University of Virginia Library

Search this document 


expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
collapse section 
Sir John Davies' Nosce Teipsum, 1599: A Bibliographical Puzzle by Gerald J. Eberle
expand section 

expand section 


Page 133


Page 134


Page 135

Sir John Davies' Nosce Teipsum, 1599: A Bibliographical Puzzle
Gerald J. Eberle

THE FULL STORY OF THE DIFFERENT ISSUES of Nosce Teipsum, 1599, a quarto, has by no means been told. I propose to add a chapter, analyzing and identifying the twenty copies I have examined or have had examined for me. Of these copies of STC 6355, all printed by Richard Field for John Standish in 1599, eleven are usually described as first editions, nine as second. Some libraries have made a further distinction between a first and second issue of the first edition, based upon variant readings in line sixteen of signature C1v. The reading "chaunge" has been used to distinguish the first issue, the second issue having "chaunce" at this point.

These editions and issues, however, are linked by more than their common imprint. In signatures K and L, for example, the first and second editions are being printed very nearly at the same time. Standing type from the first edition, signatures K4 and L3, is used with the new running-title in the second edition. More significantly, in two copies of the second edition —one in the Chapin Library of Williams College, the other in the Folger Shakespeare Library, the copy identified as STC 6355b—outer K of the second edition is perfected with corrected inner K from the first edition.

Apparently this perfecting was stopped early; the perfecting


Page 136
of outer K in other copies of the second edition was continued with only one page of standing type.

Because this linking of editions is but a small part of the central problem, I have not attempted to examine all available copies of the second edition. Nor have I reported in the charts five copies which appear to be normal second editions throughout. Those unreported copies are in the Rylands Library, Manchester; Emmanuel College Library, Cambridge; the Henry E. Huntington Library; Yale University Library; and the Detroit Public Library.

A more significant link between the first and second editions is to be seen in three copies: the Chapin Library copy, the Library of Congress copy, and the Bodleian copy. These three are usually identified as second editions, but all three have signatures B and C from the first edition. I doubt that these are made-up copies in the usual sense of the word.

These links are charted in the accompanying Table I, the full meaning of which will become apparent as this report progresses. At the moment a brief word of explanation is in order: I have used the symbol X for the first edition and Y for the second. The appended u and c indicate the uncorrected and corrected states of a forme, the combined uc indicates a partially corrected forme. The X1 and X2 indicate the first and second settings of type in outer B and C, about which more later. The absence of an appended symbol does not necessarily indicate an invariant forme but merely that I have not recorded any corrections. The gaps in the chart call attention to those copies which have not been studied fully, because I found it necessary to have them examined for me. Those fully reported I have studied either in the originals or in photostats.

As is apparent from the chart, the linking of editions is just one of several abnormalities that resulted from the complicated printing of Nosce Teipsum. That several formes appear in two settings of type has been long known, but not explained. I


Page 137
propose here to report fully what I have found and to attempt an explanation.

Though the creaking of the machinery while the guns are hauled into position might prove somewhat annoying, several more charts must be displayed and explained: the accompanying Table II indicates the distribution of running-titles in all formes of the first edition. The apparently arbitrary choice of letters to designate running-titles was a device to facilitate recognition. Each symbol indicates the letter in the running-title which is its most obvious distinguishing mark. Thus the so-called U running-title is that in which the U and M are badly spaced. It will be noted immediately that after sheet G the arrangement of these heads in sets of four is slightly varied. The earlier grouping has heads S U T E in one unit and O I N P in another.

By way of warning it must be noted here that in the rearrangement of heads after sheet G, the bad spacing of the U head was remedied. But close examination leads me to believe that it is the same head, not a new one. In passing I might add that the running-titles of Nosce Teipsum are somewhat hard to manage because they are all upper-case letters and do not show many marks of positive identification. Only the simultaneous examination of many copies made it possible for me to ignore minor variations due to faulty inking or dirty type.

Table III is involved in the distinction between the first and second settings of type in outer B and outer C, and generally in the relationship between the first and second editions. It charts an odd practice by one of the compositors in Field's shop, the man who set the type for almost all of the second edition and part of the first; this compositor habitually placed a space in front of each mark of final punctuation. The formes in both editions which are set by this compositor are marked W, the unspaced-punctuation formes are marked N. The chart refers to the first setting of type in the first edition, and momentarily


Page 138
ignores a slight exception which will be discussed in detail later.

The forme that I have listed as B(0)X2, the reset, is unusual in that the so-called wide-spaced final punctuation appears on B2v and B3, and the unspaced or narrow punctuation on B1 and B4v. This abnormality obviously suggests a resetting by two compositors, each working from half a sheet already printed on one side. No other conceivable explanation is as satisfying as this one; hence I have not hesitated to label the second setting positively.

This reset of outer B is superior to the original setting in several important readings, and includes on B4v the use of italic instead of roman type in lines 4, 5, 7, 8, 9, 11, 13, 14, and 15. Ordinarily we expect reset pages to be inferior to the original setting in appearance at least.[1] But here the evidence of two compositors at work, each on a half sheet, is, I think, decisive. Such a division of copy is expedient when the press is standing by, waiting for a forme which had been mistakenly distributed too soon.

Because the reset in B(o) is marked by improvements in readings and by the substitution of italic for roman, a needed correction, the same touchstones ought to be valid in C(o).

In outer C the reset forme substitutes italic for roman in nineteen words or phrases; it adds twenty commas; it improves seven readings, adds one typographical error and rectifies another. This analysis is re-inforced especially by one faulty reading in the first setting: "Aƈt" for "Art" on sig. C3. There can be no question that the correct reading is "Art". A compositor setting from manuscript could misread and set "Aƈt". But it is unlikely that he should set "Aƈt", using the ligature, from a printed page reading "Art". The more likely explanation is that he was resetting from corrected copy that consisted of the printed word "Aƈt" with marginal proof corrections to indicate the proper reading.


Page 139

When we turn to examination of the combinations of formes in B and C, we note several oddities. In sheet B, for example, all copies of the uncorrected state of the inner forme are perfected by the first setting of type, and all copies of the corrected state by the second setting.

In inner B the corrected state is marked by these readings:

Sig. B1v, line 14  Good and Truth  
Sig. B2, line 4  retain'd 
Sig. B4, line 4  threat 
The uncorrected readings are "God and Truth", "retained", "thereat". Only the last is a manifest error, though all the corrected readings are somewhat superior to the uncorrected.[2]

The corrected state of inner C (not a new setting of type at all) is distinguished by the readings "chaunce" and "Doctrines".

As is noted in Table I, the wholly uncorrected state of C(i) in the first edition, with the reading "chaunge" on sig. C1v line 16, and the reading "Doctrine" on sig. C2 line 6, has been found in only one copy, the so-called first issue of the first edition in the Huntington Library. Four copies retain "chaunge" but correct to "Doctrines". The remaining copies make the further correction to "chaunce".

Quite the opposite of what we found in sheet B, the five uncorrected and partially corrected formes of inner C are perfected by the second setting of type. One additional copy, that in the Carl Pforzheimer Library, perfects the corrected state of inner C with the second setting. All other copies have corrected inner C perfected by the first setting of type.

The central question, as I take it, is this: How could a miscalculation go unnoticed until two successive outer formes had been printed and their type distributed? The notion that the formes had pied may be discounted immediately, for the running-titles were not disturbed.


Page 140

Any explanation must, I believe, take into account the odd order of perfecting in sheets B and C, as well as the apparent over-printing of these sheets that manifests itself in the three copies of the second edition which have sheets B and C from the first edition.

It seems to me that we must assume, first of all, that the outer formes of sheets B and C were printed successively or perhaps simultaneously before perfecting was begun. Putting together all the evidence of running-titles, variants, and oddities, and making use of traceable broken letters, I offer one possible solution of the problem.

In the following tentative explanation of what perhaps happened in Field's shop, the number of copies printed is, of course, pure conjecture, however reasonable the figures might appear to be. Because the time element is so important at the start at least, I have numbered the various steps in the process.[3]

1. Forme B(o)X1 [S E T] and Forme C(o)X1 [O P N] were first imposed and run off. After about 800 copies had been machined, the type was distributed. It is possible that two presses worked off these formes simultaneously, and in view of what happened later it is perhaps probable.

2. Forme B(i)Xu [S U T E] perfected the 800 copies of Forme B(o)X1 [S E T]. Here in sig. B3v line 8 appears for the first time a distinctively broken lower-case roman letter g in the word "sight".

At this point it was decided to enlarge the edition to about 1400 copies. The decision was reached before the perfecting of sheet C was begun.

3. For some reason Forme C(i)Xu [O I N P] started to print white paper before perfecting C(o)X1. After about 100 copies had been printed, the first correction was made. After about 400 copies of C(i)Xuc [O I N P], the second correction


Page 141
was made. The final 100 copies on white paper were printed from C(i)Xc [O I N P].

Then the now wholly-corrected Forme C(i)Xc [O I N P] went on to perfect the 800 copies of C(o)X1.

4. Meanwhile Forme B(i)Xu [S U T E] had finished perfecting and had gone on to print 600 copies on white paper, but the forme had in the meantime undergone correction. Thus we have 600 white paper copies of B(i)Xc [S U T E].

5. The resetting of already distributed type occurred here. Forme B(o)X2 [O P N] used the ornament that had been used in Forme B(o)X1 and some of the display type from the head-title: for example, the g from the word "knowledge", but not the l. This Forme B(o)X2 [O P N] perfected the 600 copies of B(i)Xc [S U T E].

6. In Forme C(o)X2 [O P N] the previously mentioned wide-spacing compositor assisted, setting sigs. C2v and C3. Here, on sig. C1, the oddly broken letter g re-appears.

7. The 600 copies of Forme C(i)Xu-Xuc-Xc were perfected by Forme C(o)X2 [O P N]. And that completed the printing of all of sheets B and C for the first edition.

At this point it was decided to enlarge the edition further, to perhaps 2400 copies, by making of it a double edition. The composition for what we call the second edition could have started at about the same time as step number 5 above, but not very well earlier. I believe that the wide-spacing compositor began to set Forme B(o) for the second edition as soon as he had completed his two pages for B(o)X2, for these reasons:

For sig. B1 of the second edition he used some of the remaining letters of display type from the head-title of Forme B(o)X1: for example, the l from the word "knowledge" and the e from the word "humane". The ornament used is a mirror image of that used in both settings of the first edition. Thus the re-appearance of certain letters and non-appearance of others gives us probable temporal termini within which the composition of the second edition may have begun. I assume that composition began after


Page 142
B(o)X1 was distributed and before B(o)X2 was completed.

The decision to enlarge again, when the last of 1400 copies of sheets B and C were in process of being machined, made it advisable to print only 1000 copies of those sheets in the second edition. Thereafter each edition would print 1200 copies. The odd distribution of running-titles in the first edition can be reasonably accounted for if we assume that one press printed one forme each day.

While the first edition machined sheet D, the second edition was working with sheet C. The distinctively broken letter g appears now in both C(o) and C(i) in the second edition. Obviously the type from C(o)X2 was distributed by the wide-spacing compositor in order to replenish a case that must have been quite short of certain sorts; so short, in fact, that it became expedient to distribute one of the formes in sheet C in the second edition before composition for the other forme was well under way. This unusual state of affairs came about probably because the second edition was being set by formes from proof-read sheets of the first edition as they came off the press. Had this delay between inner and outer formes of a single sheet occurred at the end of a working day, no special inconvenience would have resulted.

From this point on, it seems likely that the second edition was composed by formes from the first edition and printed at the same time as the first edition but slightly more than one forme behind. Tracing the broken g through the formes enables us to see some such pattern.

The broken g, last used in both formes of sheet C in the second edition, appears nowhere in sheet D. But it occurs in the first edition in sig. E3v line 14, and in sig. F2 line 8.

Then it appears again in the second edition, in G1v line 4. Then back to the first edition in H3 line 6 and in I3v line 13.

It seems reasonably certain that the second edition was at least one forme behind the first edition but probably not much more than two formes behind. After sig. I3v the distinctively


Page 143
broken g disappears. Fortunately, other links become apparent.

It is noted in Table III that the wide-spacing compositor set both formes of sheet H, the inner forme of sheet I, and both formes of sheets K and L in the first edition; his hand is also traceable in the second edition in all formes except inner and outer F, inner and outer I, and outer L. This suggests the switching of compositors wherever the pressure was greatest at the moment. I think, too, that after the second edition had machined inner I, all the resources of the shop were directed toward completing the printing of the first edition. Some evidence is available to enable us to determine the time relationship in sheets K and following in the first edition.

We note, for instance, that Forme K(i)Xc was used to perfect the first printed sheets of Forme K(o) in the second edition. The fact that the entire forme was used, running-titles and all, is significant, especially when we note that the relative position of those running-titles on the page is different in the second edition. Then, when the inner forme of K in the second edition was imposed, it used standing type from Forme K(i)Xc on sig. K4v but with a single italic word in the first line set in different type. Apparently the use of the first edition forme in the second edition took place after that forme was unlocked.

Now when we note that the running-titles used in K(i)Xc were used also in L(i)X we may assume that, unless the printing of sheet L in the first edition was delayed, the use of K(i)Xc to perfect K(o) of the second edition took place after the printing of the first edition was completed. In other words, the I T U S running-titles were re-imposed around the still standing pages of K(i)Xc for use in the second edition after they had been used for L(i)X. This explanation can be re-inforced by means of a further link between the two editions a bit earlier.

When sig. M1 in the first edition, a single page, was printed along with the preliminary matter,[4] only one running-title was


Page 144
needed. The title used was the one identified in Table II as P. Another running-title in the same O E N P group, the O title, for which there was no longer any use in the first edition, appears now on sig. I3 in the second edition. I have not been able to trace any other of the first edition running-titles in the second edition.

As a final link, the display type from the first edition title-page and from sig. A3 was re-used in the second edition but with a different ornament, again a mirror image of that which was used in the first edition. That the second edition was printed after the first here is determinable by the progressive deterioration of the central stem in the letter m in the word "Humane" on the title-page. The letter is damaged throughout the first edition copies and through part of the second edition copies. In some copies of the second edition the piece is broken off completely. In general, it seems clear that the outer forme of sheet I in the second edition was imposed while the press was machining the final formes of the first edition.

To sum up: in the outer forme of sheet I of the second edition one running-title from the first edition was used.[5] In the inner forme of sheet K an entire forme was used for some of the perfecting of the second edition. Then a single page of standing type from the inner forme of sheet K in the first edition was used with a new running-title in the newly set inner forme of sheet K in the second edition. This must have been a time-consuming process, but no simpler explanation suggests itself. The possible overprinting of the inner forme of sheet K in the first edition would not account for the different position of headlines in the normal Forme K(i)Xc and in those pages as they appear in the second edition sheets. Similarly, in sheet L a single page of standing type from the first edition was used on


Page 145
sig. L3 with a new running-title and with reset marginal notes in the second edition.

Furthermore, two compositors are at work on the final sheets of the second edition: the inner forme of sheet L is wide-spaced before final punctuation except for the first two lines of sig. L1v; the outer forme of sheet L is narrow-spaced except for sig. L3, where the wide spacing results from the fact that this is standing type from the wide-spaced first edition.

In passing it might be noted that all the evidence that the second edition was set from the first is borne out by careful collation of variant readings. I make a special point of this since we might be tempted to assume that sheet I in the first edition was set from already printed formes of sheet I from the second edition, because of the odd division of labor between compositors. But I believe that the use of the O running-title on sig. I3 precludes this. Apparently we must assume a division of copy designed to allow each compositor to set a complete forme.

Finally, the fact that the variant reading on sig. K3v— "will" for "will"—represents a miscorrection in the first edition that was repeated in the second must not disturb our belief that the second edition was in every forme set from the first.

The complicated order of formes through the press in sheets B and C of the first edition, as outlined above, would result in the following combinations of variants:



Page 146


It is unfortunate that the wide distribution of copies of the first edition makes so difficult the examination of watermarks and paper. Otherwise it might be interesting to see whether such a study would confirm or render untenable the beliefs expressed here.


Page 147

  • Copies reported in numbered columns:
  • 1. Huntington Library copy listed as first edition, first issue.
  • 2. Huntington Library copy listed as first edition, second issue.
  • 3. The Folger Shakespeare Library copy. STC 6355a.
  • 4. The Pforzheimer Library copy.
  • 5. The British Museum copy. C.34.f.22.
  • 6. The New York Public Library copy.
  • 7. The Wellesley College Library copy.
  • 8. The Harvard University Library copy.
  • 9. The Trinity College Library copy. Capell Q91.
  • 10. The Trinity College Library copy. Capell Q93.
  • 11. The Trinity College Library copy. Capell S364.
  • 12. The Bodleian Library copy. Malone 716.
  • 13. The Chapin Library copy.
  • 14. The Library of Congress copy.
  • 15. The Folger Shakespeare Library copy. STC 6355b.



Page 148




See McKerrow, An Introduction to Bibliography (Oxford, 1928), pp. 184-6.


Because the second edition was set from uncorrected copy, it corrects the obvious error to "threat", but retains the faulty "God".


The symbols used in Tables I and II have been used throughout the discussion that follows. The running-titles used for each forme are placed in square brackets after the identification of formes. Thus the symbol B(o)X1 [S E T] means "outer B of the first setting of type of the first edition with the running-titles identified as S E T."


Note that sig. A in Table I is not simply divided into inner and outer formes. The unsigned title-page has its verso blank; the next leaf, signed A3, begins the dedicatory verses that end on the verso. Signature M1 was probably printed with the preliminaries. All known copies lack A1 and A4. The M1 page from the first edition in one copy of the second is probably the result of sophistication.


The running-titles in the second edition present a hopeless picture of lack of order. I find at least fourteen different titles, some used as many as seven times; is this evidence of a pool of running-titles?