 | |  |
Notes
'The | Wheel of Life | by | Ellen Glasgow | [device] | New York Doubleday, Page & Company | 1906'. Of the 150 copies, I have examined twenty; the rest were examined by private persons and staff members of various libraries. A random followup check was made of forty of the copies not seen by me to determine the accuracy of reporting. I wish to thank the librarians and scholars who examined copies for me.
Other pages damaged in some copies are 17, 21, 96, 419, 438, 451; all of this damage occurs on the margins of pages.
The following abbreviations are used in this table: C — College, U — University, P L — Public Library.
The twenty-four patterns are represented by the following numbers: 1-5, 2-1, 3-9, 4-7, 5-5, 6-9, 7-8, 8-3, 9-12, 10-5, 11-5, 12-3, 13-1, 14-1, 15-1, 16-1, 17-2, 18-2, 19-25, 20-6, 21-2, 22-3, 23-4, 24-30.
In assuming that these 150 copies are an accurate reflection of the impression of 3,000 copies, one also assumes 1) that there are no copies undamaged in both formes of gathering 6, and 2) that the ratio of damaged states of gathering 6 found in the sample can be applied to the whole impression sheet. By accurate reflection of the impression is meant a statistically sound sample, one which will give narrow confidence limits. The 1% sample in general use by bibliographers of hand-printed books is not statistically sound for the application described in this note. For it is important here not only to have all the states of the gathering represented in the sample, but also to have them represented in a ratio which closely approximates the ratio of states in the whole impression. I am indebted to Professor Marvin Tummins of the McIntire School of Commerce of the University of Virginia for his help with the statistical problem.
Theodore Low De Vinne points out that there are inherent objections to this method of imposition (sheetwise imposition). "One of the difficulties of sheetwise imposition is that of making register when there is shrinkage of furniture in either form . . . as a rule, the form that can be printed perfect on itself as a half-sheet is printed with more ease than if the pages were imposed in two forms." (Book Composition, ed. J. W. Bothwell, 1918, pp. 154-155).
In the foregoing table, the numbered steps of the perfecting do not represent any necessary order of perfecting. That is, all three steps are interchangeable given the order of printing of the white sheets.
Supposing the damage to p. 84 occurred with the perfecting of the sheets, one would also have to suppose the order of printing and perfecting of the following table:
- 1. 1,260 full sheets printed without damage
- 2. 860 sheets printed with damage to p. 86
- 1. 1,260 half-sheets perfected with damage to p. 86 and 1,260 half-sheets perfected with damage to p. 84
- 2. 860 half-sheets perfected with damage to pp. 84 and 86 and 860 half-sheets perfected with damage to p. 86
It seems significant that both of the Library of Congress deposit copies have gatherings 28 and 29 in the undamaged state. Copy A is rebound, but Copy B is in the original binding and is cut and trimmed.
This explanation, or one something similar to it, seems the most likely to me. One would naturally be more satisfied with an explanation less conjectural.
Miss Helen Crosby of Doubleday & Co., Inc. informs me that Doubleday no longer has records for 1906.
Thus damage to tops of pages, such as on p. 451 in gathering 29, is unlikely to be the result of normal pressure of the cylinder since the tops will normally lie inside the edges of the forme. The conclusion might be drawn that damage to tops of pages is a safer indication of separate impressions than damage to bottoms.
See "Half-Sheet Imposition in 19th and 20th Century Books," forthcoming in Gutenberg Jahrbuch, 1962.
The objections raised to the usual method of half-sheet imposition of an octavo gathering apply as well to variations known to me. See the schemes discussed by De Vinne, Modern Methods of Book Composition, (1904), p. 353.
One interesting question which has not yet been answered is whether gatherings 28 and 29 were imposed together. The Wheel of Life contains thirty gatherings and, if the formes of two gatherings were imposed together beginning with gatherings 1 and 2, gathering 28 would have been imposed with gathering 27 and gathering 29 with gathering 30. Furthermore, the pattern of damage in gatherings 28 and 29 offers some support to the belief that they were not imposed together. If, in the scheme illustrated above, the pages of gathering 27 are substituted for those of gathering 29, and the entire forme turned 180°, it can be seen that the damage to gathering 28 occupies exactly the same position as the damage in gathering 29 on the leading edge of the forme. The argument here is that ordinarily the bottoms of the pages of the earliest gathering of the two in the forme were at the leading edge, but that in the case of the forme containing gatherings 27 and 28, the forme was placed on the press upside down so that the pages of gathering 28 occupied the leading edge. This argument is strongly supported by the ratio of damage in the inner and outer formes within the two gatherings. In gathering 29, page 446 (outer) is damaged in 115 copies and page 444 (inner) in 102 copies. In gathering 28, page 434 (outer) is damaged in 104 copies and page 432 (inner) in 115 copies. If, as argued, the forme containing gathering 28 was turned 180° it is easy to see why 432 is damaged in exactly the same number of copies as page 446 and page 434 in almost the same number as page 444. The same amount of stress was placed at the same points on the leading edge of the forme. I wish to thank Mr. John Cook Wyllie and Professor Bowers for their advice, and Mr. Willis Shell of the William Byrd Press of Richmond and Matthew Bruccoli for critical readings in an early stage.
| | |