University of Virginia Library

Bibliographical Evidence

When stop-press corrections are made, they come early in the run, or there is, of course, little point in making them. That is, if the aim of the printer is to have as many corrected sheets as possible, he will begin proofing and correcting as soon as he can. And if one is talking about authorial corrections, or at least an attempt to follow authorial intention, one or the other of which seems most likely in the present instance, corrections would be made early or not at all. Therefore, the mere proportions of copies with the Astraea/Pallas conclusion or the Pallas/Astraea conclusion provide evidence as to which is the corrected version. In the argument I wish to pursue here, this evidence corroborates conclusions that are drawn from a close examination of physical differences in the printing of the two versions. The sequence is this: all of the large-paper sheets of 4Q2-3v and 1v-4 were machined with the Pallas/Astraea conclusion[11] and then some of the small-paper sheets (in the neighborhood of ten percent) were.[12] Then someone intervened to reverse the order of the Pallas/Astraea speeches. The sequence is decisively established by a piece of type that was broken during the machining of 1v-4 after the speeches had been rearranged to Astraea/Pallas.[13] In all large-paper copies and in the small-paper copies with the original arrangement (i.e., Pallas/Astraea), after the word "unsold" in line thirty of 4Q4 (the penultimate line of Astraea's speech) the comma is intact (see Fig. 1). After the type was rearranged, and the line became line eight of 4Q4, for a while the comma remained intact (as in four of sixteen copies with the final leaf present that I have recently examined). Subsequently, however, during the press run of the rearranged forme (Astraea/Pallas), the comma, its face somewhat broken, went adrift. As the broken comma moved, it was pushed into the letter "d," more and more, by stages. A glance at Figs. 2 and 3 (reproduced at twice actual size) reveal that such movement did take place, even to the point of the broken comma subsequently damaging "d." The letter "d" and the disintegrating comma seem to have survived the press run; at least I am not aware of any evidence that they did not. Although one might argue that

the Pallas/Astraea ending was machined first, that during the press run, or even after, the comma was broken and subsequently replaced when the speeches were rearranged, the argument will not hold. Even punctuation marks can be distinctive. Depending upon how the comma was inked (there is of course some variation), little or almost no curve can be seen in the concave side; the convex side is flattened at an angle of about forty-five degrees at the bottom. I have taken a sample of commas in the text nearby. Of one hundred successive commas, this one being the last of the lot, none compares with it. Although there could, of course, be more than one comma thus distinctive, the chances are slim, and the chances are exceedingly remote that a compositor replaced the present one with a "duplicate" (if such a thing even existed) when the Pallas/Astraea ending was altered to be Astraea/Pallas, or at any other time. If one compares the (intact but somewhat distorted) comma following "unsold" with the commas in the half-dozen lines immediately preceding, one sees that the "unsold" comma remained the same until it was broken.

There are, conveniently, several other details which, though less conclusive, are nevertheless useful in helping to confirm the order of printing that I have just described. Sometime after the Pallas/Astraea speeches were rearranged, another variation was introduced. In line three of 4Q2 a space between the words "fraud" and "and" was pulled to the level of the type face

(no doubt by the inking ball) and subsequently inked and printed.[14] The raised space is in itself less conclusive than the broken comma because the space could have been pulled to the "surface" and then subsequently forced back down. However, with the evidence of the broken comma providing the essential underpinning for my argument, the raised inking space serves quite well in support.

Another detail is that the Pallas/Astraea version has "The End" in caps and small caps, in rather small type. The Astraea/Pallas version has "The end" in much larger type and in upper and lower case. The font used for the reset "The end" is the same as that used for the words "The Catalogue," on the first page of set type of the preliminary matter, ¶3.[15] It is reasonable to suppose that the compositor, following the customary practice of setting the preliminaries last, was drawing upon that font when the speeches were reversed and that he employed it as a matter of convenience. As it happens, the font is one that was employed but sparingly in the Folio. Aside from "The Catalogue" on ¶3 and "The end" on 4Q4, it appears at most at sixteen places,[16] each time as part of the title of a play or entertainment or masque, but in no instance running for more than two or three words.

There is also another connection between the final pages of the volume and the preliminaries. The watermarks in the preliminaries are all grape bunches. I have charted all watermarks in twenty-two small-paper copies and have found only one instance of these grapes in the body of the volume (in 4I2.5 of one copy). This single example, present perhaps by accident, does at least show that the grape-bunch stock of the preliminaries was in Stansby's shop when some of the last quires were being printed.

To return to the issue raised at the first part of this section, how many copies, then, were printed with the original Pallas/Astraea ending and how many with the reverse? I have consulted sixty-nine copies of the Folio, and add to that number six of those cited by Greg which I have not yet examined, for a total of seventy-five. Four of those (all small-paper copies) are missing both of the last leaves, and thus provide no evidence. Of the seventy-one remaining, six are large-paper copies, all of which retain the Pallas/Astraea ending. One small-paper copy, the Lowell volume in the Houghton Library,[17] has bound in with the normal small-paper sheets several of large-paper cut down to small-paper size, including all of 4Q. This can be readily determined by examining the watermarks. There are some three dozen different watermarks in the various sheets used for the small-paper printing, but large-paper

copies have only variants of one watermark, a shield with three lions.[18] Of the sixty-four copies with the final quire printed on small-paper sheets, six have the Pallas/Astraea ending and fifty-eight the reverse. As the vast majority of the small-paper copies, and therefore a substantial preponderance of all copies,[19] have the Astraea/Pallas ending, the mere probability of there being more corrected copies than uncorrected suggests that Astraea/Pallas is the corrected state.[20]