University of Virginia Library

John Payne Collier's Great Forgery
by
Giles E. Dawson

A communication in The Athenaeum for 31 January 1852, signed "J. Payne Collier," announced his purchase of a copy of the Second Folio of Shakespeare inscribed "Tho. Perkins, his Booke" and containing thousands of alterations of the text written in a hand "probably not of later date than the Protectorate." This "Perkins Folio," now in the Huntington Library, does indeed contain thousands of alterations — perhaps 25,000 — ranging in size from marks of punctuation to whole lines and pairs of lines.

In 1859 Sir Frederic Madden, Keeper of Manuscripts at the British Museum, observed that the writing in the book was "not a genuine writing, either of the 17th or 18th century." Later evidence found at the Museum indicated that Collier himself had been the fabricator, and Clement Mansfield Ingleby, in A Complete View of the Shakspere Controversy, 1861, set forth all the evidence of Collier's wrongdoing. But, persuasive though the evidence is in its sheer bulk, most of it is at best circumstantial, and the rest is open to suspicion.

Madden, though a first-class paleographer, did not record in any detail his reasons for concluding that the Perkins annotations were written in the nineteenth century. Nor, until now, has anyone else made a thorough paleographical examination of them. A good many years ago I was struck by the close similarity between the writing of the corrections in the Perkins folio and that of eighty-three ballads written in a manuscript in the Folger Library. But though I then saw good reason for believing that Collier had written the ballads, it was not until I tackled the problem recently that I came upon the proof that at first eluded me. That proof, together with a demonstration that the ballads and the Perkins annotations are the products of one hand, is the matter of this paper.

By way of introduction I might reasonably be expected to provide an account of Collier's career, emphasizing the facts concerning his acquisition of the Second Folio, Collier's several descriptions of the

written matter in it, his later dealings with these, the first suspicions about their genuineness, and the evidence of fabrication that was produced within ten years of Collier's first announcement. But since all this is readily available in Sir George Warner's admirable account in The Dictionary of National Biography and all the details may be read in Ingleby's Complete View, I may perhaps be excused for skipping the background matter in the interests of brevity.

I

My first task, then, is to demonstrate that eighty-three ballads written in a manuscript volume in the Folger Library (V.a.339) are the work of J. P. Collier. The manuscript is a commonplace book of 291 leaves measuring 5.3 x 3.6 inches, originally bound in reddish-brown calf in the early seventeenth century and rebound, in 1960, in the Folger bindery. At that time, while the book was in sheets, I collated it and pencilled a signature on the first recto of each sheet. The collation, by these added signatures, is as follows:

The book originally consisted of 300 leaves, of which nine have been removed. There is no indication that any of these extractions are attributable to Collier. While the book was in his possession he numbered the leaves, but made mistakes that resulted in his ending with 282 instead of 291. I later added correct foliation, which can be seen in the reproduction of f. 114 in Plate I, my numbers being smaller in size but greater numerically than Collier's.

Of the date at which Collier acquired the manuscript volume all that can be said with certainty is that it was in his possession in 1843, when, in his first edition of Shakespeare (8 vols. 1842-44), VIII (1843), 568, he notices a variant in a piece from the Passionate Pilgrim copied into the manuscript by its first owner, to the end of which have been added the initials W. S. in a hand and ink that, like initials elsewhere in the volume, I believe to be Collier's. Collier's statements about the date are vague and inconsistent; that he bought the manuscript from Thomas Thorpe, as he several times says, we need not question. After his death in 1883 the volume was No. 16 in the catalogue of the sale of his library by Sotheby, Wilkinson, and Hodge, 7-9 August 1884. The successful bidder, Quaritch, paid £52. In another Sotheby sale, 20-22 April 1899, the manuscript was No. 1002 in a section of "Other

Properties." A statement in the catalogue of this sale informs us that "In a note on a fly-leaf he [Collier] says "This volume, besides other poems, contains about 80 ballads not found elsewhere. I gave Thorpe £25 for it.'" No such fly-leaf was present when the manuscript came to the Folger Library; we may assume that it was the missing A1. In the same year the manuscript appeared as No. 386 in J. Pearson's Catalogue of Shakespeareana, 2 volumes, 1899, and was sold to Marsden J. Parry, of Providence, Rhode Island, after whose death it passed, through A. S. W. Rosenbach's hands, to H. C. Folger.

Who the first owner and first user of the volume was seems impossible to determine. It may have been the Joseph Hall (not the Bishop of Norwich) whose signature is on what is now the only fly-leaf, A2, but this is doubtful, since the appearance of the signature does not suggest that Hall wrote anything else in the book and does suggest that it was written as late perhaps as 1700. All but two of the 582 pages are pretty well filled with written matter. The first writer appears, from a few scattered dates, to have done his work in the decade from 1630 to 1640. He did about fifty percent of all the writing (by page-count; it would be more by word-count because his writing is always small, sometimes minute, always close packed). Starting with a blank book, he planned to classify his matter, allotting a section of the book to each kind — theology and religion, medicine, remedies, and so on — six sections in all. He began each section at the top of a page and from time to time made additions, leaving no space between old and new matter. None of the five sections of written matter fills the block of pages allotted, so that he left blocks of blank leaves, the biggest containing 142 pages. The second man who used the book did his writing (on about forty percent of the pages) during the next decade — if we can judge by the dates scattered through his portion of the writing, the latest of which is 1648. At first he seems to have continued his predecessor's classification but to have given it up when certain of the available spaces became filled. But he always started just where the first left off, not unusually at the top of a page. When the second writer quit, ff. 39v-41, 107v-179, 181v, 228v, 242-245, 276v-277 were entirely blank, while the lower halves of ff. 39, 181, and 227 were also unused and empty. All of these pages except the last two, still blank, now contain the eighty-three ballads written, as I think I can show, by Collier in a simulated hand of the seventeenth century.

Below is a list of the ballads. The first column consists of Collier's serial numbers (some counted but not written by Collier and now silently supplied). The second gives the number of the leaf on which

each ballad begins. The third is titles. The fourth gives the first three or four words of each ballad. The fifth consists of Collier's notes, with some bracketed additions, corrections, or substitutions. Most of Collier's notes consist of page numbers or volume-and-page numbers, which I will explain shortly. Nine of the notes identify or lead to printed sources from which Collier took certain ballads; these too I will enlarge on presently. Those on Nos. 54 and 77, like the query on No. 75, are typical specimens of his puckish mystification and verisimilitude; Collier knew perfectly well whence he had copied the ballads. The note on No. 1 reads "See Massinger 'Believe as you List' pr. by Percy Society p. 65," where the words "a new rigg'd pinnace" are used, as the same figure is in the ballad, of a courtesan. The dates supplied for Nos. 62, 63, 64, 73, and 74 are dates of events or subject matter. The note on No. 59 is baffling. All of these notes, Collier's serial numbers, and of course the foliation, are in pencil, in Collier's normal nineteenth-century hand. One note shown in the list, connected with No. 76, is an exception, for it is in ink and in the simulated old hand — another bit of mystification, though the possibility that there is or was another book is not to be ruled out.

                                       

1.	107v	The Pinnace	A pinnace riggd	[see above]
2.	108	To his Lady	O pity me
3.	109	*Giue Place You Ladies	Giue place you	p. 152 p. 170
4.	110	*Kitt hath lost her Key	Kit hath loste	p. 55
5.	110v	*Youth and Age	Olde age and Youth	p. 83
6.	111v	*Women best when at Reste	Women are best	p. 23
7.	112	*The Praise of Milkemaydes	Passe not for	p. 90
8.	113	*All in a garden Greene	All in a garden	p. 196
9.	114	Dames of London	Who so defames
10.	115	*Ladie Ianes Lament	Now must I lose	p. 72
11.	115v	*The Cuckoes Song	ffull merilie singes	p. 122
12.	116v	*The Kinges hunt is vpp	The hunt is vp	p. 129
13.	117	*The Batchelour	Hough for a	p. 43
14.	118v	Agaynst Idelness	Though idlenes in	See Shak. Soc.
15.	119	*Beauties fforte	When raging Loue	p. 5
16.	119v	The Maryed Mans lament	To all whose wyues
17.	120v	*The newe Hunts vpp	The hunt is vp	I. 130
18.	121	*Maides and Widowes	If euer I marry	p. 9
19.	121v	Truth hath a quiet Breast	Whatere my ffortune
20.	122	The Louer Scoffed	Attend thee goe	H. of Plea Delights

                                                                     

21.	123	*The Damned Soule in Hell	O cruell paynes	p. 117
22.	123v	*The Christians A. B. C.	All men that harken	p. 104
23.	124	*Citie and Countrey Maidens	Mall. ffaith Ione	p. 145
24.	125	*Awake and Arise	After midnight when	Sh Soc 89 p. 186
25.	126	The Clowne turnde Gentleman	A countrey clowne
26.	127	Home is still Home	Home is still
27.	127v	When raging loue	When raging loue	Surrey
28.	128	The Praise of May	When May is	Par D. Dev
29.	129	*Against Couetousnes	Noe wight in	p. 33
30.	129v	The cittie Maide & Countrey Maide	M. ffaith Cicely
31.	131v	Daintie come thou to me	He Wilt thou from	Roxb Bal
32.	132v	*The burning of Powles	Lament eche one	p. 40 1561
33.	133	Try ere thou trust	Try ere thou
34.	133v	*Wine Women and Dyce	Wine women and Dice	II.82[69]
35.	134v	The praise of the Gilliflower	If I should
36.	135	*Maide will you Marrie?	He Maide will you	II.6
37.	135v	The Tinker and ye Countreyman	T I am a iolly
38.	137v	*The wickednesse of Cruell Women	Though Helen were	p. 132
39.	138v	*Loue me little loue me long	Loue me little	p. 213
40.	139	My Prettie little One	Sweete if thou
41.	140v	To his Ladie	The rushing Riuers	Googe
42.	141v	Womens Tongues	When riuers turne
43.	143	What is my Ladie like?	My Ladys like
44.	144v	*A Caueat for Beauty	Ye dames that	[II.62]
45.	145v	Tinkers Truths	Dick Tarleton was
46.	146v	Send me thy sonne	Send me thy
47.	147v	Tarltons Toye	A countrey damsell
48.	148v	A loue Song	So much I loue
49.	149	What is my Seruaunt like	My seruaunts like
50.	150	*Defence of a bald Head	B. Baldnesse though	II.97
51.	151v	The Praise of good Ale	When cloudes of
52.	152v	O Yes: list the Cryer	If once man or
53.	154	*The Vertuous Wife	ffaine would I	p. 162
54.	154v	Dialogue betw. Venus and Diana	V. All gallant youthes	Qre by Yates, Servingman
55.	155v	When Knaues wilbe honest men	When mountains moue

                                                       

56.	157	Of his Lady	What flower is my
57.	158	A Toy of Elderton	Will Eldertons red
58.	159	The olde Man's Song	ffull long ago
59.	159v	*The choice of Friends	In choice of	MS 220 [II.189]
60.	161	A iest of Scoggin	Scoggin a street
61.	161v	Corsbies Confession	If teares could
62.	163	ffatall ffall at Paris Garden	Listen a while and	1583
63.	164	*The death of Deuoreux	Lament lament for	1576
64.	165	The life and Death of Lo. Graye	Shed teares wth	1562
65.	166	*A iest of Peele and Singer	G. Peele & Singer	[II.216]
66.	167	*Two Spanish Louers	If you list vnto	[II.200]
67.	169v	Madge Howlets Song	ffull mournfully hootes
68.	170v	The disobedient Prophet	A prophet of the
69.	172	*Gods iudgement on a Sorcerer	Of all the crimes	[II.30]
70.	173	The Praise of a Whore	I haue lou'd to
71.	173v	*Life and Death	L Nay what art	[II.43]
72.	175	The Song of a Louer	Desire hath driuen
73.	176	*Against the newe Playhouses	The fire that frō	about 1576 [II.152]
74.	176v	*The great Earthquake	Take warning London	6 Apr. 1580 [II.111]
75.	177v	Spurina and Roman Ladies	If nature beare	Qre printed
76.	178v	The Cobler of Colchester	Walking abroad in	V. my other Booke
77.	181	The Louer his Lullabie	Sing lullaby as	Qre Gasc? Yes, p. 8 1587
78.	227	Churchyardes ffarewell	With shaking handes
79.	243	The English Rose	Of euerie flower
80.	243v	Husband and Wife	Dear Wife and if
81.	245	Murther of Iohn Bruin	Murther it is a
82.	39	Ballad of his Mistresse	Shepherdes haue you
83.	40	The wise Mans Warning	In youth when teares

Of these ballads the following are those for which Collier's notes furnish or lead to identifications:

• 14. John Heywood's Works and Miscellaneous Short Poems, ed. B. A. Milligan, Univ. of Illinois Press, 1956, p. 256; also in J. O. Halliwell [-Phillipps], ed., The Moral Play of Wit and Science, and Early Poetical Miscellanies. From an unpublished manuscript, B. M. Addit. MS. 15,233, Shakespeare Soc., 1848, p. 79. The poem as printed in these

  ---

  11

  
  
  
  two editions consists of a burden and six stanzas, of which Collier's version includes four, with many alterations. He probably used Halliwell-Phillipps' collection, as his marginal note suggests.
• 20. A Handful of Pleasant Delights (1584), STC 21105, ed. H. E. Rollins, 1924, p. 12. Collier's text shows substantive alteration, including additions and omissions.
• 24. The Moral Play of Wit and Science (as in 14, above), p. 89. Halliwell-Phillipps' version consists of a burden and twenty stanzas, Collier's, of eleven stanzas, some owing little to the original, others more or less altered. Of the thirty-four ballads printed in the notes in the Extracts this is the only one for which I have found a printed source.
• 27. Tottel's Miscellany (1557), STC 13860, ed. J. P. Collier, Illustrations of Early English Poetry, I (1886-70), 18; also ed. Rollins, 1965, I, 14. Collier makes a few alterations.
• 28. A Paradise of Dainty Devices (1576), STC 7516, ed. Collier, Illustrations of Early English Poetry, I (1866-70), 13; also ed. Rollins, 1927, p. 9. There is a fair amount of Collier's usual tampering with the text.
• 31. Collier presumably took this from the broadside in the Roxburghe collection of ballads acquired by the British Museum in 1845 from the estate of B. H. Bright, who acquired them in 1813. The ballad was later printed in the Roxburghe Ballads, I (ed. W. Chappell, 1888), 629. In Collier's version almost every line is lengthened by the addition of an extra foot or syllable (a liberty characteristic of Collier), in an attempt at improvement.
• 41. B. Googe, Eglogs, Epytaphes, and Sonnettes (1563), STC 12048, ed. E. Arber, English Reprints, 1871, p. 105; also ed. F. B. Fieler, Scholars Facsimiles & Reprints, 1968. Collier's version contains several verbal alterations by way of improvement.
• 54. James Yates, The Castell of Courtesie (1582), STC 26079, sig. L3. Collier's text is essentially that of the original, but with many verbal alterations little affecting the sense, some of them improving the metre, and all looking like Collier's work.
• 77. G. Gascoigne, A Hundreth Sundrie Flowres (1587), STC 11638, ed. C. T. Prouty, University of Missouri Press, 1942, p. 150.

Four other pieces that he drew from printed sources Collier does not identify. These are as follows:

• 33. A Gorgeous Gallery of Gallant Inventions (1578), STC 20402, ed. Collier, Illustrations, III, 124; also ed. Rollins, 1926, p. 100. Collier's text agrees fairly closely with Rollins'.
• 40. Collier must have taken this either from the broadside in the Roxburghe collection (see No. 31 in the foregoing list) or from that he

  ---

  12

  
  
  
  himself had recently published, A Book of Roxburghe Ballads, 1847. Collier's manuscript consists of only eleven stanzas and differs in other repects from his own printed version with fifteen stanzas. That this latter followed the original broadside closely (except in omitting one stanza) is shown by its (for Collier) surprising textual agreement with the sixteen-stanza version printed in the Roxburghe Ballads, VI (ed. Ebsworth, 1888), 277.
• 72. A Gorgeous Gallery, ed. Rollins, p. 24. Collier's text follows the original fairly closely except for the interpolation of an erotic stanza.
• 75. A Paradise of Dainty Devises, ed. Rollins, p. 59. A stanza probably of Collier's composition is substituted for one that appears in all early editions and in Rollins'.

The many page references that Collier pencilled in his book in connection with the eighty-three ballads (shown above in the complete list) are references to a work of Collier's own, Extracts from the Registers of the Stationers Company . . . 1557-1570 and, Vol. II, 1570-1587, Shakespeare Soc., 1848, 1849 (hereafter abbreviated as Extracts). Collier's selection of entries from the Register for printing in this work is in keeping with his life-long concentration upon the popular literature of the sixteenth and seventeenth centuries. The entries are of ballads and other light verse, plays, tales, romances, and the like. For about half of his entries he furnishes notes that exhibit his quite impressive familiarity with the material. They supply bibliographical data, gloss words, discuss subject matter, suggest alternative readings, and occasionally assert that a ballad has not survived in print. It is with these notes that we are mainly concerned. Thirty-three of them, mostly notes to ballad entries, say that the piece entered in the Register is not known to exist, or does not exist, and add a reference to a manuscript volume in the editor's possession that contains the ballad in question. Then follows the ballad printed in full. The note that includes ballad No. 12 also prints one stanza of No. 17. Thus in thirty-three notes thirty-four of the ballads from Collier's old manuscript volume (which is of course Folger MS. V.a.339) are printed fully or in part. Henceforth I refer to these as "the thirty-four ballads." They are the ballads marked with asterisks in the list on pages 8-10, above. Strictly speaking, they are not all ballads in any proper sense, but Collier always so calls them, and no confusion will result from my using the term for the sake of brevity.

My primary interest is not in the ballads themselves, but in the way they came to be written in the old commonplace book. The evidence will show that before "finding" the thirty-four ballads and identifying

them in the Stationers' Register, Collier himself wrote them — them and the rest of the eighty-three.

Here we may pause to consider what Collier asks us to believe. In his preface to the Extracts (I, p. vii) he writes as follows:

If Collier's date were correct — if the ballads had been written in the commonplace book in the reign of James I — we would find nothing odd in the fact that twelve of the ballads that are on datable historical subjects (Nos. 32, 47, 57, 61, 62, 63, 64, 65, 73, 74, 78, 81) are all Elizabethan and we would scarcely notice the fact that the early poetical collections in which ten of the ballads had been printed were all published before 1588. That the entries in the Stationers' Register that correspond with thirty-four of the ballads were all made before 1588 we might think a curious coincidence but would not regard as utterly incredible.

But we know that the second of two early writers in the commonplace book was still at work in 1648 and that the ballads were written after he had stopped. While Collier was working on his second volume he must have come upon the evidence of a date much later than the time of James I. Several of the notes in Volume I refer to that time as the date of his manuscript volume, but notes near the end of Volume II give the date as "about the middle of the seventeenth century" (p. 200) and "about the reign of Charles I" (p. 215), and the new preface makes a perfunctory correction of his earlier assumption. That Collier ever saw the implications of this later date there is no reason to believe. The implications are severely damaging of his whole case regarding the ballads. He would have us believe that a ballad collector of 1650 gathered eighty-three ballads, of which thirty-four (forty percent) had been printed, or at least entered, between 1570 and 1587 (just the period of Collier's Extracts) and none that were later,[1] twelve on historical events of the reign of Elizabeth (none later),

and ten that had been printed in collections before 1588 (none later).

One of the most curious features of the ballads is their relationship to the Stationers' Register, and these go deeper than we have yet seen. Further examination of the thirty-four ballads (those printed in notes to the Extracts) reveals another kind of relationship, which is made clear in the two tables that follow:

Table I

                                                                       

Ballad number	Where printed	Date of entry
3	I.152	1566-7
4	I.55	1561-2
5	I.83	1563-4
6	I.23	1558-9
7	I.90	1563-4
8	I.196	1568-9
10	I.72	1562-3
11	I.122	1565-6
12	I.129	1565-6
13	I.43	1561-2
15	I.5	1557-8
17	I.130	1565-6
18	I.9	1557-8
21	I.117	1565-6
22	I.104	1564-5
23	I.145	1565-6
24	I.186	1568-9
Ballad number	Where printed	Date of entry
29	I.33	1560-1
32	I.40	1560-1
34	II.69	1570-1
36	II.6	1570-1
38	I.132	1565-6
39	I.213	1569-70
44	II.62	1578
50	II.97	1579-80
53	I.162	1566-7
59	II.189	1584-5
63	II.35	1576-7
65	II.216	1586-7
66	II.200	1585-6
69	II.30	1576-7
71	II.43	1577
73	II.125	1580-1
74	II.111	1579-80

Table II

                                                                       

Where printed	Balled number
I.5	15
I.9	18
I.23	6
I.33	29
I.40	32
I.43	13
I.55	4
I.72	10
I.83	5
I.90	7
I.104	22
I.117	21
I.122	11
I.129	12
I.130	17
I.132	38
I.145	23
Where printed	Balled number
I.152	3
I.162	53
I.186	24
I.196	8
I.213	39
II.6	36
II.30	69
II.35	63
II.43	71
II.62	44
II.69	34
II.97	50
II.111	74
II.125	73
II.189	59
II.200	66
II.216	65

What can be seen at a glance, in Table I, is this: that of the first twenty-four ballads written in the manuscript (Nos. 1-24) Collier was to "find" that seventeen corresponded with ballad titles entered in the Stationers' Register before 1570 and that the last eight corresponded with titles entered after 1575. Thus there is a roughly chronological order in the thirty-four ballads as they stand in the manuscript. In the ballad-number column of Table II I find five pairs of numbers that are consecutive in Table I: 29 and 32, 22 and 21, 11 and 12, 74 and 73, 66 and 65. Among thirty-four ballads this kind of correspondence among ten does not look like pure chance.

If we are to believe in Collier's sevententh-century ballad collector we must be able to put forward a reasonable hypothesis to account for the unmistakable traces of chronological arrangement in thirty-four of the ballads as they stand in his manuscript. The assumption that he copied these from printed broadsides does not help, since the printers of ballads, fully conscious of their ephemeral nature, almost never dated their imprints. Our supposed collector would have had to be a tireless researcher to achieve even such a degree of chronological arrangement as we find. Yet with the twelve ballads about easily-dated events — Lady Jane Grey's lament, the burning of St. Paul's steeple, the great earthquake, and the others — he took no trouble to set them down in any order.

I can make no sense out of this. But if instead we assume that Collier himself wrote the ballads in the commonplace book we can reach a reasonable explanation of all the main features of the collection. Looking into the Stationer's Register and finding it rich in entries of popular literature, he began, in 1847, transcribing all the entries that interested him, with a view to publication. At some point during the work of editing it occurred to him that he might compose pseudo-Elizabethan ballads to fit the titles that struck his fancy as he came across them and that he could then "find" them and print them in his notes. We cannot know whether or not he wrote the ballads in the order in which he found appropriate titles. It was obviously not in this order that he copied them into his commonplace book. The first ballad listed in Table I, above, is the eighteenth in Table II, and the sixth in Table I is the twenty-first in Table II. Presumably Collier first wrote the ballads in his ordinary hand on ordinary sheets of paper; at what stage in his work he copied them into the book appears to be inscrutable and irrelevant. When he prepared for the task of copying in his "old" hand, though he did not sufficiently appreciate the problems of avoiding detection, he realized that the thirty-four

ballads must not stand in the manuscript in the fully chronological order in which they would appear in the printed Extracts. It was clear to him also that his manuscript collection must not consist solely of the thirty-four ballads entered in the Stationers' Register between 1557 and 1587. There must be others with them, and these he could of course manufacture as easily as he did most of the thirty-four. It is reasonable to suppose that he had long entertained himself by writing Elizabethan ballads and that he had some of these laid away, and he could write more as they were needed. Accordingly he selected a few suitable pieces, perhaps wrote some new ones, and, for verisimilitude, gathered a dozen or so from Elizabethan printed sources. These he interpolated among the sheets containing the thirty-four and shuffled them all together. The shuffling was adequate (as Table I shows) — or nearly so — but the interpolating was remarkably uneven, as can be seen in a list in which the thirty-four are in roman type, the interpolations in italic: The uneven interpolation, light at the beginning, heavy in the middle and at the end, does not look like chance. Two alternative and almost equally attractive explanations present themselves: (1) Collier was unable, till he had copied about twenty-four, to estimate how many ballads the available pages would take; (2) it may have been only after he had copied twenty-four of the ballads that he observed that the interpolation was too light and must be increased. Neither explanation accounts for the presence of the last nine superfluous ballads; perhaps he merely wanted to fill up the space. Since the whole eighty-three are included in a list that Collier prints in the preface to Extracts, II (1849), we cannot suppose any to be later additions.

I have been assuming that all but one of the ballads printed in the Extracts and most of the others were of his own composition. He may have drawn from printed sources more than the thirteen identified above, but I doubt that many more will prove to be of this sort. The existence of some seventy ballads apparently unique among the eighty-three argues not only against an early collector but for Collier's authorship. Further evidence, tenuous though it is, supports this assumption. In his notes, where he is vague as always in speaking of his old manuscript, Collier frequently notes that a ballad is not known to have

survived in print or that he has not heard of a copy. But four times he is positive: No. 5 "has been preserved in MS. only (penes the Editor)," and Nos. 11, 21, and 34 are known only in manuscript. These lapses, as they must be, into expressions of the positive knowledge that only the author could have had seem to reveal more than he intended.

More telling perhaps, when we take Collier's evident vanity into account, is his praise of certain ballads — praise not warranted by any special excellence that I can recognize. No. 3 "runs thus pleasantly, and by no means unpoetically." No. 11 is a "remarkable and spirited ballad," No. 12 "an extremely spritely performance." No. 15 "has considerable merit" and "must have been written by no inferior hand." About No. 21 "There is something very striking in the manner and measure," and No. 71 is "a remarkable and striking relic of the time." Collier's vanity is well illustrated in a letter to J. W. Ebsworth, 12 August 1879 (an uncatalogued Folger manuscript), where he writes of his Poet's Pilgrimage (1825), a poem in four dull cantos, "I thought, & think, that I never wrote anything else that was half as good — yet nobody cares a fig about it, but its poor old Author. . . . It is good: it is of its kind, highly meritorious and imaginative. It is good, let who will say the contrary, and by it I will live & die." His sensitiveness as a result of the neglect of that poem was such that he never again, I believe, publicly exposed an acknowledged poem of his own. Perhaps that is all the explanation we need for Collier's cryptoauthorship of the ballads published in the Extracts, there exposed in the hope that they would be praised by others.

That Collier could and did write ballads meant to be approximately Elizabethan in character is proved by several among the many specimens of his own verse scattered through his privately-printed Old Man's Diary (2 volumes, 1871-72). In another letter to Ebsworth, 27 October 1879, he writes, "ballad writing is merry work, if good, both for writer and reader." I do not wish to press the evidence of Collier's authorship of most of the eighty-three ballads, though it helps to explain the uniqueness of seventy of them. The point is not essential to my purpose.

If it was in hope of praise that Collier printed his ballads in the Extracts he was not altogether disappointed, for William Chappell thought highly enough of seven of them (Nos. 7, 8, 12, 17, 18, 32, 57) to print them in his Popular Music of the Olden Time (2 vols., London, 1859), with references to the Extracts or to Collier's old manuscript, which Chappell could have known only from the Extracts.

Collier also lived to see No. 10, "Lady Jane's Lament," printed by Furnivall in Ballads from Manuscripts (1868-72), the first volume of the Ballad Society's publications. Norman Ault thought three of the ballads (Nos. 18, 36, 39) worthy of inclusion in Elizabethan Lyrics (1949 ed.), pp. 37, 514, 61.

II

Whether Collier composed the ballads or not we can be quite sure that it was he who copied them into his old commonplace book. If, then, it can be demonstrated that the annotations in the Perkins folio are in the handwriting of those ballads, Collier's fabrication of these will have been proved. That demonstration is my next task. It ought not to be a difficult one, for, given two such ample specimens to work with, it will always be possible and usually easy to determine whether they are the work of one hand or two.

Each half century or so has its own characteristic handwriting, the norm being the work of the writing masters, the schoolteachers, and the copybooks. But every schoolboy sooner or later, depending upon how much writing he does, departs from the prescribed norm — adopting his own comfortable slant, his own penlifts and linkages, his own ways of forming certain letters. Thus by the time he reaches maturity he possesses his own peculiar hand, as much his own as his fingerprints. A close analysis of the writing of two persons will therefore bring to light many fairly consistent differences even though the two specimens bear a strong superficial resemblance. Without such a resemblance the question will hardly arise.

This is true of today's hands and of the hands of any century. It is more strikingly true of the secretary hand of the sixteenth century than of later English hands. Two men's secretary's hands are easier to distinguish than two later hands because many letters of the secretary alphabet are more complicated than those of later alphabets. This complexity provided greater scope for individuality, and men took advantage of this, for they set store by individuality in writing. A father admonishing a son at Oxford in 1622 charged him with a "barren invention" in his writing, for the son's hand lacked character and individuality — qualities that the father's possessed in abundance.

The two specimens that I am to examine are secretary only in part. They are of a style of writing common in the seventeenth century, for which the term mixed hand seems appropriate because of the mixture in it of secretary letters and italic. The secretary hand was a native English hand that for the greater part of the sixteenth century was

most widely in use for literary composition, letter-writing, and business. Since it developed from earlier hands its beginnings cannot be dated, but we find it fully developed by 1525. For the next hundred years it remained the common hand, taught to schoolboys and written by most men. But it was not the only hand of the sixteenth century in England. Before the beginning of the century the italic hand — a new hand in Italy that grew out of the humanistic script — was brought to England. It was in the sixteenth century a rather special hand regarded as more beautiful than the secretary, but at first used mainly at the universities, later adopted by many noblemen. Well before the end of the century we find other men using it as a second hand; Ben Jonson was one who did, a little later.

The increasing popularity of the italic hand near the end of the sixteenth century and in the seventeenth was, by 1700, to drive the secretary out and take its place as the common hand. The process by which this conquest took place is not clearly understood. Some schoolteachers may have been teaching it, but I suspect that it was the growing use of the italic as a second hand that gave rise to the mixed hands — half secretary, half italic — that are observable by 1625 and very common before 1650. For some letters, c and e, for example, there appears about the middle of the century to be a strong preference for the secretary in mixed hands; in several other letters the preference, more or less strong, was for the italic. Some letters can hardly be called secretary or italic because identical forms are used in both. Other considerations too limit the letters that we can profitably use in examining mixed hands to six minuscules and four capitals. In the table below, using these letters (with two kinds of s) I show the mixtures used by ten seventeenth-century writers and by Collier in the hand of the ballads. Columns 1 and 2 represent the two early writers in Collier's manuscript volume, columns 3 through 10, eight hands written between 1629 and 1665 drawn from a collection of the correspondence of Col. Robert Bennet (Folger MS. X.d.483 (1-209)). Column 11 represents Collier's hand. S stands for secretary, I for italic, SI for letters that are mixed by a single writer — now secretary, now italic.

       

1

2

3

4

5

6

7

8

9

10

11

e

S

S

S

SI

SI

S

S

S

S

S

S

f

I

SI

SI

I

I

I

I

I

I

I

I

g

I

S

S

SI

S

S

S

I

I

I

I

               

h

SI

SI

I

SI

I

I

I

SI

I

SI

S

p

SI

S

I

S

S

SI

S

S

S

S

S

init.-med.

s

I

SI

S

SI

I

I

I

I

I

I

I

final

s

I

I

S

SI

SI

SI

I

S

I

I

S

E

I

I

I

I

SI

I

I

I

I

I

S

G

S

S

S

I

I

I

S

I

I

I

I

H

I

S

S

S

I

I

I

SI

S

I

S

T

SI

SI

S

SI

I

S

I

SI

S

I

S

Of the ten genuine hands represented here no two exhibit the same mixture. If we omit f and E because each shows a tendency to be of one kind or the other and ignore some other probable but imponderable tendencies, it will appear that the chance of finding the same mixture in two different hands of the midseventeenth century is one in 6,561. Collier's ballad hand, column 11, shows, of course, a unique mixture. The hand of the Perkins annotations shows the same mixture, with the exception of one letter which I will explain and which does not reduce the significance of the similarity.

It is, however, the individual letters that must carry the burden of proof. In working with the minuscules we must look for letter-shapes or other characteristics that are eccentric, unusual, highly individual in the period under consideration — the seventeenth century because the hand purports to be of that century. If we find frequently recurring in one specimen a certain letter formed in a curious and uncommon way and then find the same curious form in the other specimen, again recurring frequently, this will suggest a single writer. If we find half a dozen such letters a single writer is strongly indicated. In what follows I attempt to demonstrate that certain letter-shapes common in Collier's ballads do not conform to standard or common practice of the midseventeenth century. To make this demonstration I refer to Plate IV, where I reproduce many examples of three minuscles and one capital letter found in the Bennet correspondence described above. In gathering these examples I took the characteristic shape or shapes of each of four letters as I found them in each of a large number of documents written by different hands. Written above each letter is the number of the document from which I took it. Several of the shapes reproduced are unique among those shown; had I gathered other letters too, I should have found more unique shapes, possibly some of them at least roughly identical with shapes common in Collier's ballads. I cannot

assert that any letter-shape in the ballads is unique. What the reproductions in Plate IV do show, I think, is what we may regard as the common, usual shapes that we might expect to find in most mixed hands of the midseventeenth century. These reproductions, as I have taken the liberty of calling them, are not facsimiles, but the best copies that I could make by hand. Each of my reproductions fairly represents the proportions and shapes of the letters; a photograph would prove no more than these prove. For the letters that I discuss but do not reproduce in Plate IV I made a careful search through the Bennet letters to assure myself that the shapes that I discuss in the ballads are not (unless I so designate them) generally standard or common shapes.

Below I examine in detail three minuscules and three capital letters that are to be found both in the ballads and in the Perkins annotations.

f This minuscule differs from the tall s of the ballads (hereafter abbreviated Bal.) and the annotations (hereafter Ann.) in having, usually, instead of the straight shank of the s (as in staring, Plate II, f. 114v, line 4), a bowed or hooked descender (finde, faire, Plate I, f. 114, lines 16, 17), in having, often, a looped head (first, Plate I, f. 113v, line 6), and in being crossed just above the baseline. The cross is what makes it an f, not an s. About this letter, as found in Bal. and Ann., there is nothing distinctive or interesting. It becomes so only when followed by and joined to a t, for here Collier almost, but not quite, always forms an s instead of an f; that is he does not cross the shank. An example may be seen in often, Plate II, f. 114v, line 14. Elsewhere I have found ten more of these in Bal. and four in Ann. — three of the four shown in Plate III: oft in (b), line 1; after in (d), line 2; left in (g), line 1. Three times (to fifteen) I have found ft properly made — one of them shown in Plate II, f. 114v, line 13 (soften). I attach considerable weight to the appearance in both Bal. and Ann. of this curious abberration.

g Collier's typical g, an italic, may be seen in all of the reproduced pages of Bal. Those in staring and louing (Plate II, f. 114v, lines 4 and 15) are typical. The common italic g of the seventeenth century is essentially the same as the g of today. The secretary g is distinguished by being open at the top, like our y, then crossed by a separate stroke that usually forms the link with the following letter. Several of the examples shown in Plate IV are secretary letters that the writer forgot to cross. Of the italic forms there shown none resembles the characteristic Bal. g with its hooked descender, which we must conclude is not a usual or standard form of about 1650. The same g is to be found abundantly in Ann. and may be seen in Plate III (c),

line 1, in the words grace and high. A variant to be discussed later can be seen in brings, in Plate III (e) and in other words in (f) and (g).

h The secretary h of Bal. is a perfect copybook shape, almost identical, at its best and most typical, with the shape shown in the secretary alphabet in De Beau Chesne and Baildon, A Booke Containing Divers Sortes of Hands (1571). But in practice few writers of any time formed the letter in that way; most men made the loop lower and rounder, like the secretary specimens numbered 2, 5, and 15 in Plate IV, which are typical Elizabethan shapes. The examples in that plate indicate that the Bal. h is not the usual one in the midseventeenth century. The common shape of h in Bal. may be seen in might in Plate I, f. 113v, line 1. It can also be seen in thankes and thanke in III (b) and (c), respectively, and it is the usual h in Ann.

s The secretary hand makes a distinction between the s used initially or medially and the final s; they are two quite different characters. That Collier makes this distinction may be seen in Plate I, f. 113v, line 3, shines. Only the initial-medial s now concernes us. This can be seen in virtually every line in Plates I and II and in Plate III (b), (c), (f), and (h). That it is not a standard or usual form is evident in Plate IV, where Nos. 3, 6, 7, 10, 31, and others are secretary. The Bal. s, also found everywhere in Ann. is not really either secretary or italic; in fact I do not remember that I have elsewhere seen such an s. So highly eccentric and abnormal is this letter that it is hard to see how Collier came to adopt it.

Before beginning the examination of individual capitals I must make one or two observations about capital letters in general. Most of these had in the sixteenth century, as they have now, more than one acceptable, almost standard, shape, and many writers used interchangeably two or more shapes for certain capitals. Again, in the sixteenth century and now a plain roman block capital, resembling roman type, may be used for almost any letter. Many sixteenth-century writers commonly used both a standard secretary C and a roman block capital, either plain or embellished, and so with other letters as well. Plain block capitals are almost worthless for the kind of investigation I am attempting, but a number of secretary capitals and some italic are useful.

F Collier's ordinary form of this capital, a secretary form, has little individuality and is not useful. The F that concerns us is an italic letter. In Plates I and II, on ff. 114 and 176v, two titles of ballads are written in pure italic. From the sixteenth century onward it was a common practice in secretary writing to use italics for titles, for Latin

words, and for such other purposes as printers used italic type for. Following this custom, Collier's ballad titles are almost all in italics, though with some scattered mixture, presumably unintentional, of secretary letters. Likewise in the Perkins annotations stage-directions and speakers' names are by and large italic, and here I find three examples (Comedies p. 274, Histories pp. 56 and 79 [2H4]) of an eccentric and therefore useful F. The peculiarity of this F (not shown in Plate III) is that the lower, short bar is not horizontal but a mere dot elongated vertically. No such F is to be found in the ballads, but an E of exactly the same sort can be seen in Plate II, f. 176v, in Earthquake, and another example of the E is on f. 154 (not illustrated).

G Typical examples of the italic Bal. G are to be seen in Plate II, f. 176v, lines 5 and 9-up. Of the twenty-eight examples of this letter shown in Plate IV, most are, like the Bal. G, italic, but only No. 31 is of the same type. The same G as that common in Bal. occurs throughout Ann. and may be seen in Plate III (a).

H This capital, shown in Plate I, f. 113v, line 5, occurs throughout both Bal. and Ann. with little variation. Though a perfectly acceptable secretary letter, its shape is not a common one at any time. Of eighteen examples of H that I found in the Bennet collection only three were secretary, and none bore close resemblance to Collier's.

Several capital letters must be approached in a different way, since what is interesting about them is not any striking eccentricity of form, but the variety of forms in which they appear. This could be shown to hold true in both Bal. and Ann. of the letters B, C, D, L, M, N, P, R, S, and W. A single one of these letters will perhaps be sufficient to show how any or all of them would confirm the evidence that I have already produced. Though the N exhibits fewer varieties than either M or W it will, through the exigencies of the Bal. reproductions, serve well. I find five major types of N, all of which can be seen in Plates I and II. They are as follows: (1) a type common in the sixteenth and seventeenth centuries and the commonest of the five in Bal. — Plate II, f. 176v, line 3-up; (2) almost as common as the first, in Bal. — Plate II, f. 114v, line 9-up; (3) Plate I, f. 113v, line 1; (4) Plate I, f. 113v, line 8-up, and f. 114, line 4-up; (5) the only secretary form of the letter in Bal. and one of only two examples there — Plate I, f. 114, line 13-up, a form that I have not found in the seventeenth century. All of these forms of N occur in Ann., including (5), which I have found there only once. Similarly all the types of M and W that I have seen in Bal. I have found also in Ann., together with at least one additional type of each and several variants not found in Bal.

In discussing the minuscule g above I pointed out one of several examples of a g not found in Bal. In Ann. this variant occurs about as frequently as does the Bal. g. The only other minuscule of which I have observed a new variety peculiar to Ann. is the w. The Bal. w may be seen in Plate I, f. 113v, lines 2, 5, 6, and 7; it may also be seen in Ann. in Plate III (a), with, (c), wthout, and (g) way. The new Ann. variety is shown in III (b), wth , (c), which, and (e), what. Like the g, the old Bal. w and the new variant appear to be evenly mixed in Ann.

Of the capital letters that we have examined, F, G, and H, no significant variants are to be found in Ann. (always excepting wholly italic stage-directions and speakers' names). But I find in Ann. three other capital letters, A, E, and I, that are either unknown in Bal. or almost unknown. Of the A the Bal. type (of which some half dozen are to be seen in Plate I, f. 114) is common in Ann as well (Plate III (b)). But there is a new Ann. type, visible in Plate III (b), line 2, (e), first word, and (f), first word, and this type I have not seen in Bal. Apart from stage-directions and speakers' names, the old Bal. A appears to outnumber the new variety by about five to three. The E common in Bal. is a secretary letter, a rather malformed example of which may be seen in Plate II, f. 114v, line 3. In Ann. this is replaced by an italic E that can be seen in Plate III (g). The switch from the secretary E of Bal. to the italic of Ann. would be perfect were it not for a curious single appearance of an italic E in Bal., at f. 159 (not illustrated) and a balancing secretary E in Ann., where it appears twice (at 3H6, p. 164, and Tragedies p. 127. With the letter I the story is different, for the switch from the Bal. I to the new I of Ann. is, so far as I have discovered, flawless. Actually two types are common in Bal. — a crossed form (Plate II, f. 114v, line 17) and a less correct form without the cross (f. 176v, line 6). Neither of these have I seen in Ann., where instead we everywhere find the quite different shape that can be clearly seen in Plate III (h), line 1.

In view of the overwhelming evidence that one hand wrote the ballads and the Perkins annotations, we are forced to look for an explanation of the new types of three capital letters and two minuscules that make their appearance in the annotations that will be consistent with the view that Collier wrote these as well as the ballads.

For this explanation we must go back again to the time when Collier began to learn the writing of an old hand. Clearly we can ascribe to him a superior manual dexterity and perhaps superior coordination of hand and brain. Even so it was a long and taxing labor —this learning to write a new hand. I can see him bent over his task

for an hour or two at a time, day after day, week after week, slowly acquiring the automatic translation of thought into written words. For when a man writes he is unconscious of the shapes of letters, or only half conscious of them, and yet they come out in his own peculiar hand. Such facility Collier obviously possessed when he wrote, with remarkable consistency, more than a hundred and fifty pages of ballads in his old commonplace book. To have stuck at the learning of that hand he must have been strongly urged forward by an objective that was very important to him. It is not necessary to suppose that he had the Perkins folio in view from the start; in fact this is highly unlikely, since he was writing the hand with at least some facility more than fifteen years before he announced the discovery of the annotations. Ingleby reproduces, in his Complete View, many documents which Collier had "discovered" in the Bridgewater collection, in Dulwich College, and in other collections and had announced and printed, and which N. E. S. A. Hamilton had tracked down and found to be forgeries (as Halliwell-Phillipps and others had earlier suspected). One of these, in the Bridgewater collection, which Collier announced and printed in New Facts regarding the Life of Shakespeare (1835), is in a hand differing from that of the ballads only slightly. Several others are in hands differing only a little more, and in others the ballad hand can be seen, as it were, bleeding through the attempt to disguise it. For a short document Collier could in various ways quite consistently and with fair success give his hand a number of different appearances. But for an extended job, like the ballads, he had only one hand.

When, in 1851 or 1852, Collier began work on the Second Folio he had no strong reason to fear detection, and it is therefore a little hard to understand why he wanted to alter his hand. It is even harder to see how he could have thought that altering two minuscules and three capitals could effectually disguise the hand. Nevertheless he appears to have made the attempt from the moment he began work on his new adventure in fabrication. But even in the small attempt he made at alteration, his inability to make the switch in his capital A and the minuscules g and w exhibits the difficulty of the task. Even so he must have been helped by the nature of the job, for few of his additions consist of more than one line, and not more than a couple of dozen are of more than four or five words. Probably eighty percent of the corrections, as he called them, consist of single words, a letter or two, or a mark of punctuation. There are certainly not less than 20,000 of them in all. This kind of writing would never be automatic; the writer would be working slowly and carefully. On the whole, I find

it a little surprising that Collier's success in altering was not greater than we find it to have been.

The fact, then, that a few letters in the annotations are different from those in the ballads does not weaken the strong evidence that both were written by one hand. The paleographical case against Collier is therefore complete. If weakness is found in the whole argument of this paper, it will hardly be found, I think, in the paleographical part of it, and if the demonstration that it was Collier who copied the ballads into his book falls short of mathematical proof, there is another point to be thought about. If Collier acquired his manuscript volume about 1840 and it then contained the ballads, he could hardly have failed to see, three or four years after he had copied out thirty-four of the ballads, that his remarkable annotated folio, assuming it to be genuine, was in the same hand. And if he saw it, why should he not have mentioned such a very remarkable coincidence?

Notes