University of Virginia Library

Search this document 


expand section 
expand section 
expand section 
expand section 
expand section 
expand section 



Page 97

Daniel W. Mosser

In this essay[1] I undertake a reconstruction of four paper manuscripts— British Library MS Arundel 140 (Ar), British Library MS Harley 2382 (Hl3), Magdalene College, Cambridge MS Pepys 2006 (Pp), and British Library MS Sloane 1009 (Sl3)[2]—the collation of which has been characterized variously as “impossible” (Manly-Rickert 1:52, 516), “impracticable” (James 60; Manly-Rickert 1:406), “unknown” (Seymour 1966, p. 184), and “too tightly bound to be determined, even by watermarks” (Seymour 1995, p. 134). All four MSS do present interesting and challenging problems for the reasons described by Manly-Rickert and others: in Ar and Sl3, the material structures have lost their original integrity and the leaves are now mounted on guards (the guards can present problems for the detection of watermarks in 4° format); Pp is tightly bound, making problematic the detection of some of the watermarks in 4° structures; textual lacunae in Hl3 and Pp can be difficult to determine because some of their texts are extracts.

The establishment of the present structures of these MSS and the reconstruction of their original structures (to the extent that these goals are attainable) provide important information for textual editors about their texts' original integrity or incompleteness. Texts of the Canterbury Tales that circulate independently of the frame provided by the poem's General Prologue may or may not include links referring to that larger framework. The structure I suggest below for Sl3, for example, suggests that some or all of the “Words of the Host” linking Sir Thopas and Melibee could have been present in this MS. While another Chaucer text, the House of Fame, occurs as a fragment in the first part of Pp, the structure postulated below could have


Page 98
accommodated all of the poem (Edwards' collation, detailed below, notes telegraphically that the MS “lacks 3 leaves containing end of House of Fame” [p. xxiii], but the structure he proposes does not indicate how these might be included). Other texts in Pp may or may not have been fragmentary; only a clear sense of the structures available for those texts will provide us with an understanding of whether they could have been complete in their original form.

Beyond the significance of these MSS as conveyors of important medieval texts is their importance for the evidence they provide about the processes of fifteenth-century book production. Edwards raises the issue of “booklet construction”[3] for the first part of Pp and concludes that:

[Pp] does not show many indications of booklet construction on the—admittedly inconclusive—evidence of quiring. Insofar as the quire divisions are determinable, the scribes do not appear to have been constrained to define texts around them. It is, however, quite likely, given the variegated contents of the first part of Pepys, that it does reflect the collocations of booklet exemplars at some stage prior to the transcription of the exemplar of Pepys itself.

(p. xxviii)

The revised collation I offer below for Pp indicates that the first part of the MS does provide clear evidence of booklet structure for just that part of the MS where one might expect to find it: the Lydgate and Chaucer texts. This in turn clarifies the picture of how the exemplars circulated: together, assembled as a booklet.

The expectations generated by the processes of paper manufacture in the fifteenth century are key to determining the structures of these books. The paper stocks in these MSS have not been well described and have in some instances been misidentified, even with respect to their generic morphology. Little, if any, regard has been paid to the principle of symmetry as it applies to the distribution of watermarks in gatherings, whether it be in folio construction (2°), quarto (4°), or some combination of these two formats (the only formats employed in these MSS). One expects, in fifteenth-century paper, to find watermarks in the left or right half of the full sheet, so that when folded 2°, one half of the sheet will bear the impression of the wire design of the watermark and the other will be blank. In 4° format, the portions of the watermark not obscured by the binding will appear at the spine-edge (i.e., in the gutter) and, of the four folios constructed from the full sheet, two will share the watermark (usually a top half and a lower half) and two will be blank. These facts create certain expectations about the structure of a gathering and, together with other facts about the manufacture of paper and its use in the production of books—whether print or manuscript—provide the bibliographer and cordicologist with invaluable tools in constructing a description of the structure of a given book.


Page 99


British Library MS Arundel 140 contains:

  • Part 1
    • 1. “The Legend of Ipotis” (IMEV[4] 220): fols. 1r-5r
    • 2. Mandeville's Travels (B text: IPMEP 233): fols. 5v-41r
    • 3. Prick of Conscience (East Midlands recension; IMEV 3429): fols. 41v-146v
    • 4. Speculum Gy de Warwyke (also titled Speculum Mundi and Speculum vtile istius mundi; IMEV 1101): fols. 147r-151v
    • 5. Seuyn Sages of Rome (A text; IMEV 3187): fols. 152r-165r
  • Part 2
    • 6. Chaucer's Melibee (VII 967-1777/B2 *2157-*2967; missing VII 1459-1562/B2 *2650-*2752)[5]: fols. 166r- 181r

The volume consists of two structural sections, the final item having been produced independently of the rest of the MS.[6]

Manly-Rickert (1:52) describe three paper stocks: “Mont (2 or 3 varieties) and Basilique” in the first section, and “Ancre, Briquet No. 360 (1459) is perhaps nearest” in the Melibee. Seymour (1966, p. 184) concurs with Manly-Rickert's assessment of the paper stock in Melibee (folded 2°), as do I, but he suggests the first two watermarks are “unidentified, possibly ancre (ff. 1-27) and aile (ff. 28-165).” Manly-Rickert are correct here in their identification of the first watermark as a “Mount surmounted by a Cross” (“Mont”), although it is the same type throughout, occurring in a 2° format in fols. 1-20, 23-27, and 149-165. The second watermark is a Unicorn's Head in Profile ( Tête de Licorne, cf. Briquet 15803 [1406-8]) folded 2°, occurring in fols. 21.22, and 28-148.

Manly-Rickert (1:52) believe a collation of the MS to be “[i]mpossible; all leaves now mounted on modern paper. Two folios missing between 177 and 178 (B 2650-2752).” Seymour (1966, p. 184) provides the following collation for the first section:

5+112 (wants 1, 12), 2-412, 512 (wants 1, 2), 612, 712 (wants 4), 812 (wants 5, 6), 912, 1012 (wants 1, 10), 1110, 1212, 1312 (6, 7 fragments), 1412 (wants 4-8), 1512 (wants 12; two fragments bound before f. 165 do not belong to this quire).

Of the section containing Melibee, Seymour's assessment in 1966 (p. 184) was that the structure was “unknown,” but in his most recent description he states: “Perhaps 112 (lacks 1), 28 (lacks 1-2, 7- 8)” (1995, p. 137). I propose the following collation[7]:


Page 100

Part 1: ॠ1-5 1 12 (-1, 12) 2-4 12 5 12 (-1, 2) 6 12 7 12 (-4) 8 12 (-5, 6) 9 12 10 12 (-1.12, 2.11) 11-13 12 14 12 (-4.9, 5.8, 6.7) 15 12 (-12; 10+१2). Part 2: 1 12 2 8 (-1.8, 2.7)

This collation differs from Seymour's analysis of Part 1 of the MS with respect to his Quires 10 and 11. Quire 10 was originally a gathering of twelve leaves, folded 2°, now lacking the outer two bifolia, and Quire 11 is a gathering of twelve.

(The following structural representations generally consist of four columns. The presentation is intended to suggest the symmetry of the gatherings' physical structures: the tops of the two center columns correspond to the inner and outer folios of the gathering and the folios at the bottom of these two columns correspond to the center bifolium. Commentary on material in the first half of a gathering is placed in the leftmost column, and commentary on material in the latter half of a gathering is placed in the rightmost column. The abbreviation “WM” denotes the half of a bifolium bearing the watermark; a period between folio numbers signifies conjugate structures [inferred where they are no longer physically joined]; the designation “OUT” is used to signify folios that can inferentially be construed as lost; “CW” abbreviates “catchword[s].” The quire numbers are inferred and thus italicized. Hereafter, “Q” abbreviates “Quire”; “Qq” abbreviates “Quires.”)

Q10 12 (-1.12, 2.11):8

OUT  10 1.10 12   OUT 
OUT  10 2.1011   OUT 
107.114  WM 
WM  108.113 
WM  109.112  WM  110.111 

Q11 12:

WM  115.126  CW on fol. 126v  
WM  116.125 
WM  117.124 
118.123  WM 
119.122  WM 
WM  120.121 

The Speculum Gy de Warwyke ends on fol. 151v, at line 892 (“Do nouʒtaʒeyne his lordes wyll”), 142 lines shy of its usual conclusion. One folio (144)[8]


Page 101
could have accommodated these lines (double column format, ca. 45 lines per column). Seymour's collation of Q 14 suggests that 14 4-8 are wanting; there are, however, only six folios surviving from the original gathering (fols. 149-154, with watermarks on fols. 151, 153, and 154; 154v has a catchword), and the distribution of watermarks suggests, rather, that fols. 14 4-9 are missing. If the first of the missing folios held the end of Speculum Gy de Warwyke, then the remaining (now-missing) five folios could have held the lost beginning of the Seuen Sages (figuring ca. 180 lines per folio in a double-column format, with 40-51 lines per column; the Ar text begins at what corresponds to line 951 in the text found in Egerton MS 1995).[9] The missing final folio of Q 15 would have been sufficient to hold the lines lacking at the end of Seuen Sages.

Q14 12 (-4.9, 5.8, 6.7):

149.145  WM; CW on fol. 154v  
150.153  WM 
WM  151.152 
OUT  14 4.14 9   OUT 
14 5.14 8   OUT 
OUT  14 6 14 7   OUT 

Q15 12 (-12; 10+χ2):[10]

155.15 12   OUT 
WM  156.165 
157.164  WM (+2 non-integral fragments) 
WM  158.163 
159.162  WM 
160.161  WM 

The structure of Part 2 is perhaps less problematic than previous analyses have contended. Sixteen folios survive, two are missing between fols. 177 and 178 (containing lines VII 1459-1562/B2 *2650-*2752), and Melibee lacks 111 lines at the end, which could have fit on two additional folios (this section of the MS is in a single-column, prose format, with ca. 35 lines per page). The first twelve folios contain a symmetrical distribution of the Anchor watermark and so can reasonably be inferred to be a gathering:

Arundel 140, Part2:
Qr13 [Seymour's 112 (lacks 1)"][11]


Page 102

WM  166.167 
167.176  WM 
168.175  WM 
169.174  WM 
170.173  WM 
171.172  WM 

If the first two folios of the following quire are missing, as well as the final two (needed for the last 111 lines of text), the original gathering must have consisted of at least eight folios. The surviving four folios are 178-181.

Q2 8 (-1.8, 2.7) [Seymour's “28 (lacks 1-2, 7-8)”]:

OUT  2 1.2 8   OUT 
OUT  2 2.2 7   OUT 
178.181  WM 
WM  179.180 


British Library MS Harley 2382 contains:

  • 1. Lydgate's Life of Our Lady (IMEV 2574): fols. 1r-74v
  • 2. The Assumption of Our Lady (IMEV 2165:) fols. 75r-86r
  • 3. “Oracio ad Sancta Mariam,” extracted from the Speculum Christiani (IMEV 2119; Patterson 1971): fols. 86v-87r
  • 4. “The Testament of Dan John Lydgate” (IMEV 2464): fols. 87v-96v; 108r-v; 128v-129v [The text is spread out, exploiting blank leaves left from the original process of copying. References to the discontinuous sections occur on fols. 96v and 108v.]
  • 5. Prioress's Tale: fols. 97r-100r
  • 6. Second Nun's Tale: fols. 100v-108r
  • 7. “Passio Sancti Erasmi” (IMEV 173): fols. 109r-111r
  • 8. “The Long Charter of Christ” (B text; IMEV 4154). fols. 111v-118r
  • 9. The Child of Bristowe (IMEV 1157): fols. 118v-127r
  • 10. Latin quatrain (“O tu puricia...”): fol. 127r
  • 11. “Of al þe merueile of merlyn how he makys his mone” (SIMEV 2613.5): fols. 127v-128r

The paper stocks in the MS are all folded 4o, distributed as follows:

  • 1. Cart (Char, near Briquet 3547 [1467]): Qq 1-3, 5, part of 6: 52.61, 55-58, 62
  • 2. Crown (Couronne, near Briquet 4646 [1473] and 4645 [1459-69]): Qq 4, part of 6: 53.60, 54.59
  • 3. Cart (Char, a pair, somewhat like Briquet 3544 [1434-79]; see also the addendum to no. 3544 in Briquet 1968 (The “Jubilee” edition), p. *65: Otranto, 1458) but not the same: Qq, 7, 8, 10, 11
  • 4. Bull's head ( Tête de Boeuf, a pair of twins, near Briquet 14970 [1454-61]; cf. Piccard, Ochsenkopf VII.210-222 [1469-75]), with one of the twins having a “droopy” right horn): Qq 9, 12

The quires are structured as follows[12]:

1 12 (-1, 2, 3, 4, 5, 6) 2 12 (-7, 8, 9) 3-5 12 6 12 (-1) 7- 8 12 9 12 (-2, 3) 10-11 12 124 13 6 (-6)


Page 103

The missing final folio in Q 13 was likely a blank. Q 12 consists of two unmarked folios (121.124) and the upper and lower halves of the second Bull's head (with symmetrical horns) (122.123). Q 13 has two unmarked folios, followed by the upper and lower halves of the first Bull's head (droopy horn), and another unmarked folio:

Q 12 4; 4°:

121.124  CW on verson 
WM, upper half: reco=mold side  122.123  WM, lower half; verso=mold side 

Q 13 6 (-6); 4°:

125.13 6   OUT 
WM, upper half; verso=mold side  127.128  WM, lower half; recto=mold side 

This differs from Manly-Rickert's analysis of these final nine folios (1:246), which they believed were all part of an original gathering of ten (their Q 12). There is a catchword on fol. 124v, which I construe as the final folio of Q 12. Seymour (1995, p. 143), who likewise combines the folios distributed in my Qq 12-13, suggests a final gathering of “1212 (wants 10-12).” The distribution of watermarks militates against this. As Manly-Rickert, Seymour, and Keiser (private communication) all argue, the MS consists of three structural sections, with the scribe subsequently exploiting blank space at the end of them for added material. As the sequence of paper stocks indicates, the scribe combined two stocks to produce fols. 1-62 and used a third supply to finish Life of our Lady (through fol. 74, end of Q 7) and The Assumption of Our Lady (through fol. 86, end of Q 8). He then continued using the same supply of paper to copy Prioress's Tale, Second Nun's Tale, “Passio Sancti Erasmi,” “The Long Charter of Christ,” and the beginning of Child of Bristow (through fol. 120, end of Q 11). A fourth paper stock was required to finish Child of Bristow, the scribe for some reason making Q 12 shorter than necessary and overcompensating in his construction of Q 13. Utilizing the blank space on fol. 86v, the scribe then added a gathering between fols. 86 and 97 (Q 9), copied the “Oracio,” and began the “Testament,” running out of space and spreading the remainder out over other blanks left from original stint of copying. These blanks point to three “booklet” structures, consisting of Qq 1-8 + 9; Q 10; and Qq 11- 13. [13]


Magdalene College, Cambridge MS Pepys 2006, is paginated, and contains:

  • Part 1:
    • 1. Lydgate, The Complaint of the Black Knight (IMEV 1507): pp. 1- 17


      Page 104
    • 2. Lydgate, Temple of Glas (IMEV 851): pp. 17-52
    • 3. Chaucer, Legend of Good Women, ends at l. 1377 (IMEV 100): pp. 53-87
    • 4. Chaucer, An ABC, ends at line 59, with spurious added line following (IMEV 239): pp. 88-90
    • 5. Chaucer, House of Fame, ends at line 1843 (IMEV 991): pp. 91- 114
    • 6. Chaucer, Complaint of Mars (IMEV 913): pp. 115-122
    • 7. Chaucer, Complaint of Venus (IMEV 3542): pp. 122-124
    • 8. Chaucer, Fortune (IMEV 3661): pp. 124-126
    • 9. Chaucer, Parliament of Fowls (IMEV 3412): pp. 127-142
    • 10. The Three Kings of Cologne (IPMEP 290): pp. 143-189; 190 blank, with later additions[14]
    • 11. Lydgate, Serpent of Division (IPMEP 835): pp. 191-209
    • 12. Lydgate, Envoy to Serpent of Division (IMEV 3625): p. 209
    • 13. Benedict Burgh, Parvus Cato, lines 1-50 (IMEV 3955): pp. 211- 212
    • 14. Benedict Burgh, Cato Major, ends at line 360 (IMEV 854): pp. 213-224.
  • Part 2:
    • 15. Chaucer, Melibee: pp. 225-275
    • 16. Chaucer, Parson's Tale: pp. 276-376
    • 17. Chaucer, Retraction: p. 377
    • 18. Chaucer, Complaint of Mars, lines 1-28; 57-84; 29-56 (IMEV 913): pp. 378-380
    • 19. Chaucer, Complaint of Venus, begins line 45 (IMEV 3542): pp. 381-382
    • 20. Chaucer, Anelida and Arcite, lines 211-289; 299-311 (IMEV 3670): pp. 382-384
    • 21. Chaucer, Fortune, lines 78-79 (IMEV 3661): p. 385
    • 22. Chaucer, Lenvoy de Scogan (IMEV 3747): pp. 385-386
    • 23. Chaucer, An ABC, ends at line 59, with spurious added line following (IMEV 239): pp. 386-388
    • 24. Chaucer, “The Complaint of Chaucer to His Purse” (IMEV 3787): pp. 388-389
    • 25. Chaucer, Truth (IMEV 809): pp. 389-390
    • 26. Chaucer, Merciless Beauty (IMEV 4282): pp. 390-391
    • 27. A seventeenth-century collation of the MS's contents: pp. 392-394

Previous descriptions of the watermarks in Pp reflect a naiveté about the structure and appearance of watermarks in a 4o format (see also Keiser). Manly-Rickert attempt only an identification of a “Tête de Boeuf, which predominates in CT,” suggesting that it “bears some resemblance to several in Briquet (cf. 14238, 14239, 14244, dating mainly 1469-87)” (1:406). Edwards (p. xxiii) determines that two of the watermarks— Tête de boeuf —approximate “Briquet 15043” and “Briquet 15204.” For the remainder of the MS (the second part), he observes that “all the watermarks are in the gutters,” provides diagrams of three “ascenders,” and says that “[t]here are possibly additional, unidentifiable watermarks on pp. 239-40 (possibly a tête de boeuf ), 291-2 and 359-60.” McKitterick and Beadle (p. 42) echo Edwards' suggestion about the two Tête de boeuf watermarks “approximately resembling Briquet nos 15043 and 15204,” and further note “a croix (cf. Briquet nos 5590 and 5706), a five-pronged ascender, a forked ascender, and an arrow-headed ascender (for page references see Edwards, p. xxiii).” Both descriptions are


Page 105
misleading, especially the association of what is the upper or lower portion of a 4° watermark with a Briquet representation of a full watermark.

The paper stocks in Pp occur as follows:

  • 1. Mount with jagged bottom edge surmounted by a Cross (Monts, unidentified), 4° (a pair of twins, one with a crossed ascender, and one on which the ascender appears to lack the horizontal elements of the cross): pp. 1-44; 73-189
  • 2. Bull's head ( Tête de boeuf ), 2°, close to Briquet 15068 (1462): pp. 45-70
  • 3. Bull's head ( Tête de boeuf ), 2°, nearest Briquet 15204-15206 (1440-51), but it is definitely not any of those; the watermark in Pp has two ears: pp. 193-226
  • 4. Bull's head ( Tête de boeuf ), 4°, unidentified, with only the nostrils and (sometimes) the lower portion of the nose of the lower half, and a five-pronged ascender and (sometimes) the tips of horns, approximately 2.7 cm apart, of the upper half visible; similar to the second mark listed above, but with attendant chainlines 4.2 cm apart as compared with 3.8 cm apart in number two). Possibly similar to Picard Ochsenkopf VII.852; VII.564 (chainline spacing and alignment are closest to the latter) (1440-55); VII.435; VII.294 (1466-70); it is also very similar to a mark found in British Library Canterbury Tales MS Royal 17.D.XV, including the bent tip of the ascender (similar to Briquet, Tête de Boeuf, 15054 [1441-1445], with attendant chainlines 3.8 cm apart): pp. 228-290
  • 5. Cart (Char), 4°, near Briquet 3544 (1433, with variants to 1473): pp. 291-390

The structure of this manuscript—really two separate manuscripts bound together—remains indeterminate in some sections. McKitterick and Beadle (pp. 42-43, with the suggested collation attributed to McKitterick) propose (“[h]aving regard to the sewing, stubs, and disposition of the texts”) the following:

I4 (pp. 1-8), II8 (pp. 9-24), III6 (pp. 25-36), IV8 (pp. 37-52), V16 (pp. 53-84), VI14 (pp. 85-112), VII16 (wants 16; the remaining stub is pasted to 15) (pp. 113-142), VIII24 (pp. 143-190), IX6 (pp. 191-202), X6 (4 and 5 pasted together) (pp. 203-212), XI6 (pp. 213-224), XII8 (pp. 225-240), XIII16 (pp. 241- 272), XIV4 (pp. 273-280), XV8 (pp. 282-296), XVI8 (pp. 297-312), XVII4 (pp. 313-320), XVIII8 (pp. 321-336), XIX4 (pp. 337-344), XX12 (pp. 345-358) [sic [15]], XXI14 (pp. 369-376), XXII8 (pp. 377-392, but pp. 377-378 and 391-392 consist of two leaves pasted together).

Edwards (pp. xxiii-xxiv) offers these “very tentative suggestions for the first part of the manuscript” (I omit the prose description of the contents and restructure the collation for more ready comparison with that of McKitterick and Beadle):

18 (pp. 1-16), 214 (pp. 17-44), 38 (pp. 45-60), 48 (pp. 61-74; “lacking 2 leaves after p. 70, one of which is supplied in a later hand”), 58 (pp. 75-90), 612 (pp. 91-112; “possibly a missing [?blank] leaf before p. 91”), 78 (pp. 113-122; “lacks 3 leaves containing end of House of Fame”), 810 (pp. 123-142), 9- 118 (pp. 143-190), 1212 (pp. 191-210; “+ 1 leaf pasted to p. 190 + 1 leaf pasted to p. 210”), 138 (pp. 211-224; “lacks 1 leaf after p. 224”)

Second part: 18 (pp. 225-240), 28 (pp. 241-256), 38 (pp. 257-272), 48 (pp. 273-288), 58 (pp. 289-304), 68 (pp. 305-320), 78 (pp. 321-336), 88 (pp. 337-352), 98 (pp. 353-368),


Page 106
108 (pp. 369-380; 2 leaves glued together after p. 377 and one leaf missing after p. 380), 118 (pp. 381-392; 2 leaves missing after p. 384)

Neither of these collations makes specific reference to the distribution of watermarks (i.e. the symmetrical structure resulting from folding the paper 4° and 2°). While there are uncertainties, as must be the case with any hypothetical reconstruction, I believe the following best represents the structure for the first part of the MS:

Part 1: 1 24 (-1, 2) 2 14 (-14) 3 22 (-1, ±2) 4 20 (-2.19, 3.18, 20) 5 24 6 10 7 8 (-8)

Part 2: 1 16 2 16 (2 16 + χ1) 3 14 4 18 5 6 6(?)

This collation is motivated by the physical evidence and the relationship of that evidence to the distribution of texts. The sewing is through the sides, rather than through the center of bifolia, and so does not indicate the middle of gatherings (“oversewing,” rather than the more usual “flexible sewing”).[16] The first part of the MS consists of two discernible sections signaled by changes in paper stocks, textual divisions (with a blank verso at the end of the first section), and breaks between gatherings. These sections consist of: pp. 1-189 (190 is blank), with runs of two paper stocks; and pp. 191-224 (a single paper stock, distinct from those in the first section). Further, in the second section, each of the texts occupies a single gathering in the structure I propose (although the Cato text may have continued into another gathering). The second part of the MS does not contain any correspondences of text, paper stock, and quiring suggestive of booklet construction. I detail this evidence more fully in diagrammatic format below.

Pepys 2006, Part 1:

Q 1 24 (-1, 2), pp. 1-44; Mount, folded 4°:

OUT  1 1.43/44  WM, upper half; CW on 44v  
OUT  1 2.41/42 
WM, upper half  3/4.37/38  WM, lower half[17]  
WM, lower half  5/6.35/36  WM, upper half 
WM, lower half  11/12.29/30  WM, upper half 
WM, upper half  13/14.27/28  WM, lower half 
WM, lower half  17/18.23/24  WM, upper half 


Page 107

Q 2 14 (-14), pp. 45-70; Bull's head, folded 2°:

WM  45/46.2 14   OUT (blank?) 
WM  47/48.69/70  (“Legend of Cleopatra” ends on p. 70) 
49/50.67/68  WM 
(Temple of Glas ends on p. 52)  51/52.65/66  WM 
(LGW begins on p. 53) WM  53/54.63/64 
WM  55/56.61/62 
57/58.59/60  WM 

Q 3 22 (-1 ±2), pp. 71-112; Mount, folded 4°:

OUT  3 1.111/112  CW on 112v (Guard obscures gutter) 
(LGW ll. 706-776 are missing)
(LGW, ll. 777-845, added in different hand) 
WM, lower half  73/74.107/108  WM, upper half 
WM, lower half  77/78.103/104  WM, upper half 
WM, upper half  79/80.101/102  WM, lower half 
WM, upper half  83/84.97/98  WM, lower half 
WM, lower half  89/90.91/92  WM, upper half 

Q 4 20 (-2.19, 3.18, 20), pp. 113-142; Mount, folded 4°:

(HF breaks off at l. 1843, at the foot of 114)  113/114.4 20   OUT (blank?) 
(HF ll. 1843-2003?) OUT  4 2 4 19   OUT (blank?) 
(HF ll. 2004-2158?) OUT  43.418   OUT (PF, ll. 668-669?) 
115/116.141.142  (guard; PF breaks off at l. 667 
WM, lower half  117/118.139/140  WM, upper half; guard 
WM, upper half  121/122.135/136  WM, lower half 
(PF begins on p. 127)  127/128.129/130 

Q 5 24, pp. 143-190; Mount, folded 4°:

(Three Kings begins)  143/144.189/190  (two pages glued together, both 4°)[18]  
WM, lower half  145/146.187/188  WM, upper half 
WM, lower half  149/150.183/184  WM, upper half 
WM, lower half  161/162.171/172  WM, upper half 
WM, upper half  163/164.169/170  WM, lower half 


Page 108

Q 6 10, pp. 191-210; Bull's head (near Briquet 15204 & 15206), folded 2°:

(Serpent of Division begins)  191/192.209/210  WM (two sheets glued together) 
193/194.207/208  WM 
195/196.205/506  WM 
WM  197/198.203/204 
WM  199/200.201/202 

Q 7 8 (-8), pp. 211-224; Bull's head (near Briquet 15204 & 15206), folded 2o:

(Cato begins)  211/212.7 8   OUT (stub; end of Cato Major missing, 360-? 
213/214.223/224  WM 
WM  215/216.221/222 
217/218.219/220  WM 

The first section of Pp, containing texts by Lydgate and Chaucer (pp. 1-142, with two blank leaves left in the original structure at the end of my Q 4), appears to have employed examplars that were structured as booklets and to have emulated that structure, a fact obscured by previous attempts to collate the MS. In the remainder of Pepys Part 1, textual boundaries coincide with structural boundaries; previous collations have also failed to capture this important point.

Pepys 2006, Part 2:

In pp. 225-290, the lower portions of the 4° watermark can be very difficult to detect because they are often deeply buried in the gutters, often showing only parts of the nostrils. Between my second and third examinations of the MS, slightly over a year apart, I constructed a collation that originally had to rely on a series of hypotheses about where the lower portions of the mark should be expected. During my third examination, I was able to confirm the presence of those marks where I predicted they should be, but they are easily overlooked if one does not know what to look for and where to look.

Q1 16, pp. 225-256; “Bull's head,” folded 4°:

WM, upper half  227/228.253/254  WM, lower half 
WM, upper half  229/230.251/252  WM, lower half 
WM, lower half  235/236.245/246  WM, upper half 
WM, lower half  239/240.241/242 

Q 2 16 (2 16 + χ1), pp. 257-290; “Bull's head,” folded 4°:

WM, upper half  257/258.287/288  WM, lower half 


Page 109

WM, upper half  265/266.279/280  WM, lower half 
WM, upper half  271/272.273/274  WM, lower half 

Q 3 14, pp. 291-306; “Cart” (near Briquet 3544), folded 4°:

WM, lower half  291/292.317/318  WM, upper half 
WM, upper half  293/294.315/316  WM, lower half 
WM, upper half  297/298.311/312  WM, lower half 
WM, upper half  301/302.307/308  WM, lower half 

Q 4 18, pp. 310-338; “Cart” (near Briquet 3544), folded 4°:

WM, lower half  321/322.351/352  WM, upper half 
WM, upper half  327/328.345/346  WM, lower half 
WM, lower half  329/330.343/344  WM, upper half 
WM, upper half  333/334.339/340  WM, lower half 

Q 5 6, pp. 355-366; “Cart” (near Briquet 3544), folded 4°:

WM, lower half  355/356.365/366  WM, upper half 
WM, upper half  359/360.361/362  WM, lower half 

The following section has suffered several losses of text. The text of the Complaint of Mars (ll. 1-28 [p. 378], 57-84 [p. 379], and 29-56 [p. 380]) was originally copied consecutively. The folio containing pp. 379/380 has been reversed, with the lower half of the “Cart” watermark appearing at the fore-edge. Thus the entire text might once have been present and the final 214 lines of text subsequently lost, along with the first 45 lines of the Complaint of Venus. At 28 lines per page, allowing for an explicit and incipit, this would require 4.6 folios. The final 43 lines of Anelida & Arcite and the first 77 lines of Fortune are also lost; at ca. 32 lines per page in this section, another two folios can be postulated as missing. There is, however, a “vacat” in the gutter of p. 384 that appears to be written in the scribe's informal hand, and since the last line of Fortune provides a rhyme for the last line of text on p. 384 (“soueryn:atteyn”), it is entirely possible that this was the scribe's solution to a lacuna in the copy text. It is also possible that if more of the final part of the MS were extant, the postulated gathering 56 might be seen to associate with a larger structure.

Q 6?; “Cart” (near Briquet 3544), folded 4°:

WM, lower half  369/370 


Page 110

WM, lower half (Parson's Tale ends p. 376)  375/376 
(Two leaves pasted together: Complaint of Mars, pp. 378-80  377/378 
leaf reversed; WM, lower half, at fore-edge (Seymour (1995, p. 134) suggests this is the second leaf of a gathering of 12, with 3, 6, 7, 12 missing)  379/380 
(Complaint of Venus, pp. 381-2)   381/382 
WM, lower half (“Vacat” in the lower gutter of p. 384)  383/384 
(Final 43 lines of Anelida & Arcite lost)  OUT? 
WM, upper half  385/386 
WM, lower half (fly leaf glued to verso)  391 


British Library MS Sloane 1009 contains:

  • 1. “Penaunce is the seconde medycyne of syn[n]e after Noees flode” (Jolliffe, p. 63: A.4.ii; Memoriale Credencium, ed. J. H. L. Kengen; IPMEP 448): fols. 2r-16v [19]
  • 2. “[w]E knoweth well by comyn exp[er]ience...” (Reclusorium Anime [extract]; Jolliffe, p. 114; I.41; Hartung, vol. 7, p. 2535 [150]): fols. 17r-25v
  • 3. “[a] grete Clerke Richard of Seynt Victores seyeþ in a boke| of contemplation þt ther be þre worthynges of eu[er]y c[ri]sten| soule Thought Thenkyng & contemplacou[n]” ( De contemplacione, Ed. Hayes[20]; Jolliffe, p. 127: M.5/p. 138: 0.15; IPMEP 5): fols. 25v-26v
  • 4. “Nonnulla de hominum natura prava” (Latin): fols. 26v- 27r
  • 5. “[t]her be vj thynges þt will bring a mannys soule to hevyn” (“Of the Six Religious Duties”: Jolliffe, p. 114: I.36; Hartung, vol. 7, p. 2318 [141]): fols. 27r-28r)
  • 6. Chaucer's Melibee (ending at VII 1846/B2 *3036): fols. 29r-48r
  • 7. “Rubrica ¶De vita & honestate Clericor[um]”:[21] fols. 49r-57r
  • 8. “Computacio Danielis Prophete” (Latin): fols. 58r- 58v


    Page 111
  • 9. “Apes videre vel cap[er]e lucru[m] sig[nifica]t” (Somniale, Latin: fols. 59r-61v

The rest of the manuscript consists of collection of fiction and nonfiction texts written by Gideon Bonnivert of Oxnead Hall, Norfolk in the second half of the seventeenth century (fols. 62-437), and is not relevant to the present discussion. (See Manly-Rickert 1:515 and the Catalogus Librorum Manuscriptorum Bibliothecæ Sloanianæ [British Library, vol. 1, pp. 202- 204]).

The paper stocks are distributed as follows (cf. Manly-Rickert 1:515):

  • 1. Unicorn, a pair of twins, one 9.3 cm long from tip of the raised hind foot to tip of horn, aligned parallel to chainlines 3.9 cm apart; the second mark measures 9.9 cm from tip of the raised hind foot to tip of horn, also with attendant chainlines 3.9 cm apart. Very close to Briquet Licorne Simple 9997 (1477-78), but in both marks the front leg descends below the chainline (this watermark also occurs in British Library Canterbury Tales MS Royal 17.D.XV): fols. 1-48
  • 2. Unicorn, a pair of twins, the first (e.g., fol. 51) measuring 9.5 cm from tail to tip of horn, aligned parallel to chainlines 3.6-3.7 cm apart; the second (e.g. fol. 53) measuring 10 cm from tail to tip of horn, aligned parallel to chains between chainlines 3.9| 3.5 cm apart; the second has a fatter and squatter appearance (cf. Briquet Licorne Simple 10024 & 10026 (this watermark also occurs in British Library Canterbury Tales MS Royal 17.D.XV and is the same paper stock used by Caxton in his first edition of the Canterbury Tales: (STC 5082 [1477]): fols. 49-57
  • 3. Circle, near Briquet Cercle 3107 (Augsburg, 1455; Magdebourg, 1458), 4°. Since the mark is symmetrical, it is impossible to determine whether the top half or bottom half appears on a given folio: fols. 58-61

Seymour (1995, p. 144) based on “contents and watermarks,” suggests “18 (wants 1), 28 // 312 4-510 // 610 (wants 1) // four leaves of uncertain structure, ff. 58-61. One leaf lost after quire 5.” The watermark distribution, however, does not support this hypothesis. I propose instead the following:

1(?) 2 12 3 12 4 24 (- 1, 22, 23, 24) 5(?) 6 4

The reasons for proposing this collation are detailed below:


Page 112


Page 113

The four new collations provided here derive from a consideration of the kinds of physical structures that can result from the folding of paper into gatherings. This and other kinds of physical evidence (scribal stints, textual lacunae, sewing, changes of paper stock, catchwords), taken together with the layout and distribution of texts in the MSS, argue for a limited number of possible structures. From among those structures, I believe the ones I propose here are the most probable and most reasonably account for the evidence provided by the testimony of these fifteenth-century witnesses.


Benson, Larry D., ed. The Riverside Chaucer. 3rd ed. [Based on The Works of Geoffrey Chaucer. Ed. F. N. Robinson]. Boston: Houghton Mifflin Co., 1987.

Briquet, C. M. Les Filigranes: Dictionnaire Historique des Marques du Papier Dés Leur Apparition vers 1282 jusqu'en 1600. Facsimile of the 1907 ed. with supplementary material contributed by a number of scholars. Ed. Allan Stevenson. 4 vols. Amsterdam: Paper Publications Soc., 1968.

Brown, Carleton, and Rossell Hope Robbins, eds. The Index of Middle English Verse. New York: Columbia UP, 1943. [IMEV]

Brunner, Karl. The Seven Sages of Rome (Southern Version). EETS o.s. 191. London: Oxford Univ. Press for the Early English Text Society, 1933.

Edwards, A. S. G., introd. Manuscript Pepys 2006: a Facsimile, Magdalene College, Cambridge. The Facsimile Series of the Works of Geoffrey Chaucer, vol. 6. Norman, OK: Pilgrim Books; Woodbridge and Suffolk: Boydell and Brewer, 1985.

Hanna, Ralph III. “Booklets in Medieval Manuscripts: Further Considerations.” Studies in Bibliography 39 (1986): 100-111.

Hartung, Albert E. A Manual of the Writings in Middle English: 1050- 1500. Based upon a Manual of the Writings in Middle English 1050-1400 by John Edwin Wells, New Haven, 1916 and Supplements 1-9, 1919-1951. Volume 3. New Haven: The Connecticut Academy of Arts and Sciences, 1972.

Hayes, Stephen E. “Of Three Workings in Man's Soul: A Middle English Prose Meditation on the Annunciation.” In Vox Mystica: Essays on Medieval Mysticism in Honor of Professor Valerie M. Lagorio. Ed. Anne Clark Bartlett, with Thomas H. Bestul, Janet Goebel, and William F. Pollard. Cambridge: D. S. Brewer, 1995. 177-199.

James, Montague Rhodes. A Descriptive Catalogue of the Library of Samuel Pepys. Part III: Medieval Manuscripts. London: Sidgwick & Jackson, Ltd., 1923.

Jolliffe, P. S. A Check-List of Middle English Prose Writings of Spiritual Guidance. Subsidia Mediaevalia II. Toronto: Pontifical Institute of Mediaeval Studies, 1974.

Keiser, George R. Review of A. S. G. Edwards, introd. Manuscript Pepys 2006: A Facsimile Magdalene College, Cambridge. Studies in the Age of Chaucer 9 (1987): 212-215.

Kengen, J. H. L. Memoriale Credencium: A late Middle English Manual of Theology for Lay People edited from Bodley MS Tanner 201. 1979.

Lewis, Robert E., and Angus McIntosh. A Descriptive Guide to the Manuscripts of the Prick of Conscience. Oxford: Society for the Study of Mediæval Languages and Literatures 1982.

Lewis, R. E., N. F. Blake, and A. S. G. Edwards. Index of Printed Middle English Prose. New York & London: Garland Publishing, Inc., 1985. [IPMEP]

McKitterick, Rosamond, and Richard Beadle. Catalogue of the Pepys Library at Magdalen College Cambridge, vol. V.i. D. S. Brewer, 1992. 39-44.

Manly, John M., and Edith Rickert, eds. The Text of the Canterbury Tales: Studied on the Basis of All Known Manuscripts. 8 vols. Chicago: Univ. of Chicago Press, 1940.

Patterson, Frank Allen, ed. “Oracio ad Sanctam Mariam.” In The Middle English Penitential


Page 114
Lyric. Columbia Univ. Studies in English, Series Z, vol. 14. New York: Columbia Univ. Press, 1971. 139-141.

Piccard, Gerhard. Ochsenkopf Wasserzeichen. Findbuch II.1-2 of Die Wasserzeichen Piccard im Haupstaatsarchiv Stuttgart. Stuttgart: Verlag W. Kohlhammer, 1966.

Robbins, Rossell Hope, and John L. Cutler, eds. Supplement to the Index of Middle English Verse. Lexington: Univ. of Kentucky Press, 1965. [SIMEV]

Robinson, Pamela R. “The `Booklet,' A Self-Contained Unit in Composite Manuscripts.” Codicologica 3 (1980): 46-69.

Seymour, Michael C. “The English Manuscripts of Mandeville's Travels.” Edinburgh Bibliographical Society Transactions 4 (1966): 167-210.

—. A Catalogue of Chaucer Manuscripts: Volume 1. Works before the Canterbury Tales. Aldershot, England & Brookfield, VT: Scolar, 1995.


I wish to acknowledge the help provided to me by Svato Schutzner, Juris Lidaka, Josephine Tarvers, and Stephen Hayes in response to queries posted to Medtextl and Chaucernet. I am also grateful to Dr. Richard Luckett, Pepys Librarian and Keeper of the Old Library (Magdalene College), and Aude Fitzsimons, Assistant Pepys Librarian, for allowing me to examine Pepys MS 2006 on several occasions, to the British Library and the staff of the Students' Room for making the three manuscripts described in this essay available to me, to Julia Boffey for answering several questions about the identification of some of the texts contained in these MSS, and to George R. Keiser for his suggestions on the description of Harley MS 2382. I have also benefitted from the thorough and detailed comments provided by the editor of this volume.


The sigla provided here in brackets are those used by John M. Manly and Edith Rickert (editors of The Text of the Canterbury Tales, hereafter “Manly-Rickert”) and are in standard use in Chaucer studies.


On the subject of the use of booklets in fifteenth-century book production, see especially Robinson and Hanna. Hanna defines the booklet in this context as “a group of leaves forming at least one quire, but more likely several, and presenting a self-contained group of texts” (100-101).


The abbreviations “IMEV,” “SIMEV,” and “IPMEP” refer to The Index of Middle English Verse, the Supplement to the Index of Middle English Verse, and the Index of Printed Middle English Prose, respectively.


References to the works of Chaucer follow the lineations of the Riverside Chaucer (ed. Benson).


The “independence” of the two parts of the volume are attested to by the differences in paper stocks, formats, and scribal hands and dialects.


There is no reason not to characterize the first five folios as part of a fragmentary gathering (probably a twelve; fols. 2, 4, and 5 bear watermarks), but so as not to make my references here unnecessarily confusing, I will maintain Seymour's quire numbering, referring to these folios as “π1-5.” The two added fragments in Quire 10, which are misbound, are foliated as “165.1” and “165.2.” The second gathering in Part 2 was, given the losses of text at the beginning and end of the gathering, at least a gathering of eight.


Lewis and McIntosh note the following omissions of text: “(1) two leaves between ff. 51 and 52 containing ll. 856-1011; (2) one leaf between ff. 76 and 77 containing ll. 2965- 3034; (3) two leaves between ff. 88 and 89 containing ll. 3976-4140; (4) two leaves between ff. 106 and 107 containing ll. 5674-841; (5) two leaves between ff. 114 and 115 containing ll. 6475-636; (6) two leaves between ff. 144 and 145 containing ll. 9304-467, but they remain as scraps and are bound between ff. 143 and 144” (1982, pp. 57-58).


See Brunner, p. 37. The Ar text of Seuen Sages begins “Hys comaundement þei dide belyue,” and ends “ʒef I wyll late þe wtowte oþe.”


The two fragmentary leaves signified by “χ2” are misbound in Q 15 following fol. 164.


Seymour's suggestion that the first folio is missing in the first gathering of Part 2 (despite there being no loss of text) is an inference based perhaps on the traces of “signatures” on fols. 167v (“c”), 169v (“e”), 170v (“f”), and 171v (“g”?). If these were accurate indications of the original structure (and their placement on the versos is itself odd), then it is conceivable that the first gathering in this section was one of fourteen leaves, with unknown material on the first folio, and with the final folio containing part of the text missing between fols. 177 and 178. The following gathering would still require at least seven folios to complete the text of Melibee.


Fol. 61 is now fol. 88 of British Library MS Sloane 297.


At the foot of fol. 96v, the scribe writes “Quere plus inxijo folio post,” picks up copying at the foot of 108r (following Second Nun's Tale), fills the verso, writes “Quere plus in xxo folio post,” resumes the “Testament” on fol. 128v, and completes the text of the “Testament” on fol. 129v. I am very grateful to George R. Keiser for suggesting several corrections in the description and analysis of Harley 2382 (personal communication). I accept and follow his analysis of the scribe's progress of copying.


Although there is an explicit on p. 183, the material that follows, through p. 189, is the concluding section of The Three Kings of Cologne (see Edwards, p. xix).


Note that “XX12” should include pp. 345-368.


Sewing, through the side of the pages rather than through the center fold, can be seen, e.g., between pp. 4/5, 16/17, 30/31, 44/45, 68/69, 89/91, 98/99, 130/131, 146/147, 218/ 219, 256/257, 288/289, 304/305, 328/329, 340/341, 358/359, 372/373, 382/383.


The jagged edge of the lower half of the mark can barely be seen. Approximately 6.3 mm of the ascender shows on pp. 3/4. By comparison, on pp. 13/14, approximately 6 mm of the ascender are visible, and the corresponding portion of the conjugate lower half that is visible on pp. 27/28 is approximately 3 mm. The expected portion on pp. 37/38, then, could be predicted to be right at the edge of the gutter, where the edge is indeed just visible.


A blank quarter of a sheet is glued or pasted to what would have been p. 190 and is itself numbered “190.” Whether this was done for structural reinforcement or to obscure what has been written on the back of the folio (originally left blank) is not clear. Since the same procedure has been carried out on p. 210, at the end of the following gathering, and there is also writing on the blank space left there, I suspect the latter possibility.


Kengen: “[The text] corresponds to MS. O [Bodley MS Tanner 201] ff.66r/6-87r/15; 88v/6-89r/4, and 90r/18-97r/18, i.e. the section on Penance, Tribulation, Temptation, the Divine Virtues, the Cardinal Virtues, Meditation, and part of the section on Contemplation, omitting the last degree of contemplation and ending with the seven gifts of the Holy Ghost. Even in these shared sections there are occasional omissions and additions.... There is at least one leaf missing corresponding to MS. O ff.70r/12-71v/18, from `dedly' to `purgatorye'” (10).


Hayes characterizes the text as “an abridged Middle English paraphrase translation of Richard of St. Victor's Benjamin minor” (p. 185).


Not the Speculum sacerdotis (Title 1 of Book III of the Decretals of Gregory IX), as the rubric might suggest, but “a sermon, or perhaps a treatise in the form of sermon, by somebody used to preaching. It is pretty obviously a sermon by a man working in canon law, and probably addressed to other people enamored with that lore. Oxford would not be an unlikely place.... It sounds Wycliffite, very much so. It seems to belong to the genre `University sermons.' In this case, the preacher is taking a rubric as his theme, rather than a verse from the Bible. [It is] an unrelenting attack on the established clergy. The text is continuous, no gaps found between end of one leaf and beginning of another” (Svato Schutzner, e-mail communication 8 February 1997).


The text in Sloane 1009 begins with “Penaunce is the seconde medycyne of sy[n]ne after Noes flode.” While this could be the intended beginning of an extract, it omits an amount of text equivalent to 59 folios in Bodley MS Tanner 201 [= “O” in Kengen]. The watermark structure suggests that this gathering is fragmentary at the beginning and end. See Kengen, p. 10: “There is a least one leaf missing corresponding to MS. O ff.70r/12-71v/18, from `dedly' to `purgatory'.”