decadence
for L}'ons as for the rest of Gaul, and in 732 the town was
I'lundered by the Saracens. Finally, however, at the close
ot that centur}-. it shared in the prosperit}' of the Carolingian
epoch and once more became a leading cit^.
Thus, in all its ups and downs, there was no complete
upheaval, nothing to put an abrupt stop to the continuity
of its culture. Under the Roman Empire it was a Roman-
Hellenic centre ; and Greek, no doubt, kept a footing there
tor man\' centuries. Its famous Rhetoric-schools still flourished
in the fifth century, as we
tVom the letters of Sidonius
see
Apollinaris and in the beginning of the sixth centurv a rhetor
:

of L\"ons could correct a grammatical error of the bishop of

Vienne (cf. Alcimus Avitus Epistulae in Mon. Germ. VI- xx).
' "

How these Rhetoric-schools passed into the Church-schools

we cannot say. but the secular element in its teaching is em-
phasized in the famous panegyric of Lyons by Heiric of Auxerre
(Mon. Germ. Epist. \ I 124); ibi quas dicunt discipHnarum li-
beralium peritia, quasque ordine currere hoc tempore fabula
tantum est, eo usque convaluit ut, quantum ad scholas, pu-
blicum appellaretur citramarini orbis gymnasium. Heiric is
speaking of the latter part of the fifth centur}-, the time of
Constantius of Lyons, his predecessor in the biograph\- of
St Germanus of Auxerre. His words shew at least the be-
lief, current in his time, that L\-ons was a great transmitter

of ancient culture. In the seventh century, under Bishop Dal-

finus (alias Annemundusi, Wilfrid, subsequently Archbishop
of York, studied for three \ears there et a doctoribus valde *

eruditis multa didicit (Was one of the doctores that

'. '
'

'
grammaticus anonymus mentioned in my next paragraph .-).
"

The first care of the reformer of the Lyons Church in Char-

lemagnes time. Archbishop Leidrad, was the restoration of
68 S. Tafel

the Church-schools (see his letter to Charlemacrne in Mon.

Germ. IV 5 39).
I need not enumerate the earlier literarv celebrities. Pagan
and Christian, of Lyons and that whole region: the S\agrii,
Sidonius Apollinaris, Eucherius and the rest. They are treated
in p. de Colonia Histoire littéraire de la ville de L}on
•
(2 vols..
'

L3-ons, 1728) ; volumes of the Benedictines' His-

in the earlier '

toire littéraire de la PVance in M. de la Saussure

; Les six *

premiers siècles littéraires de la ville de L^'on Lyons and ',

Paris, 1876; partly too in the Mémoires de l'Académie de Lyon.

Classe des Lettres. 1858-61. I will onl}- mention that Lyons
seems to have been the centre of that great juristic labour
which produced the Lex Romana Burgundionum and it is ;

Lyons' librar}" which has preserved for us a great part of

Roman juristic literature. Also that there appears to have
been at L\"ons a succession of hagiographers from Constan-
tius and Eucherius on to the ninth century. But I would
call the readers attention to F'eiper's very convincing argu-
ment (Jahrb. klass. Phil, Suppl. XI 297) that a small gram-
matical treatise which conserves ancient tradition was written
at L}ons, the De dubiis nominibus
'
Keil. Gramm, lat. \' 571).
'

It deals with the gender of nouns, illustrated b)' quotations,

and enables us to recover much of that important grammar

of the second century. Caper's De dubiis generibus.
The ninth century is justly called by P. de Colonia the
'
âge d'or' of Lxons Hterature. We find a band of learned
writers working in harmon}- for the defence of true religion.
First in time Archbishop Leidrad, the reformer of the L}ons
Church, author of opuscula and of letters to Charlemagne.
798-814. First in importance his successor in the bishopric,
Agobard (814-840), one of the leading figures of the ninth
century, a voluminous writer of extraordinär}- learning. With
him worked the erudite deacon Florus, the famous antagonist
of Johannes Scotus and of Amalar. Florus was equally lo}al
to Agobard s successors, Archbishop A molo 840-852) and Re-
migius (852-875), aiding them in their literary war against the
Jews and Godescalc's Predestination-theories. The works of
these four authors will be found in these volumes of Migne's
Patrologia latina: Agobard, vol. 104, p. 29: Florus, vol. 119,
p. 11; Amolo, vol. 116. p. 77; Remigius, vol. 121, p. 985.
 The Lyons Scriptorium 69

Their letters, in Mon. Germ. Epist. the poems of Fl >rus in

;

Mon. Germ. Poet. II 506. In all probability they were the

teachers and inspirers of not a few of the Carolingian theo-
logian.s and literary writers. And a large colon\- of Spaniards,
driven out by the Saracen invasion, seems to have settled at
L^ons, e. g. Felix of Urgel and his pupil Claudius, subse-
quentlv bishop of Turin.
We
ma\' take it that the treasures of the Lx'ons librarv
provided material for much of the ninth centur\- literature, so
copious in citation of the early Christian authors. That a
man's learned writing depends upon his reading is a truth
that was truer in the ninth century than it is now. At that
revival of learning was the contents of a cathedral librar}'
it

which made the literarv work of the region possible and

guided it this way or that. That is how we should view the
L\ons MSS. And we should try to make these silent witnesses
of the Carolingian literary activity yield to us their secrets :

•
This was the volume from which Agobard drew ammunition
for the good tight against heresv these marginal entries came
:

from the pen of Florus in his hours of studv ".

Cathedral libraries habent sua fata

•
Three were fortu-
'.

nate: Lyons. \'erona, Würzburg. Their treasures were to a

wonderful e.xtent preserved at the original home. True we ;

dare not assume that each and every one of their ancient MSS.
was actually written in the home scriptorium. X'^erona, as
Beer has convincingly shewed (Die Handschriftenbestand von
Bobbio, in the Anzeiger der K. K. Akad. d. Wiss. in Wien,
1911, p, 89), took from Bobbio some old MSS. which Bobbio
in its turn had taken from Cassiodore's famous library at \'i-
varium. Still these immigrants, transcribed Verona
in the
scriptorium, studied b\' successive generations of Verona ec-
clesiastics, became to all intent \'eronese. Whether palaeo
graphic investigation will enable us to discriminate the autoch-
thones from the advenae in the twelve uncial and half-uncial
MSS. now in the Lyons library remains to be seen. At any
rate we have little difficult}' in connecting each of them with
the learned work of Lyons
in the ninth centur\-. And from
an intelligent survey of that learned work we can reasonably
infer the presence there of other ancient MSS., since lost or
straved.
70 S. Tafel

So the ninth century must be our point of departure. To

investigate the earlier and the later history of the Lyons li-

brary we must first tr\- to reconstruct in imagination the li-

brary's contents in Leidrad's time.

II. — The Library of He Barbe.

At the eve of the great Revolution of 1789 there were

in L3ons, beside the Cathedral-library, a fair number of other
church-libraries. Man\- of them, were possibl}- ver\' old,
such as the libraries of St Just, St Irénée (from which in the
sixteenth centur}' Th. Beza took that Graeco-latinus codex
of the Gospels called the codex Bezae '
and perhaps the',

codex Claromontanus of the Pauline Epistles), St Paul, St Martin

d'Ainay. IVIanN' of them ma^' at the time have owned old MSS.
To a previous period we ma}' assign the dispersion of the
library of a monastery in the vicinity of L3ons, on an island
of the Saône, Ile Barbe, an ancient foundation (Insula Barbara)
mentioned b}- Gregor\- of Tours.
In the late Middle Ages a legend arose regarding the He
Barbe librar}- and was, from the beginning of the nineteenth
centur\', applied also to the older MSS. of the Cathedral, a
quite false legend that Charlemagne s librar}- was lodged at
He Barbe under Leidrad's care. I find it first in Paradin's

Mémoire sur l'histoire de L}on, L}'ons, 1573, p. 99. It is re-

peated b}' man}' writers of the following centuries, with this
or that variation with the monastery-tradition of the
e. g,
presentation of a MS.
of Dionysius Areopagita written by
Charlemagne's own hand Celandine, the first modern histo-
.

rian of the Lyons xMSS. (Manuscrits de la Bibliothèque de

Lyon I [18121, p. 2i actually identifies the older of them with
the remains of this He Barbe librar}'. The older, the more
fragmentary a Lyons MS. is, the more unhesitatingly does he
ascribe to it this provenance. He adds that Archbishop An-
toine d'Albon, also Abbot of He Barbe, in the }ear 1562, after
the plundering b\- the Calvinists, had the monaster}- MSS.
conve}'ed to the Archives of his cathedral. This statement,
for which I have been quite unable to discover the slightest
authority, I believe to be a mere guess b}- Delandine.
 The Lyons Scriptorium 71

Now the whole legend is based on a passage of the letter

(Mon. Germ. Epist. IV 542) written by Leidrad, probably near

the end (814) of his archbishopric, to Charlemagne. He
makes a report to the emperor on his mission, the reform
of the L\-ons Church. He speaks of the monasteries in the
town and the neighbourhood, and how he had restored them
to order. The passage regarding He Barbe ran, in the im-
perfect version first published of the letter, so Et mmiaste- :

riuvi regale insidae Barbarae situm in medio Araris flumine,

quod antiquitus est dedicatum in honore sancti Andreae apo-
stoli et omnium apostolorum, nunc autem in honore sancti
Martini recens videtur esse fundatum mssii donini Kar olì ini-
pei'atoris, qui ibidem praefecit Benedictum abbatem. cum quo
simul direxit ibi silos codices) ita restauravi ut tecta de novo
fièrent et aliqua ex maceriis a fundamentis erigerentur, ubi
nunc monachi secundum regulärem disciplinam numero no-
naginta habitare videntur (Diimmler Mon. Germ.). The
bracketed sentence, from situ7n (or from nunc autem) to codices
betra}-s itself as a later interpolation b\" its introduction of
Charlemagne (to whom the letter is addressed) in the third
person. And we have actuallv a version, hitherto little
known, of a part of the letter, without the interpolation. It
is in the Rhone Archives at L}-ons (Fonds de St Pierre) and

is ascribed by M. C. Guigne to the thirteenth century. Though

unluckil}- it was not visible at my visit, it has been printed
by Guigne in the Cartulaire Lvonnais I (1885) p. 4, where the
passage reads Monasterium quoque (so regale too is pro-
:

bably interpolated' Insule Barbare situm. in honore sancti

Martini dedicatum, ita restauravi ... So the transcripts which
have preserved Leidrads letter for us appear to be the ver-
sions of difterent monasteries, the monasteries whose privi-
leges or claims to fame are mentioned in it, and which there-
fore were interested in its preservation. And this interest
interest explains these later interpolations. Will not some
one collect all these versions and edit the letter once for all .-).
The He Barbe tradition is represented in the official transcript
of the year 1447, a transcript preserved in the paper MS.
Lyons 1488, with the above mentioned interpolation. This
paper MS. was sought in vain hx G. Morin (Revue Béné-
dictine IX [1892[ 339 and was not used b\- Diimmler. It shews
72 S. Tafel

US that the forg-ery was as old as the fifteenth centun-, pro-

bably older.
Indeed the interpolated sentence itself is nothing more
than the usual stor\', so common in monastery-chronicles, of
a present of books b\' Charlemagne, and cannot be stretched
to a transfer of Charlemagne's library. Details are given of
his present to He Barbe in another report, a Fundatio eccle- '

siae et monasterii Insulae Barbarae which CI. Le Laboureur

',

saw in a seventeenth century MS. and refused to believe

(cf. Les Mazures de l'Abbaye royale de l'Ile Barbe, Paris,
1681, p. 5): Illud quidem Monasterium Carolus Imperator glo-
riosus summo dilexit amore, qui posuit illic abbatem nomine
Benedictum, secundum carnis cognationem sibi propinquum ;

et misit tabulas eburneas miro opere sculptas atque politas

et libros, scilicet quatuor Evangelistarum expositiones. Bene-
dict, of course, is Benedict of Aniane. At Leidrad's request
He Barbe received monks of Aniane, as is mentioned in Ardo's
Vita Benedicti Anianensis (Mon. Germ. SS. XV 209). These
monks would probably take Aniane MSS. (or transcripts of
them) to He Barbe; so we have no reason to doubt a collection
of MSS. at He Barbe. And the close relation of the Saône
monastery with the Cathedral makes probable a certain
amount of exchange of books. We know that in the first
half of the sixteenth centur}' three earl)' MSS. were found at
He Barbe. First, the famous MS. of Ausonius in V^isigothic
minuscule, which Sannazar found (with others, as his report
suggests) in his visit to France in 1501 or 1502 It is now at
Leyden, Voss. Ill, but I have shewn (') (Rhein. Mus. of 1914)
that the first part is at Paris, lat. 8093. Next, J. de Gagny
found there, before 1536, an old MS. now lost) of CI. Marius
Victor. His account (in his edition of 1536) shews that the
monaster3--library had already suffered: quum in Barbarensis
insulae coenobium amici aliquot venissemus incidit in manus
. . .

nostras ex eius loci vetusta sed ante annos aliquot depeculata

libraria poema doctum et elegans Claudii Marii Victorii
. . .

Massiliensis Oratoris titulo inscriptum. Finally, Abbot Antoine

(•) [It is unlucky that Dr. Tafel 's discovery was not utilized for the

recent Bordeaux facsimile of the Leyden Ausonius by Ville de Mirmont,

Paris 1919). If a facsimile of the Paris leaves had been added, we should

have the complete MS. ED.].

 The Lyons Scriptorium 73

d'Albon (mentioned above) found, at some time before the

religious wars, an old MS. ot" a Commentar}' on the Psalms,
which he ascribed to Rufinus in his edition of 1596 (reprinted
in Migne, P. L. 21, 633), but which has recently been claimed
for Alcuin h\ H. Brewer (Innsbrucker Zeitschr. f. kath. Theol.
37 [1913] 66S). This MS. too is lost.
These three MSS. may be referred to the ninth century.
The Le\den Ausonius was certainly written then. A Com-
mentary b\' Alcuin cannot be older. And Gagn3-'s account
of the CI. Victor MS., in spite of his assurance that it was
extraordinaril}' old, does not give us the impression of an
earlier date.
Delandine's ascription of all the older Lugdunenses to
He Barbe breaks down on another count. A number of ninth
century MSS. now at L}"ons were presented to the Cathedral
Library by the archbishops Leidrad, Agobard, Amolo and
Remigius. We find inthem a record of the presentation. In
one, L}'ons 471 (401) Beda in Esdram, presented by Agobard,
we Rostaing who re-arranged
find also a 1511 entry b\' a Pierre
the Cathedral Librar\'. And this tells us that the MS. be-
longed to the Cathedral both in the ninth centur\' and the
beginning of the sixteenth. On fol. 2'' is a fourteenth century
title: Liber Bede super Esdram prophetam. Since a number
of other Lyons MSS. have their titles written b}' this same
hand, we may infer an Inventory-taking in the fourteenth cen-
tur}- of the Cathedral Librar}-. These titles appear in some
of the oldest MSS., e. g. n.° 468, of saec. vii. This confutes
the idea that all the old MSS. belonged to He Barbe until the
sixteenth centur}'.
To my knowledge, there is only one MS. (Paris lat. 2859,
whose presence at He Barbe in the tenth century is
saec. ix)
proved by an ex libris entry. Even it, in my opinion, was
' "

given to He Barbe bv the Cathedral.

( To be co7iti?iued).
74 Wilhelm Weinberger

BIBLIOGRAPHIE DER LATEINISCHEN BUCHSCRIFT

(bis 1050) 19n-1922
VON WILHELM WEINBERGER.

L Die von Herrn Professor Lindsay angeregte und geförderte

Bibliographie, die den Fortschritt
und die weiteren
Aufgaben der Forschung betonen möchte, schliesst
sich an Text und Bibliographie von Thompson's In-
troduction auch in dem Sinne an, dass dort nicht
erwähnte Veröffentlichungen von 1910 hier aufge-
nommen wurden. In der Anordnung schienen einige
Aenderungen notwendig, um die wechselseitige Beein-
flussung der Buch- und der Bedarfsschrift (die also
nicht ganz beiseite gelassen werden konnte) hervor-
treten zu lassen. Es folgen also auf einen allgemeinen
Teil in dem Thompson's Introduction und -wegen
seines Textbandes- Zimmermanns Tafelwerk den
Uebergang von Büchern zu Sammlungen von Schrift-
proben bilden und anhangsweise einige allgemeine
Werke über Abkürzungen verzeichnet sind, die Ab-
schnitte: II. Kapitale. III. Kursive. IV. Unziale. V. Kon-
tinentale Halbunziale. VI. Insulare Schrift. VII. Ueber-
gangsschriften (a. nord-, b. süditalienisch, c. spa-
nisch, d. merowingisch). VIII. Karolingische Minuskel.
— Für Cassiodors Hss-Sammlung in Vivarium s. un-
ten 13. 21.
2.Reproduktionen ganzer Hss werden in 12, 16 [25J und 32
erwähnt.
Für Palimpsestphotographie nenne ich hier:
R. Kögel Die Photographie historischer Dokumente. 44
Beiheft zum ZB 1914 (vgl. Stud. Mitt, zur Gesch. d.
Benediktinerordens XXXIII 304. XXXIV 127. Beri.
S. -Ber. 1914, 974); Palimpsestphotographie. Encyclo-
pädie der Photographie. 95 Heft, Halle 1920.
[G. L. Perugi Saggio di Anastasiografia, Roma 1922;
Nuovo Metodo Scientifico per la Riproduzione dei
Palinsesti, Roma 1922. ED.]
 Bibliographie der Lateinischen Buchscrift 75

Für das Erlangen von Photographien aus den einzelnen Bibl. :

die Zusammenstellungen von Rabe Ph 1915, 30; und W

in Photographs of MSS. Reports from His Majesty's
Representatives abroad respecting Facilities for ob-
taining Ph. of. MSS. in Public Libraries in certain
Foreign Countries, London 1922,
An Kürzungen:
(Neues) Archiv (der Gesellschaft für ältere deutsche Geschichte).
Bibl. = Bibliot(h)eca, Bibliothek. Bibliothèque.
C. = Catalogo, Catalogue. Catalogus.
C(odices) G(raeci et) L(atini photographice depicti).
D(eutsche) L(iteratur-) Z(eitung).
Jahresberficht für Altertumswiss.).
J(ournal of) Th(eological) St(udies).
L(iterarisches Z(entral-) B(latt .

Ms, Mss = Manuscript(us), manuscrit, manoscritto, Ma-

nuskripte usw.
N(eue) Jahrb(ücher f. klass. Philo!.).
Pal(ae) =
Paläographie, Paléographie usw. ;
pal = pa-
läographisch usw.
(Berliner) Ph(ilologische) W(ochenschrift).
R(eproductions) B(ibl.) N(ationale).
R(eal-) E(nzyklopädie d. klass. Altertumwiss.).
\^erz(eichnis).
Z(entralblatt für) B^ibliothekswesen\
Arbeiten, von denen ich nicht Kenntniss nehmen konnte, sind
mit "^
bezeichnet.
3. Für die in 1 berührte Entwicklung der Schrift kommt eine
bestimmte Ter mifio logie in Betracht. Uncialis, das an
der oft behandelten Hieronymus-Stelle nur 'zollgross'
bedeuten kann (jahresber. CXCIII 81) hält Traube
(III 117 = Münch. S.-Ber. 1900. 533) für eine Prägung
der ältesten christlichen Kalligraphie. Die anderen
Bezeichnungen der Schriftarten gehen sicherlich teils
auf Mabillon zurück, der sie in sein Werk De re di-
plomatica (Paris 1681) von Praktikern übernahm (Trau-
be 79 =
472), 'teils auf die Herausgeber neuer und neue-
stenHandbücher oder Sammlungen von Schriftproben.
Für Unziale und Kursivunziale (die Majuskelkursive
kann in Kapital- und Unzialkursive geteilt werden)
76 Wilhelm Weinberger

S. 15, für Halbunziale

16. für Halbkursive 11 a. E.
15,
und Jahresber. 135 Besonders wichtig ist es,
S. 24.
dass Minuskel nur von ganz regelmässiger Schrift
mit durchgeführter Worttrennung gebraucht werde
(vgl. unten 6, 16). Statt Nationalschriften sage ich
Uebergangsschriften (s. 22 Fraglich ist es, ob die .

insulare Minuskel hinzugenommen und etwa gr-Schrift

genannt werden soll, die süditalienisch dann er- oder
Kerbschrift, die spanische gt-Schrift (s. N. Jahrb.
XLVII Die merowingische Schift könnte als
168).
a- oder als Schnörkelschrift bezeichnet werden, wenn
mann es nicht nach dem unten in 29 Gesagten vor-
zieht, nur die Schrift von Luxeuil als ar- die von ,

Corbie als b- und den az-Typ abzusondern.

,

I. Allgemeines.

4. Ludwig Traube Vorlesungen und Abhandlungen, II (Mün-

chen 1911) 22-31. III 1920 VI-X, 95-105, 112-119. 213 ff.

Bertold B retholz Lat. Pal. (Grundriss der Geschichtswiss.

hgg. V. A. Meister P) Leipzig, 1912, 112 S. bietet
eine gute, in Literaturangaben und Beispielen auch
auf Einzelheiten eingehende Einführung.
Wilhelm Weinberger Bericht über Faläographie und
Hsskunde 1907-1910. Jahresber. CLVIII 1912, 96. -
1911-1915. CLXXII 191,5, 1. - 1916-1921. CXCIII 1922,
79. — Schrift. RE klass. Altertumswiss. II A 1921,
711-717, 723-737.
R. Ca gnat Cours d'épigraphie lat.^ Paris 1914 (S. 10
Buchstabenformen der Inschriften).
5. Paul Lehmann Aufgaben und Anregungen der lat. Phi-
lologie des Mittelalters. Münch. S.-Ber. 1918 VIII
Wichtige Gesichtspunkte für notwendige pal. Ar-
(,

beiten).
Artur Men
z Geschichte der griechisch-römischen Schrift
t

bis zur Erfindung des Buchdruckes mit beweglichen

Lettern. Ein Versuch. Leipzig 1920. 155 S. bietet
viele Anregungen, namentlich für die Beeinflussung
der Schriftentwicklung durch Schreibstoffe und ge- —
rate, ph}'siologische, ps3-chologische und kulturelle
 Bibliographie der Lateinischen Buchscrift 77

Verhältnisse; doch muss iede Behauptung nachgeprüft

werden (s. N. Jahrb. XLVII 165. Jahresber. CXCIII
80 f.). Für lat. Buchschrift kommen S. (41). 57-65, 72 f.),
76, 85, 92-116, 120-124, (130-137) in Betracht.
* Luigi Schiaparelli La scrittura lat. nell'età Romana
(Auxilia ad res Italicas medii aevi explorandas I) Como

1921; vgl. DLZ 1921, 350. Archiv. XLIV 151.

6. Edward Maunde Thompson An Introduction to Greek
and Lat. Pal. Oxford 1912. 600 S.
Die treffliche Dar-
stellung ist auf der Erläuterung der Schriftproben
(119 lat.) aufgebaut.
E. Heinrich Zimmermann Vorkarolingische Miniaturen
(Deutscher Verein für Kunstwissenschaft (*). Denk-
mäler deutscher Kunst. Ill Sektion. Malerei I.) Berlin
1916. Die 341 Tafeln enthalten auch viele Proben
von Unziale. Halbunziale und Halbkursive. Die von
Z. statt Halbkursive oder Frühminuskel (vgl. N. Jahrb.
XLVII 164) gebrauchte Bezeichnung Minuskel hat ihn
verleitet, S. 2 von einer Verdrängung der Unziale
durch die kleinteilige ner^^öse Minuskel zu sprechen
und (S. 53 Petersburger Gregor, 72 Cambridger Am-
brosius) den Begriff der StiHsierung (RE II A 712, 24)
mit vielen Worten zu umschreiben. Erläuterungen und
Hss-Beschreibungen des Textbandes berücksichtigen
Schriftbefund und pal. Literatur (*). Die Verfolgung
des vielfach auffällig übereinstimmenden Buchschmu-
ckes führt Angaben über Schulangehörigkeit, die nicht
ohne weiters in Schriftheimat umgesetzt werden dür-
fen, und zu gewissen Zeitgrenzen (so für bestimmte
Fisch- und Vogelformen, die Vorliebe für grelle Far-
ben setzt um 700 ein); das zeitliche Verhältnis ein-

(^) Die von demselben Verein hgg. Berichte über die Arbeiten an den
D. d, K. sind nun überholt.
(') S. 68 heisst es, dass die Entstehung von St Gallen 214 mit der Pro-
venienz des Kodex und der Verwandtschaft mit Bern 611 begründet wird.
Die Stelle bei Lowe (Studia pal. =
Münch. S.-Ber. 1910 XII 49, 1) lautet
aber :Weitere Forschung mag Beziehungen zwischen Luxeuil und Chur oder
einem anderen Schweizer Zentrum aufdecken. Ich vermute, dass Bern 611
und St Gallen 214 unter Luxeuiler Eiafluss in der Schweiz entstanden sind.
— Bei 122 und 251 ist II auf New. Pal. Soc. in I zu verbessern (beim Val-
liceli. B 62 nach II hinzufügen; pl. 7). Im Verz. der Hss Mailand F 87 in 84.
78 Wilhelm Weinberger

zelner Hss aber lässt sich nicht immer mit Sicherheit

bestimmen; s. Zimmermann S. 72, Ottenthai Mitt.
Inst. öst. Gesch. XXXVII 648; Lehmann Ph 1917, W
1621; unten 12, 16, 21, 23, 29; Zß XXXIV 189: I\Ian-
che Lokalisierungen werden über kurz oder lang als
imaginär zurückgestellt werden. Das Individuelle
scheint zu wenig berücksichtigt; schwache Arbeiten
gelten meist als ausserhalb der Hauptschule entstan-
den; s. S. 73 über Burne}' 340, aber auch 74 über
Paris n. a. 1619. Zu beachten sind Z. Datierungen
um so mehr, als sich mit pal. Mittel Bestimmungen
auch Jahrzehnten kaum erreichen lassen. Die Tatein —
werden unten mit der Closter Nummer angeführt.
7. Anton Chroust Monumenta pal. II. Serie 6-24. München
1911-1917.
Tabulae in usum scholarum
sub cura lohannis Lietz-
et
mann. III .Specimina codicum lat. Vaticanorum col-
:

legerunt F. Ehrle et P. Liebaert. 1912. IV: Inscrip-

tiones Latinae. —
collegit Ernestus Diehl. 1912 (in-
schriftliche Kapitale für Kursive s. 29 c).
;

Lindley Richard Dean Index to P^acsimiles in the Pal.

Society Publications. Princeton 1914 (Pal. Society
und 1 Serie der New Pal Society) kann trotz man-
cher Druckfehler gute Dienste leisten. Im 3, Index
(Bibl.) findet mann statt unter London, British Mu-
seum unter B, Westminster Abbey unter W, im 4.
Lat. cap., curs., carol min., goth. min., ins. min., lomb.
une, half une
min., mer. min., precarol. min., visig. min.,
The New Pai. Societv Facsimiles of Ancient Mss I
226-250 (1912). II 1-115 (1915-1921).
Rudolf Beer Monumenta Vindobonensia II. Leipzig
pal.
1913. Proben aus den Bobienses 16 (jetzt wieder in
Italien) und 958 (Sakramentar), die unten 9, 12, 13, 16,
23 und 32 erwähnt werden.
The Liebaert Collection of Photographs irom Latin
Mss (Flugblatt von Lindsay, 1920); ein genaues
Verzeichniss soll in den Studi e testi erscheinen.
M. R. James C. of the Latin MSS. in the John Rylands
Library. Manchester 1921. 2. (Tafel-) Band.
G. F. Warner, Jul. P. Gilson C. of the Western MSS.
 Bibliographie der Lateinischen Buchscrift 79

in the old Rovai and King's collections in the Brit.

Mus. London Ì921.
8. Luigi Schiaparelli Note pal. Segni tachigrafici nelle
Xotae iuris. Archivio storico Italiano LXXII 1914 I
24L LXXIII 1915 I 245-275. Le N. e il sistema delle
i.

abbreviature medievali, 275-322. Intorno all'origine e

ad alcuni caratteri della scrittura (s. unten 17) e del
sistema abbreviativo Irlandese. LXXIV^ 1916 II 1-126.
Reiche Literaturangaben und viele beachtenswerte
(auch über die Abkürzungen hinausgehende) Einzel-
heiten die Erklärungen, namentlich die Herleitungen
:

der Abkürzungen (per. prae. pro. quam, quia, quod,

autem. con. contra, eius. est. et. vel) aus tironischen
Zeichen erfordern genaue Xachprüfung.
Wallace M. Lindsay Xotae Lat. Cambridge 1915 ver-
zeichnet und bespricht die Abkürzungen lat. Hss aus
der Zeit von liXìSòO in 3 Gruppen: Xotae communes.
Xomina sacra. Xotae iuris. Capricious Abbreviations.
Das Hss.-\'erzeichniss ist für viele pal.
Fragen von der grössten Wichtigkeit. \'on
früheren Arbeiten sind hier zu nennen: The Abbre-
viation Symbols cf ergo, igitur. ZB XXIX 1912, 56
und der Zusatz zu Steffens" Xotiz über Abkürzungen
von St. Gallner Hss ZB XXX 488491. von späteren:
A new clue to the emendation of latin Texts, Class. Phil-
XI 1916, 270-277 [Cicero-Papyri aus Oxyrhynchus s. —
unten 11 — mit den Abkürzungen e(ss)e, q(ui)d(em).
t(a)m(en), t(i)b(i), ig(itur). Ancient notae
Pal. lat. 1753],
and Lati:i texts. Class. Quart. XI 1917, 38 (Der Aus-
dehnung der Bezeichnung Xotae iuris auf alle Xotae
antiquae wird eine gewisse Berechtigung zuerkannt.
Es scheint mir besser, Suspensionen und Kontraktio
nen zu scheiden und bei den einzelnen Worten das
Vorkommen in iuristischen Hss zu verzeichnen).
Wilhelm Weinberger Kurzschrift. RE XI 1921, 2217-
2231.Kurze Uebersicht mit Literaturangaben, es wer-
den auch die noch heute wertvollen Teile der für
ihre Zeit trefflichen Uebersicht.
Christian Johnen Geschichte der Stenographie. Berlin
1911 Kurzgefasste G. d. St 1917) hen,-orgehoben (156
80 Wilhelm Weinberger

f. Schriftkürzung, 163 f. Kurzschrift mit unrichtiger

Annahme von Volkskürzungen des Ennius) ferner die
Arbeiten von Mentz und Ruess (s. unten 32) über
tironische Noten. Mentz Die Hs C von Justins Epi-
tome. Hermes LV 196 beweist die stenographische
Vorlage nicht. Für Nomina sacra s. Jahresber, CXCIII
91 f.

II. — Kapitale.

9. Oxyrh. Pap. 1098 (VIII 1911 mit T. VI): Vergil. Beer (pri-
märe Schrift von 16': Lucan (auf diesen und die Ver-
gilhs Vat. 3256 kommen wir bei der Unziale zurück).
Die Vergilhs F (Vat. 3225) bezeichnet Sabbadini Riv.
hl XLVI 1918 wegen Rechtschreibung und Schrift
der Verbesserungen als spanisch.

III. — Kursive.

10. Henry Hartlett Van Hoesen Roman Cursive Writing.

Princeton 1915 ist nicht mehr als eine reichhaltige
Stoffsammlung (besonders dankenswert für Pap}TÌ ;

i vgl.Artur Stein, Untersuch, zur Gesch. u. Verwal-

I tung Aeg3'ptens. Stuttgart 1915, 206). Die Nachzeichn-
ungen einer grossen Zahl von Alphabeten reichen
nicht aus, ein Bild der Kursive und ihrer Entwicklung
zu gehen oder gar die Notwendigkeit der ELinteilung
in Majuskel- und Minuskelkursive zu bestreiten; für

das im 4 Jh. hervortetende Vierliniensvstem vgl N

Jahrb. XLVII 165. Die Bibliographie S. 245-265 hat
geringen Wert.
Richard P e t s c h m a n n Paläographisches. Aufsätze Fritz
i

Milkau gewidmet. Leipzig 1021, S. 281-285 weist auf

die Gewohnheit der Kursive hin, Teile von Buchsta-
ben selbständig zu machen (und wo es bequem ist,
ein Stück des vorausgehenden Schriftzeichens als vor-
dersten Bestandteil des nachfolgenden zu verwerten)
und verfolgt Formen des lat. R von den frühesten
Proben lat. Buchschrift bis zur englischen Urkunden-
schrift.
 Bibliographie der Lateinischen Buchscrift 81

11. O. Schissel von Fleschbiirg" und C. P. Lehmann-

Haupt Eine lat. Grabschrift in Kapitalkursive. Klio
XVI 1920, 129-136 (Facsimile, die Formen sind nicht
die der Papyri, sondern die der Wandinschriften und
Bleitafeln, manchmal auch die der Wachstafeln, vgl.
Tabulae IV^). — AufBedeutung der Kapitalkur-
die
sive für die Ueberlieferungsgeschichte weist Degering
hinten (unten 12).

Papiri Greci e Latini II (Firenze 1913) S. 65 mit Tafel.

Der Papyrus 142 Umarbeitung von \^ergilversen)
wird III-IV angesetzt: das 3 Jh. scheint durch das
deutliche Hervortreten des Vierliniens\-stems aus-
geschlossen. An diese Kursive (vgl. auch die Proben
in
C. of the Greek Papyri in the John Rylands Library I.
Manchester 1911 (griech.-lat. Ciceropapyrus 61),
Ostraka latins de Carthage- Journal d. Savants N. S. IX
1911, 511-522, und die aus Turin G Y 26 bei Lind-
say Notes [unten 19] und in Collezione pal. Bobb. 21)
erinnern die Ciceropapyri Ox3Th 1097 (VIII 1911 T.
VI), 1251 X
1914 (hochgestelltes a und u). die als
Halbunziale mit kursiver Tendenz bezeichnet werden,
mehr als an die kalligraphische Halbunziale. Zwis-
chen den Bezeichnungen Urkunden- oder Geschäfts-
schrift des 5-7 Jh., Minuskelkursive. Halbunzialkursive
schwankt
Brandi Ein lat. Papvrus aus dem Anfang des 6 Jh. Arch,
f. Urkundenl. \' 1914. 269 (mit Lichtdrucktafel). Vgl.
James 1.

Auf die für literarische Zwecke verwendete Buch- oder

Halbkursive kommen wir bei VII zurück.

IV. — Unziale.

12 Bick, Zeitschr. öst. G\mn. LXII

1911, 316 (Nachträge zur
Liste: Traube Oxvrh. Pap. 1073 (VIII T. VI:
I 171).
4 Jh.-). Lowe. Class. Rev.' XXXVI 1922, 154. Chroust
II 10. 3. New Pal. Soc I 230. II 82-85. 101. Beer (pri-

märe Schrift von 16 Pelagonius usw.) Liebaert. Clark

:

T. 4-7 und die unten 25 angeführten Nachträge. Zu

6
82 Wilhelm Weinberger

Zimmermanns zahlreichen Tafehi (nach 137 nur ve-

reinzelt; 204 Stockholmer codex aureus: 2224 Amia-
tinus) seien die Bedenken Ottentais 651 betreffs 44,
45 (Paris-Genfer Augustin) und Lehmanns 1624, 1626
betreffs 74, 75 (Cambridge, St. Gallen, Bern) ange-
führt.
Hermann Deger i n g Ueber ein Bruchstück einer Plau-
tushs des 4. Jh. Beri. S.-Ber. 1919, 465, 497 (Beri. lat.
qu. 784; Purpurtinte; Schriftprobe). Now rejected, as
a forgery, bv Châtelain, Bull. Acad. Inscr. Mai-Juin
1922; cf Lowe, Class. Rev 1923. ED.J.
Paul Lehmann Die Konstanz-Weingartener Propheten-
fragmente (CGL Suppl. IX 1912)^
Raccolta di codici riprodotti in facsimili della R. accademia
di Torino: Il codice evangelico k della Bibl. Univ.
Naz. di Torino 1912. —
Bei dieser Hs, der im Titel
genannten, und 2 Cyprian hss stellt.
Elias A. Lowe On the African Origin of Codex Palatinus
of the Gospels (e) JTh St XXIII 1922, 401-404 die
Eigentümlichkeit fest, dass die Buchstaben von den
gezogenen Linien durchschnitten werden.
[Turner cf. 14) has 8 plates of a Worcester Cathedral
fragment of Jerome on Matthew in Spanish uncial. ED].
13. Lowe stellt auch 2 Kapital- und 16 Unzialhss zusammen,
die am Beginn der Seite der Kolumne) und mitten
im Worte einen grossen Kapitalbuchstaben haben.
Die Beobachtung gewinnt am Wert, wenn man daraut
aufmerksam wird, dass nicht nur alle 4 Palimpseste,
der schon in 9 genannte Lucan, der Plinius von St.
Paul, dessen sekundäre Schrift schon Châtelain Pal.
class, lat. II 13 auf Verona oder Bobbio bezog, der
Veroneser Gaius (mit dem auch I App. f. 10 zusam-
menhängt) und der Ambrosianus C 73 i., sondern auch
der Ambrosius von St. Paul (Abbildung in: Die Rei-
chenauer Hss [Die Hss der... Bibl. in Karlsruhe VII]
III 2, 1918 T. 5) Verona XIII Hilarius) und XXVIII
(Augustin) auf Cassiodors Bibl. in Vivarium zurück-
geführt werden oder sich doch zurückführen lassen (*) ;

(*) Von den übrigen Hss will ich nur bemerken, dass der Pariser Livius
(3- Dekade) und der Petersburger Augustin aus Corbie stammen und der Li-
 Bibliographie der Lateinischen Buchscrift 83

Vgl. Beer, auch seine Bemerkungen über den ältesten

Hssbestand des Klosters Bobbio. Anzeiger der Wiener
Akademie 3. Mai 1911. S. 14-16 und 24 des Sonder-
druckes werden die meist palimpsestierten Veronen-
ses II, XXXVIII (allerdings schon 517 in Verona zum
zweitenmal beschrieben). XXXIX, XL, LV und LXII
herangezogen ^). Da wir uns unten 22 mit der sekun-
dären Schrift von \^erona XL beschäftigen müssen,
die grosse Aehnlichkeit mit der des Plinius zeigt, sei
gleich bemerkt, dass ihre Zurückführung auf Bobbio
(Vivarium) durch die charakteristische Verwendung
kleiner Teile verschiedener Hss gesichert erscheint.
Der reskribierte Teil zeigt Vergil in Capitalis qua-
drata, deren Aehnlichkeit mit dem Lucan Beer nach
Autopsie bestätigt (Traube I S. 162. X. 3; wo als
Bibliotheksheimat Bobbio angegeben wird, den für
Cassiodor bezeugten Euklid in Capitalis rustica a. a.
O. 171, 23 Verona) und Livius 1. Dekade in Unziale
(251, 49 Verona: die sekundäre Schrift ist versehent-
lich als Unzial bezeichnet).
14. Lowe erwähnt auch die Pliniushs Morgan M 462, s.

* E. A. Lowe and E. K. Rand A Sixth-Century Frag-

ment of the Letters of Pliny. A stud\' of six leaves
y
of an uncial Ms.. Washington 1921.
* C. H. Turner
Early Worcester Mss.. photographically
reproduced. Oxford 1916. (Bruchstücke des 8 Jh. von
Evang. halbunz.], Gregor Reg. past, [halbunz. Pa- .

terius).
E. C. Armstrong and H. T. Lawlor The Domnach
Airgid. Proceedings of the Rovai Irish Academv
XXXIV 191« S. VÌI 96. Die Hs selbst hat keine Be
Ziehung zu Patrick. Nach brieflicher Mitteilung des

vius auf Avellino bei Neapel zurückgeführt wird und Wien 1235, eben weil
er aus Neapel stammte, jetzt nach Italien zurückgeführt wurde. Eine gra-
phische Aehnlichkeit des Pariser Livius mit dem Wiener Plinius-Bruchstück
(in Titel) erwähnt Châtelain a. a. O.
(^) Es Hessen sich auch andere Veronenses mit Vivarium in Verbindung

bringen. Für einen Ambrosiaster-Palimpsest (s. Beer 17) in Wolfenbüttel s.

ThLZ 1917, 71 den Ambrosiaster von Monte Cassino weist Traube I S. 202
;

N. 116 der Bibl. des Eugippius zu.

84 Wilhelm Weinberger

Herrn Professors Lindsa}^ kann sie auch aus Rom

mitg-ebracht worden sein.
Texte und x'\rbeiten hgg. von der Erzabtei Beuron I 5,
1918: Dold bezeichnet das Palimpsest Sakramentar
des Mainzer Priesterseminars in Goldschrift als vorh-
adrianisch. Mohlberg setzt es Theol. Revue 1919, 20
in das 4 Jh; s. DLZ 1921, 1616. LZB 1920, 1461.
15. Neben der reinen Unziale, die —
vielleicht in diokletiani-
scher Zeit —
durch Aufnahme stilisierter Rundformen
der Kursive für a, d, e, h, in, q und 21 in die Kapitale
entstand, gab es eine flüchtigere Unziale mit verein-
zelten kursiven Formen (etwa r und j- am Zeilenende ;

von der .Art war nach Garrod, Lombardie. Class.

Quart. XIII 51 der Catull- Archetyp, den Scaliger lom-
bardisch nannte), aber auch eine Mischschrift, der man
etwa den Namen Kursiv-Unziale geben könnte fdazu
würde ich auch rechnen, was Lowe JTh St XIV 1913,
335 den bd-Typus der Unziale nennt. Wenn Mentz
73 mit Berufung auf Thompson, der aber S. 298 die
Livius-Epitome von O.xyrhynchos nur als ancestor of
the mixed-uncial and half-uncial Mss. bezeichnet, which
form the subject of the present chapter, von Halbun-
ziale spricht, möchte ich diesen Namen lieber der kal-
ligraphischen Halbunziale vorbehalten.

IV. — Kontinentale Halbunziale.

16. Beer (primäre Schrift von 16). Chroust II 6, 8. Clark (s. 25)
1-3, 8, 9. Liebaert New Pal. Soc. II 5, 6. Zimmermann
Ib, 29, 40-43, 85% 92, 102, 115-119, 129, 130-134, 138-
141. Mentz 96 nimmt den Vorschlag auf, die Halbun.
ziale als Minuskel zu bezeichnen; vgl RE II A 731,50.
Spicilegium palimpsestorum arte photographica paratum
per S. Benedicti monachos Archiabbatiae Beuronensis
1 1913: Codex Sangallensis 193 continens fragmenta
plurium prophetarum (primäre Schrift).
Henri Omont Fragments d'un très ancien ms lat, pro-
venant de l'Afrique du Nord. Comptes-rendus de l'Ac.
Belles-lettres 1918, 241-250 (Par.
d. Inscr, et lat. n. a.

1114 aus Tebessa; Schriftprobe).

 Bibliographie der Lateinischen Buchscrift 85

\ I. — Insulare Schrift.

17. S. oben. 14 Turner; Chroust II 6, 9 (Halbunziale. letzte

Zeile jeder Seite Minuskel = Zimmermann 254 22 ;

6-10. Xe\v Pal. Soc. I 231-234. II 11. 33-35, 58-62. 81

102.
Lehmann
10 angelsächsischer Kinfluss auf Deutschland),
39 und 50 (Literaturan^^-aben). Schiaparelli (oben 8)
glaubt dass die irische Schritt (mit ihren Abkürzungen)
unter dem Einfluss der einheimischen Ogamschrift im
5 Jh. entstanden sei, und bespricht 116 ff. den Ueber-
gang zur gewöhnlicher Minuskel (vgl. Lehmann 13).
18 H. M. Bannister Fragments of an Anglo-saxon Sacra-
mentarv. JTh St XII 1911. 451 (Köln. Stadtarchiv;
S. Jh.).'

W. M. Lindsay Breton Scriptoria. ZB XXIX 1912,264-

272 (2 insulare Hss).
— Early Welsh Script. St Andrews Universit}- Publications
X 1912. 17 T. Scharfe Charakterisierung ist nicht /
möglich, wie für die Abkürzungen ,2 Formen für est,
V mit übergesetztem Komma ^
ut) auch die Notae
lat. bestätigen. Zu T. 1 (Book of Chad in Lichfield)
vgl. Morris-Jones, Taliesin. Cymmrodor XXXMII 1918,
261-279 sowohl die Halbunziale als auch die Minuskel
:

der mit Surexit beginnenden Eintragung gehören in

das 6. Jh. Es ist gewiss denkbar, dass sich hier wie
anderswo \iür Corbie s. unten 31) Halbunziale und
Halbkursive neben einander entwickelten. Für die
L'ebereinstimmung der insularen mit den anderen
Uebergangsschriften (MI) kommt ausser der gleichen
Entwicklung der Kursive (für kursive Elemente der
insularen Schrift, namentlich z vgl. Lindsav, Irish Cur-
sive Script. Z. f ceh. Phil. IX 1913. 301-308: 2 T.)
die Einwirkung insularer Schreiber in Betracht.
19. Louis G ou gaud Répertoire des facsimiles des mss. Ir-

landais. Rev. celt. XXXI\^ 1913. 14-37. XXX\' 415-

430. XXVIII 1-14.
W. M. Lindsay Palaeographical Xotes. The Script I.

of the Cathach. II. The Colophon of the Durrow Book.

86 Wilhelm Weinberger

Appendix II (397403) zu H. T. Lawlor The Cathach

of St. Columba. Proceedings of the Royal Irish Acad.
XXXIII 1916 C 11, 241-443 (T. XXXIII Halbunziale
des Cathach. T. XXXIV Halbunziale und Halbkursive
von Turin G V 26; s. oben 11). Lindsay weist daraut
hin, dass wenige Proben der lat. Bedarfsschrift erhal-
ten sind, diese aber dafür sprechen, Minuskelelemente
weit früher anzusetzen als es bisher üblich war (vgl.
Jahresber. 135, 23 f.). Der Psalter könne also sehr
wohl aus dem 6. Jh stammen (Zimmermann S. 21 setzt
ihn an das Ende des 8. Jh.). Die Subskription des
Durrow Book dürfte eine Abschrift von Columbas
eigenhändiger Eintragung sein. —
S. 399 A. 2 kommt
L. auf die Notae 321 und 457 geäusserte Ansicht zu-
rück, Bonifaz könne Fulda 3 in seiner Jugend bei man-
gelhafter Kenntnis des Irischen in Südengland geschrie-
ben und dabei auch die Subskrij)tion seines Freundes
Cadmug kopiert haben. Später hat er sich vielleicht
eine bessere Schrift angeeignet oder doch auf die
Randbemerkungen in Fulda 1 besondere Sorgfalt ver-
wendet. Zimmermann T. 205 c setzt die Hs gegen
800, Lehmann um 750, also noch zu Bonifaz' Lebzei-
ten, aber nicht in seiner Jugend. —
Auf Lawlor, des-
sen Index (Mss.) gute Dienste leisten kann, kommen
wir bei der karolingischen Minuskel zurück.
20. Paul Lehmann Enim und autem in mittelalterlichen lat.
Hss. Philologus 1916, 542-547 (als Erweis insularer
Vorlagen von kontinentalen Hss). Die Abkürzungen
spielen auch eine grosse Rolle bei der Scheidung (s.
auch Lowe Ph \V 1920, 1174) angelsächsischer Hss ;

nur in diesen qnd, qndo für quando, t mit Durchkreu-

zung des Querstriches für tur (s. auch Lindsay —
unten 26 — S. 428), Kürzungen für omnis nur in iri-
schen.
Hermann Degering Ein Alkoholrezept aus dem 8. Jh.

Beri. S.-Ber. 1917, Die Zurückführung auf das

503.
8. Jh. beruht auf Fehlern, die sich durch insulare Vor-
lage erklären, z. B. igritum statt ignitum.
21. Nach Zimmermann hat England den übernommenen irischen
Buchschmuck ausgebildet und dem Festlande über-
 Bibliographie der Lateinischen Buchscrift 87

mittelt.Das oft für irischen Einfluss angeführte St.

Gallner Sakramentar stamme aus Chur (s. unten 33)
und verarbeite angelsächsische Motive. Aus seinen
Erläuterungen zu T. 160 ff. (irische Hss, Lindisfarne-,
Echternach-, Canterbur\'- und südenglische Gruppe)
sei ferner hervorgehoben, dass er das Book of Durrow
und das Book of Keils gleich dem Lindisfarne-Evan-
geliar an den Anfang des 8. Jh. setzt und das Lin-
disfarne-Evangeliar wie den Amiatinus auf Cassiodors
codex grandior zurückführt. Bei St. Gallen 51 scheint
mir Lindsa}s Urteil (\otae 483; dass die 3 letzten
Zeilen (Zimmermann 1920) zwar wie ein später Zusatz
aussehen, es aber schwerlich sind, noch immer beach-
tenswert. Wenn Z. S. 126 behauptet b und d seien
in der Halbunziale des Epternacensis (Par. 9389) often
fin der von IMaihingen I 2. 4.*" 2 geschlossen), so zei-

gen die Tafeln, dass das einstrichige b immer gesch-

lossen ist, das zweistrichige d allerdings erst in der
IMaihinger Hs.

VII. — Uebergangsschriften Halbkursive).

22. Das Nebeneinander unzialer, halbunzialer, kursiver und

Minuskelformen verschiedenen Schreibschulen (vgl.
in
Palae. Lat. I 1-66) zeigt wie insulare (vgl. oben 18),
merovingische, spanische und süditalienische Eigentüm-
lichkeiten auf die Kursive zurückgehen. Uns diese Ein-
heitlichkeit der Kursive klarzumachen, hindert uns ei-
nerseits die geringe Zahl von erhaltenen Stücken der
Bedarfsschrift (Lindsav Kotes), andererseits die Nach-
wirkungen von Mabillons Nationalschriften. Es scheint
auffällig, dass frankoburgundische und spanische Hs
ihre Bibliotheksheimat haben. Thompsons
in Italien
Bezeichnung: Franco-Lombardie hängt vielleicht damit
zusammen, dass Traube II 28 die Schrift von Verona
XL als scriptura Luxoviensis bezeichnet (A 4, in der
die Hs eigener Art er-
als Beispiel für Kursivschrift
scheint, Zusatz Lehmanns gekennzeichnet). Lowe
ist als

(oben A. 2) 31 hat dagegen mit Berufung auf die

gleiche Schrift von Paris 9427 (s. Zimmermann 34 a
88 Wilhelm Weinberger

— ohne Buchschmuck — 53-57) die Hs auf Luxeuil

zurückgeführt.Nach dem oben 13 Gesagten kann die
sekundäre Schrift von Verona XL und St. Paul 25.
2. 36 (Plinius; Die Reichenauer Hss. III 2 T. 4) nur
in Oberitalien entstanden sein, man könnte allerdings
an einen Mönch denken, der von Luxeuil, gleichfalls
einer Gründung Columbans, nach Bobbio kam. Auch
bei einer der " spanischen " Hss italienischer Bibl.,
der Bibel von La Cava, gilt der Name des Schreibers
Danila als gotisch. Es wird aber doch zu untersuchen
sein, ob nicht in Italien (s. unten 24) in verschiedenen
T\-pen der Kursive (mit verschiedenen Abkürzungen)
geschrieben wurde, die dann in Frankreich oder Spa-
nien herrschend wurden. Ivrea 1 (Z. 48, 49) könnte
nach der Uebereinstimmung des Buchschmuckes mit
dem oben 12 erwähnten Augustin französisch, könnte
aber auch, etwa nach französischer Vorlage, in Nord-
italien entstanden sein (vgl. Traube Archiv XXVI
230).
Wenn S. Berger, de la Vulgate. Paris 1893, 58 f.
Hist,
einen Teil der Korrekturen and zwei Zusätze im Texte
des Ambros. I 61 s. (aus Bobbio) als merovingisch
bezeichnet, so zeigt wenigstens die Kursive des jetzt
bei Steffens 27 b abgebildeten Zusatzes keine mero-
vingische Eigentümlichkeit. Die anderen Korrekturen
bezeichnen Berger und Steffens als irisch; der Text
ist in insularer Halbunziale, vermutlich auf dem Fest-

lande, geschrieben (Berger denkt an Bobbio).

a) Norditalienische Kursive.

23. Zimmermann 19 e-h (Mailänder Maximus, bisher in das 7.

Jh. gesetzt, von Lowe O. bei Besprechung einer

a. a.

z- Form ins 8. Z. setzt ihn, weil mit den Formern der

Schule in dürftiger Weise gearbeitet, nach dem Mai-

länder Gregor, mit dem der Buchschmuck charakte-
ristische Aehnlichkeit zeigt. Der Gregor wird 750 an-
gesetzt, weil der genannte Auftraggeber, Abt Ana-
stasius, durch eine Urkunde von 747 bezeugt ist. Vor
700 scheint der Maximus keinesfalls entstanden zu
sein), 24 d (im Liebaert Verz. als Ital. precaroline
 Bibliographie der Lateinischen Buchscrift 89

s. IX in. aus Novara stammt auch Steffens

bezeichnet ;

42 b 42 a aus Montecassinoj), 37 (Augiensis LVII;

CCXXIX —
verkehrte Beta-Form für assibiliertes ti ;

s. auch Zimmermann 34 b —
Chroust II 10. 10), 34
(Verona; dieselbe ti-Form). Ehrle 10. New Pal. Soc.
II 13. Beer (sekundäre Schrift von 16; S. 39, 50, 53).

Fumag-alli, Pal. greca e lat. di E. M. Thompson ^

Mailand 1911 (Manuali Hoepli 2 Serie 140) T. 2.

Ò) Süditalienische Schrift.

24. E. A. Loew (jetzt Lowe) The Beneventan Script, Oxford

1914 verfolgt die Schriftart (über den Namen vgl. Ot-
tental Mitt. Inst. Oest. Gesch XXX\1 205: Proben
auch bei James 2-7) vom 8. Jh. (für den Zusammenhang-
mit norditalienischer Schrift s. New Pal. Soc. II 13)
bis in das 13. 14. ist zweifelhaft) und geht in den
Abschnitten, die Abkürzungen und Interpunktion be-
handeln, über diese Schriftart hinaus. Für die Unter-
scheidung des assibilierten ti und den Gebrauch der
I longa in verschiedenen Schriftarten verweise ich hin

auf 22 (Palae. Lat. I 23 (s. auch Z. 45 c. 49, 57 b) und 25.

.

c) Spanische Schrift.

25. A. Bur na m Pal. Iberica. I Paris 1910 (Pariser Hs der Lex

Romana Visigothorum vom Jahre 828). TI 1920 (7 span.
Hss, s.Rev. 1922, 156).
Hist.
Charles Upson Clark
Collectanea Hispanica (^Transactions
of the Connecticut Academy of Arts and Sciences
^
XXIV'^) Paris 1920. Reichhaltige Stoffsammlung mit
guten Zusammenstellungen und 70 (verkleinerten) Pro-
ben. Nachträge N. Jahrb. XLVII 163 f. (ti, tj). Jahresber.
:

CXCIII S. 89. Hist. Rev. 1921. 465. Ph 1921, 321. W

ZB XXXIX 345, vgl. James 109-111, 120-122. 136. Bei
Datierungen scheint der Uebergang von der Kursive
zur Minuskel beachtet werden zu müssen; vgl.
Alfred He s sei Studien zur \'erbreitung der karolingi-
schen Minuskel. I. Spanien. Archiv f. Urkundeniorsch.
VII 1921, 197-202.
[Henri de Ville de Mir m ont Ausonius. Codex... Vos-
sianus Latinus 111. (phototyp.) Paris 1919. ED.].
90 Wilhelm Weinberger

d) Merovingische Schrift. Vorkarolingische Minuskel.

26. \V. M. Lindsay The Revue bibl.

old script of Corbie.
XXII 1912,405429 mit Abbildung aus Montpellier 69.
Verz. von Hss des ab-T}-pus —
die S. 412 angeführte
Hs von Karlsruhe ist Reichenau Fr. 140, für Düssel-
dorf B 3 s. jetzt Chroust II 24. 3, Notae XI 1 werden
Bordeaux 21, Köln Stadtarchiv B 140 nachgetragen
[cf Revue bibl. XXIV 24 A. 2 ED.] - a = ic, b mit
horizontalem Yerbindungsstrich zum folgenden Buch-
staben. Abkürzungen angelsächsisch. Von den 35 an-
geführten Hss erscheinen 15 auf Zimmermanns Tafeln.
— The Laon aztype XXIV 5-27. a besteht aus zwei win-
kelförmigen c, der obere Strich des z ist homeartig ver-
längert [cf Palae. Lat. I pl. I 81 ED. Die hier erwähnte
.

nordostfranzösische Gruppe: Paris 2110 (Z. 131-134),

2700 (Z. 129 f). Reg. 11,316 (Z. 135-137), Autun 20,
Montpellier Stadtbibl. 3, O.xford Douce 176 (Z. 142 f.),
fr. 1 wird Notae xi als eine vielleicht unreal combi-
nation bezeichnet.
27. S. Tafel. Un type français de minuscule précarolingienne.
Revue Charlemagne II 1912, 105-115 versucht die eben
genannte Hs von Montpellier (Zimmermann 128 g;
Antang des 8 Jh.) einzureihen (viele kursive Formen,
aber Regelmässigkeit der Minuskel), wobei er ausser
Reg. 316, Autun 20, Douce fr. 1 den Reg. 317 (Z. 44 f.),
dessen Minuskel Traube Nomina sacra 221 den 1-T3pus
nennt (vgl. Z. 102 b. 107 a) und in Burgund zu Hause
denkt, und Pal. 493 (Z. 101*- 101 ^*'***) heran. Die
Uebereinstimmung späterer Hss von Corbie würde
sich durch Luxeuil. das Mutterkloster von Corbie,
erklären.
^ D. A. Wilmart Rev. Benéd. XXX 1913, 439, 1.

28. Wilhelm Meyer Drei Gothaer Rlnthmen aus dem Kreise

von Alkuin. Göttinger Nachrichten 1916, 647, 651 (mit
Tafel) bespricht die Mischschriften, die schon viele
Eingentümlichkeiten der karolingischen Minuskel zei-
gen (vgl. Chroust II 10, 3, 5 a. New Pal Soc. II 10
Cambrai) und bezeichnet die Schrift von Corbie als
einen der im Uebergange vom 8 zum 9. Jh. in Frank-
 Bibliographie der Lateinischen Buchscrift 91

reich gemachten Versuche, eine kleine und bequeme

Buchschrift zu konstruieren.
* Henry Bradshaw Society LH 1917: Unziale und Minuskel
des schon genannten Regin. 317, mit dem nach Clark
S. 125 die Hs von Barcelona manche Aehnlichkeit
haben soll.

7 Paul Liebaert Some early scripts of the Corbie Scrip-

torium, Palae. Lat. I (s. oben 22) 62-66. 4 T. unter-
scheidet en-, Leutchar- und Maurdramnus-Typus.
29. Wenn man die wiederholt angeführten wertvollen Zusam-
menstellungen Lowes und Erläuterungen sowie Tafeln
Zimmermanns (44-74 Schule von L u x e u 1 85-1 18 i ,

von Corbie ('), 119-126 unter deren Einfluss stehende

Hss, 127-143 nordostfranzösische Gruppe 144-
149 a-T3'pus [Laon az-Typus. ED.] berücksichtigt, er-
geben sich für die weitere Forschung etwa folgende
Gesichtspunkte. Als Schrift mit besonderen Eigentüm-
lichkeiten (die Verbindung von e mit in, ?^, r ist in
der Kursive allgemein üblich) tritt uns zunächst die
von Luxeuil entgegen, für die auch die Zierbuchsta-
ben der Titel charakteristisch zu sein scheinen, ferner
a (aus dem sich —
wie Ottenthai vermutet, vielleicht
in Laon —
das a des az-T3-pus entwickelt), o und
gespaltene p, r und s. Es ist möglich dass diese Schrift
eine Zeit lang auch in Corbie verwendet wurde. Bald
aber beginnt der ab-Typ sich vorzubereiten, zunächst
mit dem b (geknicktes b Z. 91 b, für geknicktes / s.
oben 26); der horizontale Verbindungsstrich kommt
auch bei anderen Buchstaben vor. Andere Hss aber
(vgl. eine Hand des Par. 17655: Z. 90, die bei Z. 116
abgebildete Kölner Hs und einige der Hss, deren
Buchschmuck nach Z. noch unter dem Einfluss von
Corbie steht) zeigen diese besonderen Eigentümlich-
keiten nicht; s. Z. S. 78. der in einer Polemik gegen
Tafel bemerkt, dass die Provenienzen der Hss in den
Nordosten des Frankenreiches weisen, wohin auch
die Verwandtschaft mit den a(z)-T}-p und das Hinüber-

(M Lehmann's Bedenken (1623) ob alle Luxovienses, die Petersburg Q

I 14 vorangehen, später sein müssen als der Corbier Gregor (Z. S. 63 f.),

stimme ich bei.

92 Wilhelm Weinberger

greifen nach Deutschland spreche. Das Endglied der

nordostfranzösischen Gruppe (T. 141) sei sicherlich für,
Köln geschrieben; vgl. (ausser James 36,
vielleicht in
37) auch Ehrle 27, 28, deren Buchstabenformen nach
Mentz 115 für die karolingische Minuskel massgebend
waren. Seine Nachzeichnungen beweisen das freilich
nicht ; aber der Gedanke ist beachtenswert (vgl. Lowe
Stud. 35 f. über Epinal 68; Thompson 130, Z. 122).
30. Bei James findet sich ein mit beträchtlicher Krümmung
nach links ausgreifendes h, das nach Lindsay Palae.
Lat. I 48 für burgundische Hss charakteristich ist,
aber auch in anderen Schriftarten vorkommt,- vgl. bei
Zimmermann ausser Montpellier 3 und Reg. 316 T.
90 a, 91 ^ 105 b, 111 b, 219 e (Autun 4), endlich Chroust
II 10, 4 (Augiensis ccxxii).

31. In Corbie bediente man sich in denselben Hss oder in

solchen, die durch Aehnlichkeit des Buchschmuckes
zusammenhängen, der Halbunziale. Sie nähert sich
immer mehr der karolingischen Minuskel; vgl. aber
Lehmann 10: Man wird besser von der Anpassung
schon vorhandener Minuskel an ältere und jüngere
Halbunziale als voni Uebergang von der Halbunziale
zur Minuskel reden.

VIII. — Karolingische Minuskel.

32. Die Entwicklung lässt sich an vielfach örtlich und zeitlich

bestimmbaren Hss verfolgen bei Chroust II 6. 10. 7,
1-8, 6. 9, 7-10. 10, 1, 2, 5-9. 11, 1-6. 19, 6-10. 20, 1. 21,
9-22, 2. 23, 1-4. 24, 5-10; vgl. Beer (958; S. 62). fames
23-28,33 f 113-117. 124. 133. 148-153. 157. 181-185. New
Pal. See. I 236 f. u. s. w. Hessel (oben 25). Mentz
(oben 29).
Die Kasseler Hs der tironischen Noten [in Lichtdruck] hgg.
von Friedrich Ruess. Leipz-ig und Berlin 1914.
Cicero. Operum philos. Codex Leidensis Vossianhs lat.

Fol. 84 (C G L XIX 1^5).

Lucretius. Codex Vossianu.9 Quadratus (C G L XVIII 1913)
für Buchschmuck Wilhelm Kohler in den A. 1 ange-
führten Berichten I 52. II 52. III 71.
 .

Bibliographie der Lateinischen Buchscrift 93

Peintures et initiales de la première et de la seconde Bible

de Charles le Chauve (PBX 19,20 1911
* E. K. Rand und G. Howe X'atican Livv and the Script
ot Tours. Memoirs ol the American Academy in Rome.
I School of Class. Studies 1915 6.

A. Merton Buchmalerei von St. Gallen. 1911.

Franz Landsberger Der St. Gallner Folchartpsalter.
1912; und für das Prümer Evangeliar Degering in :

15 jähre k. und Staatsbibl. Harnack zum 31, III, 21

ueberreicht 122 (vgl. Archiv XLIII 628. XLV 135),
für die Palastschule in Aachen Lawlor 434 (Alkuin
seit 782: Goldhss, Inhaltsangabe von Psalmen), ThLZ
1922, 371. Mentz (oben 27).
33. Lehmann (oben auch (vgl. Münch. S.-Ber. 1920
17, 31) wirft
X 18 A.) die Frage auf, ob Eigentümlichkeiten alter
Minuskel von St. Gallen oberitalienischen, rätischen
oder fränkischen Urspsungs sind. Es dürfte sich um
eine Uebergangsschrift handeln, die von der nordita-
lienischen ausgeht und von der karolingischen beein-
flusst wird (Alkuin war mit Bischof Remedius von
Chur befreundet); vgl. ZB XXX
477 ff. über St. Gallen
272 IX frühkarolingisch mit St. Gallen Duktus. XXXII
;

qq. über Spicilegium (oben 16) und

Kunibert Mohlberg St. Gallen Sakramentarforschüng I
(Liturgiegeschichtliche Ouellen Münster 1918 S.
'
.,

LXXXIX und C.
34. Hans Foerster Die Abkürzungen in den Kölner Hss der
Karolingerzeit. Tübingen 1916. Das Verz. der Kürz-
ungen (von 14 Hss aus der Zeit Hildebalds und von
5 aus der Zeit von 820-921) in zwar umfangreichen,
aber doch nicht immer übersichtlichen Tabellen ist
dankenswert. F. findet es auffällig, dass t mit überge-
schriebenem c für tunc nur 1 mal vorkommt, tc öfter,
n mit übergeschriebenem o für vero aber einigemale
vorkommt (uo 2 mal in einer insular beeinflussten Hs).
Die Zahl der Stellen ist zu gering, als das ein sicherer
Schluss gezogren werden könnte.
 BIBLIOGRAPHIE

REGISTER

Armstrong 14 Lindsay 7, 8, 18, 19, 22, 24, 26

Bannister 18 Lowe 12-15 22, 24, 29
Beer 7 Mentz 5, 8, 15, 16, 29
Bick 12 Merton 32
Brandi 11 Meyer Wilhelm 27
Bretholz 4 Mohlberg 14, 33
Bumam 25 Morris-Jones 18
Cagnat 4 Omont 16
Chroust 7 Ottenthai 6, 24
Clark 12, 25, 27 Pietschman 10
Dean 7 Rabe 2
Degering 11, 12, 20, 32 Rand 14
Dold 14 Ruess 8 '

Ehrle 7 Sabbadini 9
Förster 34 Schiaparelli 5, 8, 17
Garrod 15 Schissel 11
Gilson 7 Society New. Pal. 7
Gougaud 19 Spicilegium palimps. 16, 33
Hessel 25 Tabulae 7
van Hoesen 10 Tafel 27
James 7 Thompson 6, 22
Johnen 8 Traube 3, 4, 13, 22
Kögel 2 Turner [12], 14
Landsberger 32 Warner 7
Lawlor 14, 19. 32 Weinberger 4, 8
Lehmann 5, 6, 12, 17, 20, 31, 33 Wilmart 27
Lehmann-Haupt 11 Zimmermann 6
Liebaert 7, 28
Yji)

.y<t«nÄiiili,-..t.-rjffUjn^R<*P"taic.uyöi'i«7=lmMix.

««4iÌHi^a^pr*K,iVi.„|.iY'[f!-c«i4òCt»?lio.if4V""'

ti
«ìsju^jìVmtìN, j X|..no^- M' ''>
' *i « f>
--
î\l 1,- -Ik UM

n'>1!'I><>fJWT<|ì>itA||''^''>'ìi'4i'«'i-<|al"'>>"'ì*,"''»/<''ì/-'-

V^ikvl,tJMiin;,^i:v^,.|,,,t.„yî,|-,„[„|,.Y-.lK,u)V<' l»i^+<"iì«|"t>#rK->»ii}'(fiJnir£-r«cL|"|<nyA.-i'Mt>mii-j\i.-j;

Uli>iin«iii|"auir-Jo. iij'.\j„*<»k<[.«tiN|>«*i-;'-n<*fi..

«whim'i tì»n]ì<. ii.lw>nt*'tä?ttm(4inx<

1 nw«^) Ti ..
'

mlq-m. n.|„U-(l•^<1fbT^4l<l<,n^l,ll«;?|^n,ll-^l» i.io

v.li|^'i.iilUY1i|^),-<.o^<4«,'v|ô<ifTv»<«cui|iiiii\in.r W*# "IpmiJ^^ |>U«»H«' hominiU t<.lril.^;;^^^J,l

^^-

litifiiuioi^^uiCii^.rOiiLii.U^m-r-WKwii^yirA;'

(i>rujT>th>\l'aû' -Ii»iii.M)»<iînlKhnmKiij?fîiMi<}li»vj

JifinfrctiiiilivlîUJir; '- -J ^^•'i'i»K|'\-|ltinilL;ii„,

n.Trr fre^'mlf.MoMtitcSJiar^.:
ifWln

1 .
- ) I . J./ol
rj^

vi*»

n
L tfi

4
^tin]
 -vili II
"''
'•* •>"*' fc-A ^^

"
k^i-um «iumai-u.M
. (I
JJ'. II
I
ill
" ""• •^>>«»-u.^

'l»1,„V,>A^-x,-u..^5.., „,„a„„j„. „,,r..

"^" •^-"" .
r- I
'»'l-« p...«-
, f

j|U,r t;..,iil. *.„„«? --L, ^ „

[.V.,,..
--""i.sj.^Jf „»^U--î.„.,„^fc}'..,,.„^.e^^
^^ ,„...r,>.„. ?

\fU,|,,»,,*^,..„.^-„,.,.,,^..,.,,,,_.,,.y„t.._^j,,,.j_, __^^^^_,^_^j___
^

O^ „„ .J»..;.Arp,^^,... ^„.J,^,, ,,„„-,,„„. ,

^„.;. ,;.,, ,
^, ,,,.j7 .>,-./, ,^,,„>,.

-"• j*-|.,-i>.-„
Ä^-L-N...» A.-. Ln.. W..- c,.;. „i..,-. ,yU JL^.^
* T'V- Vu^«-I.„x.„ .5..^-l^.™r>n„-|„.i. ^,,
\,..Lx^„,.xU.J.I,.
•""' 5--J*'

A--,-L ..L .
]...-~r.l.V ,...^„^.v,„„: Vi«.- ....,.>.v„-..-t„„-

-.v,,,; j» tó „. H-^ .=*

S,i..>.%„-

lv..v J.,..,- /.,:j„. !.,../.,.. ),.,^..

#
J— u... ...1 .4l.iU„t„, t-,f|,-. H...?U..-).M L--4„.- wLj
«If Ì.V lw.J„.n- o.J,„ jil,„-..l..n-. ^,c),d p„™i-
J...™-
Ji«i.|,- /^.-A-,.,»- p.

-, P*""
L.«-.
''*•«•"• *..,"U.!,,

p:.r,.l.„-,.^.„.,..>l„,i
T>..„«,U- ^^v,

^,„f,..J„,v/-.,vtó„v
,;/„„Y"M«';""-'f"""''-

H-"?-«»-!«,«^.™-
"
M
<"
n;
University of Toronto

Library

Acme Library
Uader Pat. "Rrf. Ino«.

Made by LIBRARY BUhJAU

 •/>
1*':^^^
tm