University of Virginia Library

Search this document 


  

expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
The Issues and States of the Second Folio and Milton's Epitaph on Shakespeare by William B. Todd
expand section 
expand section 
expand section 
expand section 
  
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
collapse section 
 1. 
expand section2. 
expand section3. 
expand section4. 
collapse section5. 
 1. 
 2. 
expand section6. 
expand section7. 
 8. 
expand section9. 
expand section10. 
expand section11. 
expand section12. 
expand section13. 
expand section14. 
expand section15. 
  
expand section 

expand section 

81

Page 81

The Issues and States of the Second Folio and Milton's Epitaph on Shakespeare
by
William B. Todd

From the time it was recognized as a distinct variant, now almost a century ago, the "Coppies" issue of Shakespeare's Second Folio has been gradually received and accepted as the predecessor of a certain sequence of titles in several states and settings, and thus the first to represent, on the leaf conjugate to the title, the earliest issue of Milton's first-printed English poem. Such, in brief, is the account of an issue considered only as a freak in the days of Lowndes and Bohn, then regarded as "one of the earliest impressions" by Lenox, Winsor, Cole, and Pollard, and finally accorded preeminence over all in the exhaustive study published by Robert Metcalf Smith.[1]

When confronted by this imposing array of "firsts," in point of issue for book and poem, in two great names now conjoined in a single text, and in the roster of scholars who have labored to establish and justify the present hypothesis, it ill becomes an upstart to trample upon the work of generations and subvert the tradition that work has produced. And yet, if the facts of printing invalidate critical theories, it is the bibliographer's unbecoming duty to present them. Certainly the facts allow us to accept the premise common to all discussions of the problem: We are indeed concerned with several issues of the Second Folio, each conveying a different setting of Milton's Epitaph on Shakespeare. Beyond this, however, I must demur and eventually dismiss all contentions as irrelevant.[2]


82

Page 82

I. The Order of Issues

The issues of this Folio are distinguished by three settings of the letterpress for A2.5 of the initial quire, a sheet containing on the recto (outer forme) of one leaf the title to the work, and on the recto (inner forme) of the other several commemorative verses on the author. Two of the three settings for the title also exist in a number of states, usually identified by alterations in the impressions for Robert Allot, one of the principal shareholders, or by the special imprints provided for his collaborators. Between the variants for both leaves there is an obvious correlation best described, in Smith's terminology, as: Allot 1 associated with Effigies A; Allot 2-3 with Effigies B; and Allot 4-5, Aspley, Hawkins, Meighen, and Smethwick, all with Effigies C.

To substantiate this sequence, insofar as it applies to the Allot variants, Professor Smith believed that we may "assume, with a high degree of probability, that the order of the title-pages follows the order of improvement, since it is readily demonstrable that the compositor or compositors tried to correct in every succeeding title-page mistakes made in the preceding one."[3] The probability, it seems to me, is at times very remote. Where the corrections occur within the setting, as in Allot 2-3 and Allot 4-5, the intent and direction of improvement is apparent. But where the corrections require a completely new setting, as in the transition from Allot 1 to 2 and from 3 to 4, the necessity of such extraordinary measures for such trifling results remains obscure—at least under the present assumption.

Even more remarkable is the fact that whenever the compositor of the title discards his forme the compositor of the Effigies page also discards his at precisely the same time. This means that the two men, working together and in the same incomprehensible fashion, twice distribute the entire letterpress for the sheet, twice prepare others, and in the process aimlessly fritter away the hours (days, I should say) shuffling and reshuffling countless reglets, wedges, quoins, quads and spaces, as well as some 4960 sorts[4]—all this, we note, to arrange a few "improvements"


83

Page 83
which any dolt could have managed in five to ten minutes with several pieces of type. In view of this circumstance, one unparalleled in the annals of bibliography, the inclination is to cast about for another hypothesis wherein these seeming absurdities are reasonably explained.

For the moment, though, let us consider what else may be said for our diligent compositors and their heroic struggle to improve the text. In the sentence following the one cited, Smith assured us that "this assumption acquires a certainty with the demonstration of typographical links between successive title-pages."[5] Between the first two settings there is assuredly, not only a typographical connection in the employment of the same fonts of type, but also, as Smith elsewhere observes, a further correlation in the use of the same variety of paper. Obviously, then, these two were prepared within a relatively short period. Between the second and the third, on the other hand, no similarity exists. In Smith's words, the type is "not identical," "in a different font," and the paper for the later setting is drawn "from another stock."[6] Thus by his own admission the "links" disappear, the demonstration fails, and the expected certainty resolves into an increasing distrust of the entire argument. If the premise is unsustained all that is based upon it is in imminent danger of collapse.

At this impasse let us now turn to another theory immediately verified by evidence within the book and thereafter supported by every indication of relevance to this inquiry. Contrary to the received opinion, this would suppose that the Title-Effigies sheet was, for various reasons, deliberately underprinted and then, as the occasion required, twice reset at some later time to dispose of remainders. The sheet identifying each of the subsequent issues might therefore, in this view, be properly regarded as a "reprint," and like all reprints would presumably convey readings inferior to those in its copytext. Possibly, then, the order is the reverse of that alleged, and the sequence one of degradation rather than improvement.

Some credence is given this presumption by the questions the publishers themselves might have raised. Why run a complete issue of the most expensive sheet[7] in a very costly book? Since this was not an original edition, why imprint their names to something which might still be on the shelves some twenty years after they were gone and forgotten? Why not print just enough for their immediate needs and let events determine


84

Page 84
whether they, or their successors, should print again?[8] Should these or other considerations have convinced them—and now convince us—of the advisability of limiting the original issue, the publishers of the Second Folio may be credited with greater foresight than those of the Third and Fourth; for while they eventually chose not to alter the original make-up in successive printings, their precaution obviated any necessity for cancellation, a necessity later enforced upon their less perceptive successors.[9]

For the thesis, as now proposed, several pieces of evidence may be advanced, all tending to the same conclusion. In his examination of the correlations among the three settings Professor Smith neglected to observe that the two which are actually connected have no relation to the rest of the book, whereas the third, though independent of the others, is very closely affiliated both in the ornamental letters and in the paper used for this setting. The ornaments in all issues, consisting of the letters "S" and "W" on the Effigies page,[10] are of three kinds, each cut in a manner that suggests their origin in three separate foundries.

           
Group   Effigies   Ornamental block  
"S" against a filigreed background 
C & B  "W" against a similar background 
Type "S" within a wreathed factotum 
"S" against a broad leaf background 
"W" against a similar background 
Among these only group 1 is represented in the Folio.[11]

As with the ornaments, so with the paper: that used for the settings commonly described as the first and second (Heawood 594) does not appear elsewhere in this book or in any other book of this date, but that used for the setting usually considered to be the third (H 1420 or 1731) repeatedly occurs in the last nineteen quires of the Folio.[12] Now since the


85

Page 85
preliminary sheet would normally be printed in conjunction with other work toward the end of the book, the identity in paper confirms the identity in ornaments and establishes a strong presumption for an order in which the "third" setting is distinguished as the first. Thus, in the absence of the slightest evidence to the contrary, the Allot 4-5—Effigies C sheet must be regarded as the original issue, carefully prepared from manuscript, and the others as mere reprints, hastily composed, badly centered and spaced, and obviously degraded in text.

The exact order of the reprints may now be determined by the textual relationship among the title-pages.

Table I

                     
Issue  II  III 
[Smith]  [Allot 4-5]  [Allot 1]  [Allot 2-3] 
Line 
1a  Mr Mr Mr
1b  VV  VV 
HISTORIES,  HISTORIES,  HISTORIES 
7ab  fmpreſsion [double s separate]  fmpreſsion [double s separate]  Impreſsion [double s ligatured] 
Tho Tho Tho  
10a  Blacke  blacke  blacke 
10b  Pauls  Pauls  Pauls  
10cd  Church-yard.  Church-yard.  Church-yard, 
From Table I[13] it will be observed that, on this score alone, Allot 1 cannot be designated as the first in a sequence of issues gradually "improving" in the direction of Allot 5, for it is intermediate between 5 and 2, having seven variants in common with the one and three in common with the other. Hence for this reason, others previously adduced, and others yet to be divulged, I suggest that we reject the outmoded "Allot 1—5" classification and adopt one corresponding to the evident order of issues. Including only

86

Page 86
what I believe to be the first of the states in the initial setting (all of which will be dealt with in the next section) the three issues, with their variants, are identified as follows:            
Order  Ia  II  IIIa  IIIb 
[Smith]  [Allot 5—Effigies C]  [Allot 1—Effigies A]  [Allot 2—Effigies B]  [Allot 3—Effigies B] 
Description  1st issue,  2d issue  3d issue,  3d issue 
1st state  1st state  2d state 
Paper  H1731[14]   H594  H594  H594[15]  
Ornaments on Effigies page  1,1  2,3  1,3  1,3 
The inspection of a photostat from copy tentatively identified as a variant of the state now described as IIIb[16] leads me to believe that this is the artful work of a restorer.

With the three issues properly identified, and properly arranged in sequence, we should now endeavor to fix the approximate time and circumstance of publication. For the later settings, as for certain Jaggard-Pavier quartos, there is some indication that the imprint is misleading in all of its particulars. Unquestionably, as the paper attests, issue I was printed and sold in the manner announced on the title page. But II and III, though distributed with a similar announcement, were produced on such unusual paper that an intensive search through the crown folios in several libraries[17] has disclosed its presence not before 1637 and then only in three books: certain leaves of the preliminary quire of Camden's Britain (1637), printed by Felix Kyngston and others; throughout in Paris's Historia maior (1640), by Richard Hodgkinson; and occasionally as a single sheet in Parkinson's Theatrum Botanicum (1640),[18] ostensibly by Thomas Cotes, the printer


87

Page 87
of the Second Folio. Moreover, as the investigation has also revealed, the distinctive mark for this paper seems to exist only in the two states required for a single pair of moulds;[19] and as neither of these states shows any sign of deterioration the inference follows that the period of manufacture approximates the dates of the books in which the paper was found. It is quite probable, therefore, that the later issues of the Shakespeare Folio were sold, not by Robert Allot in 1632, but by his successors sometime between 1636 and 1641.

Within that period occur four significant events, any one of which might have provided a suitable occasion for reissue: (1) an inventory of Allot's effects subsequent to his death in 1635; (2) a transfer of the stock by his widow Mary to Legatt and Crooke on the first of July, 1637; (3) the gradual accumulation of unsold copies returned to the new proprietors upon the demise of Allot's original collaborators, all of whom had died between 1636 and 1641;[20] and (4) still another inventory, in 1641, upon the death of Thomas Cotes, the printer of the original issue and part owner of the stock. Of these the most plausible circumstance is the last. Under any other condition we would expect to find issues with Legatt's imprint and Crooke's name as publisher. But since these do not appear, and since, in any event, neither Legatt nor Crooke was in possession of the paper, engraving, or ornaments used for these issues,[21] their origin may be traced


88

Page 88
to certain untitled lots originally reserved for Cotes and now brought forth upon their discovery in 1641. At this time, as stipulated by the will, Thomas's younger brother Richard assumed possession of the stock, the printing shop, and all implements therein.[22]

That Richard should have failed to enter his name on the titles is not surprising, for of all the established printers of the day his record is distinguished by a penchant for anonymity. So far as I can discover, within the STC period, his name appears alone as printer on only two books, both undated.[23] Only once is he associated with a printer other than Thomas, and on the occasions when he is entered with his brother he usually withdraws his name from subsequent editions.[24] Books in which he had a considerable interest, such as the nine owned jointly with his brother[25] or the nine owned exclusively by himself,[26] appear invariably with the name of Thomas alone, never his own. Even for the Second Folio, an enterprise in which his investment was twice that of his brother's,[27] comparable to Allot's, and equivalent to the total advanced by the minor shareholders, the colophon and imprints register the names of all but the self-effacing Richard Cotes. Apparently, if this practice has any significance, the greater his responsibility and—we may presume—the greater his share of the presswork, the less his inclination to acknowledge it. Much of what is nominally under the imprint of Thomas may, then, be justly attributed to his silent partner.

Of all the books to which I have alluded, only one bears any particular relationship to the later issues of the Folio, and in every respect the correlation is complete, exact, and irrefutable. This, we should now recall, is the Theatrum Botanicum (19302), a volume which contains in some copies the rare and elusive watermark also found in these issues. Here then is the crucial exhibit: a book published within a year of the inventory of Thomas's effects, with paper and ornaments like those in the sheet applied


Illustration 1

Page Illustration 1

Illustration 2

Page Illustration 2

Illustration 3

Page Illustration 3

Illustration 4

Page Illustration 4

89

Page 89
to the remainders presumably discovered in that appraisal, and with an imprint assigning the presswork to one man when it was actually the property and almost certainly the work of another. Faced with this series of coincidences, I find it impossible to avoid the conclusion that the title-pages for these issues, now reading "Printed by Tho. Cotes, for Robert Allot . . . 1632",[28] should be understood as "Printed by and for Richard Cotes . . . 1641 or later."

Thus, at last, the esteemed "first" and "second" issues—the erstwhile "Allots 1—3," so enshrined by the earlier bibliographers and so avidly pursued by collectors everywhere—are now revealed to be, in all the light that can be shed upon them, nothing more than sweepings from the warehouse floor. Whether these scraps are entitled to remain within the pale of the Short-Title Catalogue is not for me to decide. I am content to present here the circumstantial evidence and, later, a proper motive for Richard's action in reissuing the book in the manner described.[29]

II. The Order of Variants in the First Issue

Whatever his doubts and reservations concerning the hypothesis just presented, the reader will readily agree that for the issue now identified as the first the present alphabetical arrangement of the imprints may not necessarily correspond to their chronological order through the press. That is determined, we suspect, by the convenience of the compositor. Normally, as his interest, like that of us all, is in doing as much as possible with the least amount of effort, he may be expected to plan his work so that it requires a minimum of labor to effect the alterations he must perform. Both within and between the several states of the title-page we should therefore look, not for drastic changes back and forth in aimless manner, but for the slight and deliberate adjustments which link all variants in a certain progression.

As illustrated in the accompanying plate, the letterpress for the title is disposed in three groups: (1) the heading, consisting of the seven lines of type above the area later to be filled by the portrait; and, below this area,


90

Page 90
(2) the place, comprising the single word LONDON, and (3) the imprint consisting of the two lines which relate to printer, publisher, address, and date. Group (1), of the same type throughout the impression, is, with one exception, fixed in its position; (2) is also, generally, of the same type, but occasionally moved as a consequence of the adjustments immediately below it; (3) is of two distinct settings of type, the position of which differs from state to state and, in one instance, within a state. Much of the variation in (3) is, of course, explained by the constant revision required to supply each of the collaborators in this edition with a distinctive imprint. Of the two settings employed for this purpose one comprises the imprint states Allot a, Allot b (formerly designated as Allot 5 and 4, respectively), Aspley d, and Meighen g; the other, Smethwick c, Hawkins e, and Hawkins f.[30]

It will now be perceived that in any determination of the order we must contend with two kinds of alterations, hereafter arbitrarily distinguished as states and variants. The term state is used to denote a deliberate alteration in type, whether performed, supposedly, on press (b, f) or off press (c—e, g); the term variant to denote shifts within states in the position of the type, whether these appear as intentional adjustments on press (a2-3) or as accidental results off press (c2). The minimum deviation, laterally or vertically, which distinguishes a press variant has been fixed at two millimeters. Any discrepancy less than this is dismissed as a distortion caused by shrinkage in the paper.[31]

Since two settings of the imprint are employed within the same impression it is obvious that the relation between them can be decided only by reference to type elsewhere in the forme which both have in common. This other type is unchanged except for a few copies of a and all copies of g. In the exemplars of what I believe to be the first variant, a state, the interval between the heading and the place[32] measures 206 mm., an interval


91

Page 91
soon reduced and thereafter maintained at 204 mm. in all succeeding variants and states. Quite possibly this was a deliberate adjustment to bring the several groups of type in closer proximity to the portrait. In g we observe several alterations—or, possibly, substitutions—in the two Ns of the word LONDON. For the earlier states the lower right kern of the first N is bent slightly upward and the upper left kern of the second curled sharply downward. For this, however, the one kern is straight and the other only slightly curled. Another difference is evident immediately behind the descending stem of the second N where, in all but the g state, there appears a small speck. From these trivial discrepancies we may tentatively conclude that a is the first state, g the last, and that all other states including those represented by the other setting are interposed between them. Before considering the exact position of this other setting within the sequence, it will be necessary to prove the order within each setting.

Further scrutiny of copies in the earliest state of the first setting—and of the entire sequence—discloses certain other refinements, all intended, I believe, to improve the appearance of the page. Shortly after the imprint was adjusted to 204 mm. it was noticed that SHAKESPEARES, the longest and most prominent word in the heading, though extending beyond the portrait (191x160 mm.), was several millimeters short of the margin established by the plate mark (200x169 mm.).[33] To remedy this the forme was again unlocked and the word spaced out from 166 to 169 millimeters. The combination of adjustments thus produces three distinct variants within the state:[34]

Heading Interval

     
1. 166 mm.  206 mm. 
2. 166 mm.  204 mm. 
3. 169 mm.  204 mm. 
Like the earlier gradation, the later one, once arranged, persists in all remaining states.

Preceding these slight and delicate adjustments is another which is grossly botched. Instead of replacing the foul case "f" in "ſold" either at the striking of proofs, or when the imprint was moved, or later, in state b, the pressman attempted to gouge out the offending projection, but succeeded


92

Page 92
only in bending it up and into the shoulder of the piece, where it gradually collected a "pick" and continued to clog the face of the type[35] until it was eventually removed at the end of work on the second state.

While the later state b is strangely defective in this one respect, it does represent a further effort to enhance the appearance of the impression, notably in the substitution of ordinary type for the heavy-faced word "Printed". This too, we observe, is a correction retained in the later states of the imprint. Another alteration, the deletion of "at his ſhop", was arranged for a reason which will become evident at a later point in our discussion. Suffice it to say now that the order devised by Smith as b-a (Allot 4-5) disrupts the progression of his own sequence (where Allot 4, with the phrase omitted, intervenes between 3 and 5), disregards the actual relation between b and the states which succeed it (d and g, both without the phrase), and ignores the indifferent practice observed in other books printed for Allot at this time. In these the imprint may be without address,[36] with the simple reference "at the Blacke Beare,"[37] with the preliminary notation (as in b) "at the ſigne of the Blacke Beare,"[38] or with the full address (as in a) "at his ſhop at the ſigne of the Blacke Beare."[39] If the publisher's address could be so variously rendered in these imprints, and his name so variously spelled as Allott, Allot, or Allet, it would appear that such decisions were not at his discretion but entirely at the convenience of the printer. And this convenience, as we shall have occasion to remark, was very directly involved in the alteration. Hence, I repeat, the order is probably in the direction of a-b, and the inclusion of the phrase in the later issues II-III is to be simply explained by the use of a copytext in its earlier state.

The third imprint in this setting is obviously d, the Aspley state. Though this retains most of the type used for a and b (including the terminal phrases "at the ſigne" and "in Pauls Church-yard") the several revisions within the lines doubtless necessitated its preparation off press. Again as in b (and occasionally in a) the type is clogged in several places and the "n" of "in" is above the line. Among the idiosyncrasies not evident in b but later transmitted to g are the substitution of "ſ" for the previously mangled "f" in "ſold", a depression in the "e" and "t" of "be" and


93

Page 93
"at", and a battered "e" in "the". Peculiar to d alone, apparently, is the fact that in later copies of the impression in this state the title is a cancel.[40] What this signifies, I suspect, is an overrun on the lot of paper which had been reserved for all states through d, and the consequent necessity for the excision of the titles in excess of order and the impression of certain Aspley imprints by half-sheet imposition.

In Meighen g, the final imprint in this setting—and in the sequence—the modifications are more extensive. Besides the usual substitution in the name, these include the two Ns previously mentioned, on the next line, the use of "middle" for "ſigne", the word employed in all preceding variants, and in the last line an entirely new reading. Other than these the Meighen state has only one distinction, a broken comma after Cotes, doubtless caused by the constant shifting of the type below it in the preparation of the earlier states.

For Smethwick and Hawkins, the several states comprising the second setting of the imprint, the compositor (or the overseer) again grouped together addresses which would necessitate only slight revision between impressions. As the other imprints had, generally, certain readings in common, so these share in the reading "at his ſhop". At some time during the printing of the earlier, Smethwick state the press was stopped, the type removed from the forme, and then returned to a position in which both LONDON and the imprint were three millimeters to the left (or right on the printed page) of the space they had formerly occupied. Very probably this interruption was of some duration—overnight, at least—for while the first run was on a heap of paper turned in one direction, the second is on a heap turned in the other. Both type and paper, then, confirm the existence of two variants:

     
Position of LONDON (as printed)  Watermark[41]  
c 63 mm. to right of margin  In Effigies leaf [A5] 
66 mm. to right of margin  In title leaf [A2] 

94

Page 94
The "margin" to which I refer is a hypothetical line abutting the left edge of the word SHAKESPEARES and extending perpendicularly from that word.

Next and last is the Hawkins imprint, with LONDON disposed as in c2, but with the beginning of the final line reading first (e) "Chancery Lane", and then, as usual in Hawkins' address[42] (f), "in Chancery Lane". For the other instance of a correction at press (b) the deletion of three words required a readjustment of the imprint. For this insignificant addition, however, the compositor was content to leave the line slightly off-center.

At this juncture I must confess that while I have, to the best of my ability, defined and ordered all the points that appear on the title-page of the first issue, I have established a sequence which pertains only to the order of composition within each setting, and not to the order of variants through the press. Were the imprint of a single setting the order for both would be, of course, the same. But since there are two, a question arises concerning the necessity for the second. If eleven words are, as we see, applicable for all states, why does the compositor go to the trouble of preparing another setting?

The only plausible explanation is that the provision of this other setting enables two men to work efficiently and simultaneously at their separate tasks. While the compositor is revising and adjusting the imprint in one slug the pressman can work off copies on the other. Thereafter, as each finishes his assignment, the slugs may be transposed, and both can then continue to revise and print as before. The obvious advantage of the system must have immediately recommended it to the overseer, who doubtless classified the imprints in the two groups just described, ordered a separate setting for each, and arranged a schedule for their regular alternation through the press.

To reconstruct this schedule we may now recall that if a and g have certain distinctions not common to the other states, then these two are at the extremes and the others alternately disposed between them.

illustration

As ordered, this sequence is justified not only by the peculiarities in the terminal states, or by the progression within each setting, but also by certain links between the settings. In some copies of b [43] and all copies of c1, for example, LONDON is displaced three millimeters to the right (or left


95

Page 95
on the printed sheet) of the position retained in the other states. Again, in some copies of c2 and all copies of d the exemplars share an odd lot of paper (H1420)[44] not evident elsewhere. With c thus established as intermediate between b and d the inference follows that e-f is similarly intermediate between d and g, and that all states were therefore imposed in the order now assigned.[44a]

III. The Size and Distribution of Issues

Through various discoveries and acquisitions in the years since Smith's census the record of extant Second Folios in perfect or impaired condition[45] may now be extended to include some 196 copies, or approximately one-fifth of the number estimated in the original edition. The existing copies constitute a rather large percentage, large enough, I believe, to provide a reliable index of the proportions among the states at the time of publication. I therefore offer, in Table II, several calculations, one projecting the ratio fixed by the extant copies and, as a check against this, another indicating a hypothetical distribution based on the investment of those participating in the edition.

    Table II

  • Allocation on Evidence of Known Copies
  • Number of perfect exemplars: 163
  • Percent of 1000 copies: 16.3
  • Allocation on Basis of Copyright
  • Number of plays: 36
  • Copies per share 36/1000 or 27.8

                     
&c.mmat; 1000  Copies  &c.mmat; 1000 
Variant imprint  Copies  Copies  owned  Copies 
c Smethwick  16  102.0  111.2 
d Aspley  43.8  55.6 
ef Hawkins  56.3  27.8 
g Meighen  37.5  27.8 
"Cotes & Allot"  [125]  [791.4]  [28]  [778.4] 
Subtotals: a   31  203.8 
b   73  456.3 
II  14  87.5 
III  43.8 


96

Page 96

The remarkable correlation between these two sets of figures, each derived from a different premise, undoubtedly brings us very close to what must have been the actual distribution. Only in Hawkins is there an appreciable discrepancy, and this, I believe, can be compromised by reference to the property held by the minor shareholders. Any evaluation, whether of plays or of real estate, involves a consideration not directly related to the size of the piece. Hawkins had only one play, true, but this was Othello, a possession certainly equivalent in value to both of the plays owned by Aspley (Much Ado and 1 Henry IV) and twice the value of the single play owned by Meighen (Merry Wives). Between these two figures, then, the one which measures only the size of the property should very probably be ignored in deference to the one which evidently signifies its importance.

With this much rationalized (at least to my own satisfaction) we may confidently reconcile the slight disparity in the estimates for all states by adjusting the figures to the practical limits determined by the quires of paper, i.e., to the nearest multiple of 24.

       
Ia   b   c   d   ef   g   II  III  Total  
Quires  19  42 
Issue  192  456  96  48  48  24  96  48  1008 
Total issue  864  96  48  1008 
This tabulation serves to enforce earlier considerations and leads to others still unresolved. For one, it will be observed that the order of imposition in the original issue, as previously determined, corresponds both to the extent of the copyright and hence to the size of the allotment for each of the proprietors. First off the press were the 27 quires of title-sheets for Allot and Cotes, then 4 for Smethwick, 2 each for Aspley and Hawkins, and finally a single quire for Meighen. For another, it should be noted that the total for all issues is within one quire of what the stationer calls a "perfect bundle,"[46] the most convenient set-out for an edition of this size. Both in whole and in part the figures therefore conform to the general estimate of the extent of the edition and of its various states.

The matters unresolved and deliberately evaded until now relate to the occasion for the alteration in the Cotes-Allot imprint a-b and the reason


97

Page 97
for the excessive delay in the production of the later issues II-III. Since neither of the imprints in the first issue can be regarded as a correction of the other, we may surmise that the two were used so that the warehouseman could readily distinguish the lots to be sent forward to Allot from those to be retained by Cotes. In the preparation of the edition both of these partners, like Jaggard and Blount before them, were acting in behalf of certain unnamed associates as well as for themselves. The 28 "copies" assigned to both by imprint or, rather, the stock represented by these copies, was therefore not entirely theirs by right but by purchase from owners who had relinquished any interest in the sale of their property. Table III identifies the owners of these copies, the proportionate share of their investment, and, for comparative purposes, certain other figures from our earlier estimate.

Table III

                         
Entitlement  Represented by Cotes-Allot 
Owner[47]   Interest  Plays 
[multiples of 24]  Titles in state 
Robert Allot (½)}  As indicated,  {8  192  Ia (192 copies) 
Thomas Cotes (¼)}  part ownership  {4}  96} 
Richard Cotes (¼)}  in 16 plays.  {4} 
Richard Cotes  exclusive title  4}  192}  Ib (456 copies) 
Thomas Hayes  " "  24} 
Nathaniel Butter  " "  24} 
[Surplus: appropriated by Cotes brothers?]  120} 
Unassigned, presumably derelict  144  II,III (144 copies) 
---  --- 
Total:  792  792 
Since Allot was entitled only to 192 copies, the equivalent of those estimated in state a, it would seem that this imprint, describing the place of sale as "at his ſhop", is the one that identifies his consignment. Apparently, then, the lots with the abbreviated imprint b, those not sold 'at Allot's shop', were reserved for the Cotes brothers and their unnamed associates. Included in these, we note, are the 336 copies belonging to them by right or purchase and, as our speculations lead us to suppose, 120 others over

98

Page 98
and above their pro-rata share. Possibly, if there was any such surplus, some or all of it may have been baled up at the Cotes warehouse and dispatched to booksellers outside London for sale in the provinces.[48]

Besides these, however, there was still another lot which they could neither publicly claim nor privately appropriate for themselves—the 144 copies subject to seizure by the unknown owners of the six derelict plays. This group was discreetly laid aside, without titles, until such time as the claimants might identify themselves and secure their proper share with the imprint they might desire. For nine years or more the brothers Cotes awaited the heirs of the registered owners. Finally, after the inventory of 1641, Richard Cotes assumed the rights which others had allowed to lapse, and provided the remainders with new imprints copied from a sheet of the original issue.[49] These then appeared at the bookstalls with the titles identified as II and III.

Richard's motive in disguising these two issues was, we may believe, very prudent, if not entirely commendable. Quite apart from his customary habit of operating under his brother's name, there was now another and more compelling reason, that of avoiding a confiscation of all remainders under the Ordinance of 1641. This specifically prohibited the printing, and hence the sale, of any book without the express consent of the owners.[50] But who were the owners of these six plays? Whoever they were, Cotes must have reasoned, their failure to contest the issue of 1632 amounted to an abandonment of the property to which they were then entitled. What was apparent to Cotes, however, as he doubtless realised, might not be apparent in the eyes of the law. Hence, rather than invite litigation, he passed these remnants off as the original issue and stood ready to defend them as part of an edition published long before the Ordinance was in effect. Even in his most sanguine moments, though, he could hardly have anticipated the ruling of Shakespeare's bibliographers that the special sheets were to be regarded, not only as coeval with the original issue, but as a series of states preceding it.


99

Page 99

IV. The Effigies Leaf

In this and the following section I approach two considerations, one of such importance that a single commentary on the subject attained a circulation, if we may believe the author, of some 20,000,000 copies,[51] the other of equal importance but generally disregarded, both intimately related to all that has gone before, and both easily dispatched (the reader will be glad to know) on the warrant of evidence already adduced.

The first of these pertains to leaf A5, the recto of which bears two commendatory verses on Shakespeare, one by an unknown author "Vpon the Effigies of my worthy Friend," the other Milton's earliest printed English poem, titled as "An Epitaph on the admirable Dramaticke Poet". Since this page is conjugate to the title and, like its mate, appears in three different settings, the established order for the one leaf enforces a similar order for the other. Without further ado I therefore present, in Table IV, an account of the textual and typographical variation among the issues.

Table IV

                                           
Issue  Ib   II  III 
[Smith]  [C]  [A]  [B] 
Line 
VV 
["S" initial 1]  [2]  [3] 
Comicke   Comicke   Comick  
paſsions [ſs ligatured]  paſsions [ſs separate]  paſſions [ſſ separate] 
11  Shake-ſpeare  Shakeſpeare  Shake-ſpeare 
13a  Poet,  Poet  Poet, 
13b  VV 
13c  Shakespeare  Shakespeare  Shakeſpeare  
14a  ["W" initial 1]  [3]  [1] 
14b  honour'd   honour'd   honor'd  
17a  Vnder   Under   Vnder  
17b  -ypointing   -ypointed   -ypointing  
18  Fame  Fame   Fame 
22a  whil'st [st ligatured]  whil'st, [st ligatured]  whil'ſt [ſt ligatured] 
22b  -endevouring   -endevouring   -endeavouring  
22c  Art   Art,   Art  
25a  Impreſsion   Impreſsion   impreſsion  
25b  tooke   tooke:   tooke  
28  lie   lie,   lie  


100

Page 100

From this it will be immediately observed that, while Smith's representation of "each succeeding version [in the order A-B-C as] being printed from its immediate predecessor" is manifestly untrue, the statement is just as inapplicable for the order I-II-III.[52] But where Smith was forced to this contention by his arbitrary arrangement of the title-pages, no such obligation rests upon us. Though the order for title and Effigies page is, we reaffirm, identical, the copytext furnished the two compositors need not always be the same. For II both followed I, the only available text. For III, on the other hand, two different copies were passed to the compositors, a sheet of II to the man setting the title, a sheet of I again to the man setting the Effigies page. Thus the text descends directly for one page and collaterally for the other.[53]

To clarify the relationship I offer four stemma, the first illustrating my own explanation, the second conforming to Smith's, the third and fourth exemplifying others which might be advanced.

illustration

The difficulty with (2), we note, is the absence of any textual link between Title III and I or between Effigies II and III; with (3) and (4) the presence of an eclectic text in Effigies I; with all three the assertion of priority to issues containing paper and ornaments not found before 1637. Aside from all other considerations, these impediments alone are insurmountable.

A comparison of Effigies Ib [54] with the page I have reproduced now


101

Page 101
leads to another discovery,[55] another predicament, and another solution. The facsimile provided here is of an earlier state a in which the readings for the Effigies poem differ slightly from the usual version.      
Poem line: 3 
Ia Commicke   Laughe riſe  
Ib Comicke   Laugh riſe,  
In the correction of this page the compositor dutifully moved the single word after Comicke to close the gap created by the excised m, but neglected to justify the deletion in Laugh, perhaps because there were seven words in this line. Thus between the comma and the following word an en space appears as witness to his indolence. In a the peculiar readings are such that they cannot intervene between others (e.g., between III and Ib in Smith's sequence) but must precede those in all three issues.[56]

The predicament arises when we attempt to correlate state a of the Effigies page with state a of the title for this issue. The order in one, presumably, should correspond to the order of the other. Except for a chance conjunction in a made-up copy,[57] however, this page occurs only with titles in state b.[58] If a single pressman handled both of these formes in succession then, despite all the evidence to the contrary, it would be necessary to reverse the order in one or the other sequence. But since the Cotes establishment is known to have had at least two presses,[59] we may adhere to the evidence for both states and argue that the formes were machined


102

Page 102
simultaneously. In this manner each man would receive half of the sheets for the issue, and as these were worked off the two heaps would then be turned over and exchanged for the reiteration. The man printing the titles would therefore begin to perfect his companion's heap (topped, presumably, by the several sheets with Effigies in state a) at a point midway through his own operation. And that point, as a reference to our schedule will show,[60] occurs just after he has printed 240 copies of b and while he has 216 yet to go. Again, it would seem, apparently irreconcilable facts combine to provide an almost certain conclusion. Indeed, the case is so neat in this instance that I venture to predict that state a of the Effigies leaf will be found conjugate only with state b of the title.

V. Historical Associations: The Charles I Copy

In his account of certain notable copies of the Second Folio Professor Smith hurries over what was, for him, a relatively unimportant exemplar of "Allot 5" to humble himself before two others which he had exalted in the rank of "Allot 1". One of these was a copy first owned by King George III, then given to Dr. Burney when the latter withdrew his bid for the Allot 5 in question, eventually acquired by James Lenox, the American collector, and finally deposited at the New York Public Library. The other, once in the library of Thomas Howard, second Earl of Arundel, was donated to the Royal Society and thence passed through Sotheby's into oblivion. Both of these "treasures," as Smith calls them, were at various times offered to the British Museum, and both were as many times refused. Such persistent evidence of disrespect, though excused by Smith as occurring before the enlightenment of his "scientific bibliography," was nevertheless deplored on the grounds that the Museum had thereby denied itself the opportunity of acquiring copies immeasurably superior to those already in its collection.

A reversal in the order of variants puts an end to this solicitude as quite ill-advised and allows an entirely different interpretation. Where the Museum authorities had once been admonished they are now to be congratulated, for with a sagacity not given to Shakespeare's professed bibliographers they have been careful to accept only the states proved to be the first—an a, two bs, and a c—and to reject on every occasion the unsightly specimens now demonstrated to be of subsequent issue. The same discrimination appears in the selections made for the other principal British libraries: all possess one or more copies of a or b, only a few the


103

Page 103
copies identified as of issue II (and these, no doubt, only at the insistence of misguided benefactors), and none the copies described as of issue III. Most of II and all of III were cast aside as the scraps to be gobbled up in the American market.

Another consequence of this reversal is that the "Allot 5" Smith had slighted many years ago finally assumes its rightful position, not only as a, the first in the series of states in the initial issue, but as the only known exemplar with a provenance extending from the very date of publication to the present time. This is the book formerly in the possession of King Charles I and now, after many vicissitudes, in repose with other remarkable volumes in the Royal Collection at Windsor Castle. For the privilege of offering a summary account of this historic copy and a record of the inscriptions which it bears I am deeply indebted to Sir Owen Morshead, Librarian to Her Majesty, Queen Elizabeth II.

As a representative of the a state this book, along with others similarly distinguished, was separated from the b group retained by the printers and dispatched first to Robert Allot, the chief proprietor. Some time before the day of publication Allot honored the custom[61] of forwarding the selected copy to the King's bindery, where it was suitably covered in morocco and gilt and then presented with a ceremony befitting the occasion. As even Milton was to admit, the King soon became a confirmed admirer of Shakespeare, no doubt as the direct result of this presentation, and in the enforced seclusion of his latter days often resorted to the book as "the closest companion of these his solitudes."[62]

Toward the end of his stay at St. James's Palace Charles entrusted to Sir Thomas Herbert, his retainer, all the possessions which were then his to bestow—his books, his cabinet, his cloak, and his large silver watch[63]— and all of these Herbert faithfully preserved until his own death in 1682, when they were dispersed and, with the exception of the Folio, subsequently lost from view.

An inscription in the copy testifies that it next appeared at the auction of Dr. Richard Mead's library in 1754, where it was purchased by Mead's friend and disciple Dr. Anthony Askew. Upon the death of Askew it was again placed on sale and bought by George Steevens at what he justly considered to be an "enormous price," unequalled at this time,[64] but soon to


104

Page 104
reach over thrice that amount when it was eventually purchased in 1800, at the Steevens sale, by a representative of George III. Not until the bidding had gone to £18-18-0 was it made known that the King desired the book, at which disclosure the participants, Charles Burney among them, immediately withdrew their offers. And so, after its strange and varied adventures, this precious volume was returned to the Royal Library, there to remain until today.

Table V

Notations in the Charles I Copy

Numbers indicate order of inscriptions, letters their position as R right, C center, L left.

    Notation

  • I. Recto of end-paper
    • 1C C/L
    • 2C Aut Æternum | Aut Nihil
    • 2R αα | T H
    • 3L 2.54
  • Comment
    All notations apparently in the same hand. alpha alpha | Thomas Herbert? Shelf mark?

II. Recto of flyleaf between end-paper and A1

       
1C   Dum spiro Spero | CR  By Charles I 
2C   347 Shakespeare's Works, bound in Morocco leather, gilt leaves, | 1632 | In this Book, is the writing of King Charles the First in | these Words; Dum Spiro Spero, C. R. also in Mr. Her- | bert's Hand; ex Dono serenissinti Regis Car. servo suo | Humiliss. T. Herbert   Clipping from Askew catalogue 
3C   Bought at the Sale of Dr Antony Askew | Feby. 14th. 1775, at the enormous price of | Five pounds Ten shillings | George Steevens. | It appears that Dr. Askew purchased this | book at Dr Mead's Sale for £2..12..6.  Note by Steevens 
4R   Ex dono serenissi. Regis. Car. | servo suo humiliss°. | THerberto.  Herbert's acknowledgement: Ex dono serenissimi Regis Caroli servo suo humillimo. 

105

Page 105
     
5R   1 {Sr. Tho. Herbert was Master of | the Revels to K. Charles. I.  Note by Steevens. Parentheses and numerals in another hand. 
6R   2{This is a mistake he having been | Groom of the Bedchamber to K. Charles I.| but Sir Henry Herbert was Master of the | Revels  Correction by George III 
7C   The two lines marked No. 1. were written | by Mr. Stevens, at whose Sale this Book was | purchased; and the Correction marked | No. 2. was made and written by his late | Majesty George III. | JH. Glover. Fred: Aug: Barnard | Librarian.  Note in the handwriting of John Hulbert Glover, assistant and successor to Frederick Augusta Barnard, the librarian to George III. 

III. Title-page, A2r, reproduced in this study. Notation, by Herbert, is "Pawb yn eu Aruer," the motto of one branch of the Herbert family (teste L. W. Hanson).

IV. A Catalogue of all the Comedies . . ., * 4v. Marginalia indicating page numbers for the several plays and, opposite those in which they appear, the names of the following characters: "Bennedik & Betrice", "Piramus & Thisby", "Rosalinde", "Mr Paroles", and "Malvolio". It has been thought, and so reported, that as these are the only references in an unknown hand they might be in the script of Ben Jonson; but Sir Owen Morshead informs me, on the authority of Dr. Percy Simpson, that this is not the case.

[Since writing this I have received, through Sir Owen's good offices, a communication supplied by J. F. Kermode of the University of Reading to the effect that these inscriptions may also be in the hand of King Charles. See Farmer's note as corrected in Steevens' edition of Shakespeare (1778), IV, 283-284, and Warton's note as corrected in Todd's edition of Milton (1842), IV, 371.]

VI. Tabulation of Variants

A typical designation Ic2 indicates 1st issue, 3rd state in order of imposition, 2d variant. Measurements, all in millimeters, are of (1) linear width of word SHAKESPEARES, (2) interval between heading and LONDON, (3) interval between left "margin" and LONDON, (4) interval between chainlines in leaf conjugate to the one containing watermark, (5) vertical measure of the mark.

FIRST ISSUE: Printed in 1632, state a for Robert Allot, state b for Thomas Cotes and others not identified by imprint.


106

Page 106

illustration

SECOND ISSUE: Remainders supplied with sheet printed ca. 1641 by and for Richard Cotes.

illustration

THIRD ISSUE: Final lot of remainders, ca. 1641 or later, again by and for Richard Cotes.

illustration


107

Page 107

VII. Register of Copies

1. Following the reference I list, within parentheses, the total number of copies known to me of this variant, and any title or imprint reproduction in the Church Catalogue (c), Smith's monograph (s), or the present study (t). Item numbers in the Church Catalogue identify copies now at the Huntington Library.

2. As a matter of convenience, all copies reported as having a genuine title, whether or not this is accompanied by an Effigies leaf, are classified as "perfect". The sigil τ introduces the list of those I have personally examined, ‡ those for which I have an adequate description, and * certain others not sufficiently described or of questionable status as to variant or location. In the third group I have indicated by a number within brackets all exemplars reported in Smith's list as then (1928) in the possession of booksellers or private owners. In all groups errors and omissions are to be expected, and notations of these will be gratefully received.

3. Symbols are taken from the Short-Title Catalogue and Library of Congress index. Figures within parentheses designate the number of copies on deposit. DFo figures refer to copy number.

  • Ia1 (3) sCLU-C τNNP; ‡CLU-C, MiU.
  • Ia2 (3) ‡CSmH, PBm, MWelC.
  • Ia3 (25) tRoyal Library tDFo 48 τDFo 1, 48; NN (1 Reserve, 1 Berg); ‡Royal Library, MdBWA, MiD, PU-F, Pforzheimer; * G2, L, L18(2), M, O (2), Birmingham Public Lib, DFo 16, [7].
  • Ib (73) sPBL tDFo 34 τDFo 3, 8, 14, 21, 27, 30, 31, 34, 35, 47, 44; NN (4 Reserve, 1 Berg), NNP; ‡CLU-C, CSmH, CtY (4), DFo 22, 24, DLC, IU (2), MdBJ-G, MeB, NjP, PBL, RPJCB, TxU (2), ViU, Charleston (S.C.) Lib. Soc., Grolier Club (NY), Toledo Museum of Art; * C2 Eton, E2, G2, L (2), L2, L18 SH (2), Leamington Public Lib., Metheun facsimile; MB (2), Ohio Wesleyan, [19].
  • Ic1 (11) c615 sNN τDFo 7, 40, 43; NN; ‡CSmH (1 Chew, 1 Church), Pforzheimer; * L, DFo 4,20,[1].
  • Ic2 (5) tDFo 52 τDFo 52, MH; ‡CLU-C (2), MWiW-C. CSmH (Wilson) copy in facsimile?

  • 108

    Page 108
  • Id (7) c612 sNN tCLU-C τNN; ‡CLU-C, CSmH, NBuG; * DFo 23, [2].
  • Ie (2) tDFo 11 τDFo 11, NNP.
  • If (7) c613 sNN tDFo 41 τDFo 41, NN; ‡CLU-C, CSmH, DFo 5, Players Club (NY); * [1].
  • Ig (6) c614sCSmH tDFo 46 τDFo 46, NN, NNP; ‡CSmH, DFo 2; * [1].
  • II (14) c611sCSmH tDFo 36 τDFo, 9, 10, 36, 42; NN; ‡CSmH, DFo 6, PP; * C2, D, L12, O2, [2].
  • IIIa (1) sNN τNN.
  • IIIb (6) sNNC tInI τDFo 15, NNC; ‡InI; * DFo 38, 56, [1].
  • Defective copies: DFo 12, 13, 17, 18, 19, 25, 26, 28, 29, 32, 33, 39, 45, 47, 49, 50, 51, 53, 54, 55, 57; ICN, and 11 reported by Smith.

Notes

 
[1]

"The Variant Issues of Shakespeare's Second Folio and Milton's First Published English Poem: A Bibliographical Problem," Lehigh University Publications, II, No. 3 (1928). The death of Professor Smith on January 15 of this year brings to an end in time, but certainly not in memory, an association extending from my early days as an undergraduate. As it was he who first introduced me to the ways of scholarship and thereafter encouraged my interest in bibliography, I am confident that the present expression of this interest would have received his hearty approval.

[2]

With this disavowal I now assume responsibility for all errors and omissions in the presentation of my case. In the gathering of the necessary evidence, though, I gratefully acknowledge the interest of the Modern Language Association Committee on Research Activities for a grant in aid of my investigation, the helpful advice of Dr. Allan H. Stevenson on the variant paper in the Folio—a matter of crucial importance to my argument—and the generous assistance of curators throughout the country in reporting the peculiarities of the copies in their custody. To Dr. Giles E. Dawson and Dr. James G. McManaway of the Folger Library, Mr. H. Richard Archer of the Clark Library, and Mr. Herman R. Mead of the Huntington Library I am under especial obligation for the time and energy they have devoted in my behalf.

[3]

Op. cit., p. 13.

[4]

This is the total of sorts in the title page (247) and the Effigies page (993), each twice distributed and twice re-composed.

[5]

Op. cit., p. 13.

[6]

Op. cit., pp. 20, 47.

[7]

As this sheet would require laborious composition (in the selection, spacing, and centering of type in different fonts) and a considerable amount of presswork (three operations, one for the inner forme [Effigies], one for the outer [title], and one again for the outer [portrait]), the total cost would be far in excess of that for any other.

[8]

Not now, but pertinent to a later stage of our discussion is another question: Why print titles for stock to which, in fact, they were not entitled?

[9]

As this contention has not been justified in the present description of the Fourth Folio, I ask the reader's indulgence until such time as I may offer my own account.

[10]

Since the Archer headpiece and similar blocks from the same matrix were commonly used by Cotes and others throughout this period, I disregard its appearance here. It has been noted in such books as STC 4510, 14813, 19302, 19654, and 22273.

[11]

In A6r, adjoining Effigies leaf A5r, both of these ornaments were used again, a clear indication that they were in the hands of the compositor as he prepared first one and then the other forme for the press. Elsewhere the "S" occurs at D2r and d5v, the "W" at alr and 2p2v, and other letters of the same kind throughout the book.

[12]

So in my control copy DFo35. Heawood 1420 is a small fleur-de-lis, always in paper with chainlines approximately 23 mm. apart, and evidently existing in at least three different states. According to Dr. Stevenson's method (see fn. 15), I describe these as A 53x25mm. [7(13:12)10]—in DFo copy 2i3; B 54x25mm. [10(12:13)8]— in 2m1, 204-5-6, 2r3, 2y1; C 57x27mm. [8(15:12) 10]—in 2y4, 3a6, 3b1, 3b3, 3d1-2. Heawood 1731 is a large shield, usually indistinct, and measuring vertically either 137 or 141x54 mm. It appears in DFo copy at 3a5 and 3c2. So far as I have observed, H1420 is used as the title-effigies sheet only for some copies of the Smethwick-Aspley imprints, H1731 for all others. See further, fns. 40, 41.

[13]

Reference "ab" indicates two variants within the same word. Not cited are the unique variants "Coppies" in Allot 1, "according" in Allot 2, and the omitted phrase "at his ſhop" in Allot 4.

[14]

As stated before (fn. 12) H1420 is found only incidentally in this issue.

[15]

The sheet in InI copy of this state is not H594, but apparently of another Guesdon mould in the style of H576 without initials or date. The mark is on paper with chain-lines 21 mm. apart and measures 90x38 mm., dimensions which vary widely from those for H594 (fn. 19), or for still another variant of H576 recently described by Allan H. Stevenson (Studies in Bibliography, IV [1951], 81). For an account of this copy and a sketch of the mark I am indebted to Miss Frances Stalker, Head of the Reference Department, Indianapolis Public Library.

[16]

See Giles E. Dawson's account of DFo38 in TLS, Feb. 1, 1947, p. 65. The copy is now in storage.

[17]

By Dr. Stevenson at the University of Chicago library, by myself at the Harvard and Folger libraries. Heawood had previously specified Britain as containing the paper.

[18]

The mark is on leaf 3K5 of the Chicago copy, not observed in Folger exemplar. It is also noteworthy that this volume (STC 19302) contains all but one of the ornaments present in the several settings of A5r of the Second Folio: the 1 "S" on four pages [3R5v, 4D1r, 6Y1v, 7A4r], 1 "W" on four [4F2v, 6H6v, 6M1r, 6S1r], 3 "S" on one [7A1r], and the peculiar 2 "S" factotum on no less than thirteen [B2v, B4v, C3v, D3r, G4r, H6r, 16r, 3Y2r, 4M2r, 4R2v, 5G2r, 5H6v, 6E6r].

[19]

Mark H594 is a small Guesdon shield, 70x63 mm. The states appearing in the Folger copies are these: A 18[3:22:20:18]3 OUT and B 16[4:22:20:17]5 IN. A is in Second Folio copy 36 [issue II] A5; copy 15 [issue III] A5; Camden [Latham imprint] a1, a2; Paris C1, D2, F4, G2, G3, H5, I1, and on. B is in Camden a3; Paris G6, H6, I2, I4, and on. Quite unnoticed by me, but quite apparent to Dr. Stevenson's gimlet eye is occasional evidence indicating the transfer and resewing of the mark in the A state.

[20]

Aldis and Plomer's Dictionaries record the dates as Hawkins c1636, Aspley and Smethwick 1640, and Meighen c1641. The return of any copies would, of course, be in satisfaction of liens against the estates of these associates. Copies so recovered might then be reissued with A2.5 intact or (see fn. 21) with the inappropriately imprinted A2 removed.

[21]

It is not inconceivable, though, that Legatt and Crooke may have acquired stock in the other ways described and remaindered this in the condition in which it was found. How else may we account for the extraordinary number of copies in defective condition? Altogether these comprise about sixteen per cent of all copies now extant, and practically every one lacks only the interior leaves A2 or A2.5. Surely the restorer is not to be held responsible for such an extensive and such a specific loss as this. More likely is the supposition that some copies, representing lots returned by Allot's associates (condition 3), were issued with the inapplicable titles deleted, and some others, representing lots discovered in the warehouse (1 and 2) were distributed in their originally imperfect condition, with both leaves absent. All we need do now is to differentiate the copies in these several categories from those whose deficiencies are of later date!

[22]

Plomer, op. cit., p. 53.

[23]

STC 15594, 16738.

[24]

17366-67, 24700-01, 25837-38. The other association is 19192.

[25]

267, 891, 900, 3174, 6724, 14704, 19389, 22274, 25789.

[26]

4694, 19248, 19302, 19846, 21068, 22339 [also in Greg's Bibliography, item 284f], 22502, 23541-42. One other (12017) appears without name in imprint.

[27]

Richard shared one-half interest with his brother in 16 plays and was the sole owner of 4 others: Titus Andronicus, Henry 5, and "Yorke and Lancaster" or 2-3 Henry 6. The Register entry for these (Nov. 8, 1630) also includes Pericles, a play not incorporated in the 2d Folio.

[28]

As Dr. Dawson has remarked, this in itself should be interpreted as "Printed by Thomas Cotes for himself and Robert Allot." Studies in Honor of A. H. R. Fair-child (U. of Missouri Studies, XXI, No. 1 [1946]), p. 19.

[29]

Since the date now assigned is not beyond dispute, even if supported by a motive, I urge those who remain unconvinced to search out copies of certain provenance or others with dated inscriptions. Of some interest in this connection would be a positive identification as to the issue of two copies for which the time of acquisition is known, i.e., one bought by Sir Thomas Barrington about 1637 and the other purchased for the Earl of Huntingdon in October 1638. [Francis R. Johnson, "Notes on English Retail Book-prices, 1550-1640," The Library, 5th ser., V (1950-51), 91-92.]

[30]

Hawkins imprint e, unrepresented in Smith's account, has since been described by Dawson in TLS, February 1, 1947, p. 65. Professor Smith did remark the identity in type between the Hawkins f and Smethwick imprints (p. 26), but as he was here, as elsewhere, bound by the traditional arrangement, failed to group the two together.

[31]

In view of McKerrow's warning against the reliability of any discrepancy amounting to less than 4 per cent, I should defend the validity of this precise distinction by noting, first, that the A2.5 sheet is fine quality crown paper, of uniform weight and consistency, second, that the differences appear only among a few copies in two of the seven states in this issue, third, that when they do appear they concern only a portion of the letterpress on the page, and fourth, that the deviation is generally corroborated by other evidence. For these several reasons I have no hesitation in citing such minutiae as indications of some disruption in the presswork.

[32]

Measured from the base of The ſecond Impreſsion (excluding descenders) to the top of the word LONDON. Only one copy (NN-Hawkins state) measures less than the common interval of 204 mm., but as the discrepancy is no more than 2 mm., and unconfirmed by any other difference, it has been disregarded.

[33]

It would seem that the engraving mangle, though carefully adjusted at first so as to center the impression between the two areas of type (see initial variant copies NNP and CLU-C [the latter reproduced in Smith, p. 23]), soon developed a "wobble" which scattered the impression at various angles in all succeeding variants and states.

[34]

The measurements apply to all copies seen and reported in this state except for CLU-C, described as 167-205, and CSmH, described as 166-203. These differences are so minute, however, that I have classified the copies as a1 and a2 respectively.

[35]

Especially toward the end of the run on state b. See the reproduction of the Lehigh University copy in Smith, p. 21.

[36]

STC 20687-88-89 (Greg 431-32, d-f), 21728-29, 24156 (454A).

[37]

12613, 22808, 24155 (453A).

[38]

4911 (439A1), 10886 (482), 14755 (455-56), 26068.

[39]

291 (439AII), 3075, 17638 (470AI), 17642 (424A), 22801. In this and the preceding references it will be understood that the imprints are not, in all cases, identical but only equivalent to those cited in the text.

[40]

The paper previously identified as H1420 with chainlines 23 mm. apart (fn. 12) appears in all copies of this state, either as a full sheet (CLU-C, with w/m state C in Effigies leaf) or as the title-leaf (NN, w/m state A; other copies unwatermarked). In these others the disjunct leaf A5 has lines spaced either at 25 mm. (NN), the normal interval for the H1731 stock used elsewhere in the impression, or at 20 mm. (CSmH, NBuG), an interval for stock as yet unidentified.

[41]

Variant 2 has w/m as indicated in all copies known to me except for the two at CLU-C. In one of these the sheet is of the usual paper (H1731), but with w/m turned as in variant 1. In the other the sheet is of paper (H1420) found elsewhere only in copies with the Aspley imprint. The w/m for this, in state C, is in the title leaf. See preceding note.

[42]

Cf. Greg, items 171b, 308, 360c, 363c-d, 379c.

[43]

In a few copies of state b (e.g. DFo37) the position is normal, 66 mm. to the right of the "margin"; but in most the position corresponds to that in c1. Perhaps this displacement is to be accounted for, in part at least, by varying degrees of pressure applied against the side of the forme in lock-up.

[44]

See fns. 12, 40, 41. A reversal of this situation occurs in the last four quires of the book, where a few odd lots of H1731 intrude upon a run of H1420.

[44a]

Through design or coincidence the order of imposition arranged by the overseer corresponds, with one exception, to the order of names in the colophon. In this Allot's name is entered at the end rather than at the beginning of the sequence.

[45]

By "perfect" I mean all copies with a title-page, whether or not this is integral to the book in which it is found.

[46]

Defined as containing two reams or 43 quires, with 24 sheets to each quire. Caleb Stower, The Printer's Grammar (1808), p. 402. I here refer, of course, to the bundle provided for each sheet of the edition except the underprinted A2.5.

[47]

See Stationer's Record, as corrected in Greg's Bibliography, for November 16, 1630 (to Allot), June 19, 1627 (to Cotes Brothers), November 8, 1630 (to Richard Cotes), October 28, 1600 (to Hays), November 26, 1607 (to Butter). The derelict plays were Richard II, Richard III, and I Henry IV, last assigned to Law on June 25, 1603; Midsummer Night's Dream assigned to Fisher Ovtober 8, 1600; Troilus and Cressida entered to Bonian and Walley January 28, 1608/9; and King John, a play generally confused with Troublesome Reign.

[48]

In the several studies of the significance of variant imprints it has not been sufficiently emphasized that a considerable portion of the copies of any one edition may have been sold by persons unidentified in the imprint. One may recall several examples of this practice, notably the sale of a 2nd Folio by Richard Whitaker to Sir Thomas Barrington in 1637 (Johnson, op. cit., p. 87, 91), and Robert Allot's consignment to the Edinburgh bookseller Samuel Hart, on August 11, 1635, of some 13 books, only 3 of which carry his name as publisher. See Johnson's list, items 60, 115, 213, 252, 257, 258, 368, 390, 428, 454, and for the Allot books, 64, 234, 517.

[49]

The copy was, presumably, a proof-sheet of the earliest state a and not, under the present hypothesis, of the state b which had previously designated the Cotes allotment of copies. The earlier distinction would of course have been quite irrelevant to an issue in 1641.

[50]

Frank MacKinnon, "Notes on the History of English Copyright," Oxford Companion to English Literature, 2d ed. (New York, 1944), p. 879.

[51]

See Smith's account (pp. 40-45) of Sir Edwin Durning-Lawrence's unorthodox publicity methods. Beginning with Sir Edwin's peculiar thesis at the turn of the century, the "Effigies" controversy is the story of error steadily compounded into the distorted form now accepted by Milton scholars. Among others, see Harris W. Fletcher (ed.), John Milton's Complete Poetical Works (Urbana, 1943), I, 365-366.

[52]

Smith, op. cit., p. 34. Elements of text and type common to I and II: 6, 13c, 14b, 22b, 25a; common to I and III: 11 13a, 14a, 17a, 17b, 18, 22a, 22c, 25b, 28. Type not available upon the composition of I but present in II-III: 3, 4, 13b. Ligatured type peculiar to I: 9.

[53]

Under this arrangement starre-ypointing (17b), the crux of the Effigies controversy, occurs as the correct reading in the original issue I and in one of the two reprints of I. This simple deduction relieves us from the necessity of maintaining the usual argument that it was originally starre-ypointed, "corrected" in a new setting at Milton's insistence and then, for some inexplicable reason, "corrected" yet again in still another setting. Are we expected to believe that, for this page alone, it requires the handling of 3972 sorts to alter a suffix of two letters? See fn. 4.

[54]

Reproduced as "Effigies C" in Smith, p. 39; also in SQ, II (1951), 330.

[55]

The existence of this variant (then considered as an early state of the 3rd issue) was first noticed in a description of the Viscount Mersey copy (Sotheby catalogue, June 29—July 1, 1938, item 699) and subsequently, with a facsimile of the state, in Book Hand-book, I, nos. 7-8 (1950). I am indebted to Mr. Reginald Horrox, the editor of the Handbook, for the reference to Sotheby and for his strenuous but unavailing efforts to trace the Mersey copy.

[56]

For any other order the only explanation is one which depends upon the "lapse and recovery" hypothesis. The supposition for this is that the unattended compositor conscientiously endeavors to improve the readings from one reprint to another, and if he slips at any time supplies a reading at least as good as the original as he strives on nearer and nearer to perfection. The theory is very appealing, often proposed, and as often contradicted by strictly bibliographical evidence.

[57]

Pierpont Morgan copy 5126. Fortunately for the bibliographer, this copy is sophisticated, since neither A2 nor A5 has a watermark.

[58]

So in the reconstructed Mersey copy, as described by Sotheby, in Folger 14, another made-up copy, and in Folger 34, apparently genuine. Except for these three and the one mentioned in the preceding note all copies of this issue known to me carry the Effigies page in the later state.

[59]

Allan H. Stevenson, "New Uses of Watermarks as Bibliographical Evidence," Papers Biblio Soc. U. of Virginia, I (1948), 160 and fn. 33.

[60]

The mid-point in an issue of 864 would be 432, or 192 copies of a and the first 240 of some 456 sheets of b.

[61]

A reasonable assumption, I believe, considering the importance of the book and the person for whom it was designated. By a provision in the act of 1666 (17 Car. II, cap. 4) the practice shortly became mandatory.

[62]

Eikonoklastes, ch. 1, par. 15.

[63]

Entry for Sir Thomas Herbert, DNB, IX, 666-68.

[64]

The cost, £5-10-0, is equivalent to the highest price Smith records (op. cit., p. 54) for a copy sold in 1799.