University of Virginia Library

Search this document 


  

collapse section 
 1. 
 2. 
 3. 
  
collapse section 
 1. 
 2. 
 3. 
 4. 
 5. 
  
Notes
collapse section 
  
  
collapse section 
  
  
collapse section 
 1. 
 2. 
 3. 
  
collapse section 
 1. 
 1. 
 2. 
 3. 
 4. 
 5. 
 6. 
 7. 
 8. 
  
collapse section 
  
  
collapse section 
  
  
collapse section 
  
  
collapse section 
  
  
collapse section 
  
  
collapse section 
  
  
collapse section 
  
  
collapse section 
 1. 
 2. 
 3. 
  
collapse section 
  
  
collapse section 
 1. 
 2. 
  
collapse section 
  
  
collapse section 
  
  
collapse section 
  
  
collapse section 
 1.0. 
collapse section2.0. 
collapse section2.1. 
 2.1a. 
 2.1b. 
collapse section2.2. 
 2.2a. 
 2.2b. 
  
  

collapse section 
  
  
  
  
  
  
  

Notes

 
[1]

These quotations happen to have been taken from, respectively, Alfred P. Lee's A Bibliography of Christopher Morley (1935), p. 31, and Dorothy R. Russo and Thelma L. Sullivan's A Bibliography of Booth Tarkington (1949), p. 36. But similar phrases are found in a number of other bibliographies.

[2]

Yost, A Bibliography of the Works of Edna St. Vincent Millay (1937), entry 3.

[3]

Gallup, T. S. Eliot: A Bibliography (1952), entry A39.

[4]

Although actual descriptions of type are not common in bibliographies of post-incunabular books.

[5]

The classic discussion of cancels is R. W. Chapman's Cancels (1930).

[6]

The principal guide to the technical literature is Pulp and Paper Manufacture: Bibliography and Patents (edited, under various titles, first by Clarence J. West and then by Jack Weiner), which covers the period 1900-55 in five volumes and the years since 1955 in annual volumes; it includes — especially under the headings "History" and "Watermarks" — material of interest to bibliographers which is not always reported in the checklists that bibliographers more frequently consult. Another basic guide to this field is the Abstract Bulletin of the Institute of Paper Chemistry (1930- ); a useful shorter collection of abstracts is Jack Weiner and Lillian Roth, Paper and Its Relation to Printing (Institute of Paper Chemistry Bibliographic Series No. 164; 2nd ed., 1962). A helpful selective checklist appears at the front of E. J. Labarre's Dictionary and Encyclopaedia of Paper and Paper-Making (2nd ed., 1952), pp. xi-xx.

[7]

Greg, "On Certain False Dates in Shakespearian Quartos," Library, 2nd ser., IX (1908), 113-131, 381-409.

[8]

Carter and Pollard, An Enquiry into the Nature of Certain Nineteenth-Century Pamphlets (1934), esp. pp. 42-55 ("The Analysis of the Paper").

[9]

Stevenson, The Problem of the Missale speciale (1967).

[10]

Also, at roughly this time, Henry Thomas expressed the need for a "hand-book" of paper study, particularly the use of watermarks in bibliographical analysis; see "Watermarks," Edinburgh Bibliographical Society Transactions, II (1938-45), 449-450. And James G. McManaway, in his contribution to Standards of Bibliographical Description (1949), commented on the neglect which paper study had suffered (p. 68).

[11]

An important treatment of the general subject is Graham Pollard, "Notes on the Size of the Sheet," Library, 4th ser., XXII (1941-42), 105-137. See also Lawrence Wroth, "Formats and Sizes," Dolphin, I (1933), 81-95; and David Foxon, "Some Notes on Agenda Format," Library, 5th ser., VIII (1953), 163-173.

[12]

Nor about the number of leaves in a gathering, for a gathering can consist of half a sheet or of several sheets and still represent the same format (the same number of leaves per full sheet). For some discussion of the distinctions among format, signature, gathering, and size, see John R. Hetherington, "Signatures and Sizes," Times Literary Supplement, 14 October 1965, p. 928; and G. T. Tanselle, "The Sizes of Books," AB Bookman's Weekly, XXXIX (5-12 June 1967), 2330, 2332. Methods of analysis for detecting half-sheet gatherings are taken up in William H. Bond, "Imposition by Half-Sheets," Library, 4th ser., XXII (1941-42), 163-167; Luella F. Norwood, "Imposition of a Half-Sheet in Duodecimo," Library, 5th ser., I (1946-47), 242-244; and Kenneth Povey, "On the Diagnosis of Half-Sheet Impositions," Library, 5th ser., XI (1956), 268-272.

[13]

Blades, "On Paper and Paper-Marks," Library, 1st ser., I (1889), 217-223.

[14]

R. B. McKerrow, An Introduction to Bibliography for Literary Students (1927), pp. 164-174; Esdaile's Manual of Bibliography, rev. Roy Stokes (4th ed., 1967), pp. 237-244; Paul S. Dunkin, How to Catalog a Rare Book (1951), pp. 31-56.

[15]

In rare cases additional kinds of evidence may turn up, such as the printed lines which showed the binder where to make the cut-off in duodecimo; see Giles E. Dawson, "Guide-Lines in Small Formats (About 1600)," SB, XIV (1961), 206-208.

[16]

See A. T. Hazen, "Eighteenth-Century Quartos with Vertical Chain-Lines," Library, 4th ser., XVI (1935-36), 337-342; and Kenneth Povey and I. J. C. Foster, "Turned Chain-Lines," Library, 5th ser., V (1950-51), 184-200. Several books from the fifteenth century have been noted in which some leaves appear to represent different formats from other leaves, as a result of certain sheets having been cut in half (or quartered) before printing; see Curt F. Bühler, "Chainlines versus Imposition in Incunabula," SB, XXIII (1970), 141-145.

[17]

If one could count on the presence of a single main watermark in the center of one half of every sheet, whether a double sheet or one of the companion sheets from a double mould, one could use watermarks as a guide; but the lack of regularity in the placing of watermarks makes this test unreliable.

[18]

Bowers's discussion of turned chainlines is in Principles, pp. 193-195. A related example is taken up in Richard J. Wolfe, "Parthenia In-Violata: A Seventeenth-Century Folio-form Quarto," BNYPL, LXV (1961), 347-364.

[19]

See Edward Heawood, "The Position on the Sheet of Early Watermarks," Library, 4th ser., IX (1928-29), 38-47.

[20]

Although handmade papers can be made in unusually large sizes by having more than one man manipulate the mould.

[21]

Charles Evans in 1876 considered it "practically impossible at the present time to correctly define the size of a modern book in the old manner"; see "The Sizes of Printed Books," Library Journal, I (1876-77), 58-61.

[22]

McKerrow, p. 164n.; Bowers, pp. 429-430.

[23]

Steele, "On the Imposition of the First Edition of Hawthorne's The Scarlet Letter," Library, 5th ser., XVII (1962), 250-255. He further demonstrates that the book was printed by half-sheet imposition, with each forme containing the inner and outer subformes of a single quire.

[24]

Steele, "Half-Sheet Imposition of Eight-Leaf Quires in Formes of Thirty-Two and Sixty-Four Pages," SB, XV (1962), 274-278.

[25]

Practically all of the many printers' manuals published during the nineteenth and twentieth centuries contain imposition diagrams applicable to machine printing; a convenient one for reference is Theodore Low DeVinne's Modern Methods of Book Composition (1904), or his Book Composition, ed. J. W. Bothwell (1918).

[26]

Steele, "Evidence of Plate Damage as Applied to the First Impressions of Ellen Glasgow's The Wheel of Life (1906)," SB, XVI (1963), 223-231; part of his study also consists of an effective statistical analysis of the reliability of the sample of 150 copies which he examined. For some comment on the "leading edge" in modern half-sheet imposition, see Oliver L. Steele, "A Note on Half-Sheet Imposition in Nineteenth and Twentieth Century Books," Gutenberg Jahrhbuch 1962, pp. 545-547.

[27]

J. D. Thomas, in "A Modern Instance," PBSA, L (1956), 302-304, describes this kind of error in the folding of a sheet in the second edition of Besterman's World Bibliography of Bibliographies; and Matthew J. Bruccoli and Charles A. Rheault, in "Imposition Figures and Plate Gangs in The Rescue," SB, XIV (1961), 258-262, demonstrate the intended 64° format of the second impression of the second edition of Conrad's The Rescue through an analysis of the surviving imposition figures which identify the four-page plate gangs.

[28]

For folio, the shorter dimension of the leaf would be doubled; for quarto, both dimensions would be doubled; for octavo, the shorter would be quadrupled, the longer doubled; and so on.

[29]

The standard general history is Dard Hunter, Papermaking: The History and Technique of an Ancient Craft (1943), which includes a highly selective checklist of other historical treatments. Some additional checklists, which can serve as partial guides to the mass of historical research, are mentioned in footnotes 6 and 59; items before 1800 are described in Dard Hunter, The Literature of Papermaking, 1390-1800 (1925); and work since 1949 can be located through the annual SB checklists.

[30]

See R. W. Chapman, "An Inventory of Paper, 1674," Library, 4th ser., VII (1926-27), 402-408; cf. Chapman, "Notes on Eighteenth-Century Bookbuilding," IV (1923-24), 175-177 esp. Allen T. Hazen, in "Eustace Burnaby's Manufacture of White Paper in England," PBSA, XLVIII (1954), 315-333, reproduces a 1691 announcement of a paper auction, showing many names (but not dimensions) of paper sizes.

[31]

See the examples from 1821 reported by Lyman Horace Weeks in A History of Paper-Manufacturing in the United States, 1690-1916 (1916), pp. 119-120.

[32]

Bibliographers who consult Labarre's Dictionary should also be acquainted with Allan Stevenson's review of it in the Library, 5th ser., IX (1954), 59-63, which makes comments on some size-names not included in Labarre. See also E. J. Labarre, "The Sizes of Paper, Their Names, Origin and History," in Buch und Papier, ed. Horst Kunze (1949), pp. 35-54; and E. G. Loeber's Supplement (1967) to Labarre's Dictionary.

[33]

Gaskell, "Notes on Eighteenth Century British Paper," Library, 5th ser., XII (1957), 34-42, and John Baskerville: A Bibliography (1959), p. xvi; Stevenson, Catalogue of Botanical Books in the Collection of Rachel McMasters Miller Hunt, II (1961), ccxxvii. Cf. D. C. Coleman, The British Paper Industry, 1495-1860 (1958), p. 351.

[34]

A copy of this book, Illustrations of the British Paper Manufacture, can be found in the Wing Foundation of The Newberry Library.

[35]

Such as, for the sizes of fifteenth-century printing paper, Conrad Haebler's The Study of Incunabula, trans. Lucy E. Osborne (1933), pp. 49-54.

[36]

For an illustration of the method, see Curt F. Bühler, "The Margins in Mediaeval Books," PBSA, XL (1946), 34-42. Bühler mentions two sizes of medieval paper, 43 x 32 and 45 x 30 cm.; and he estimates that the average type page was 68% x 45% of the total height of the paper, that the inner margin was 8-10% of it, and that the height of the type page was equal to the breadth of the leaf. Cf. A. W. Pollard, "Margins," Printing Art, X (1907-8), 17-24; and "Margins," Dolphin, I (1933), 67-80.

[37]

For general comments on this subject, see G. T. Tanselle, "Tolerances in Bibliographical Description," Library, 5th ser., XXIII (1968), 1-12.

[38]

Cf. Bowers, Principles, pp. 308, 430.

[39]

See G. T. Tanselle, "The Identification of Type Faces in Bibliographical Description," PBSA, LX (1966), 185-202. The British Federation of Master Printers has officially adopted the metric system; cf. Eugene M. Ettenberg, "Is Type Measurement Overdue for Change to the Metric System?", Inland Printer/American Lithographer, CLXII (January 1969), 48.

[40]

See British Standard 3176: 1959. Cf. "The DIN System of International Paper Sizes," British Printer, LXXI (December 1958), 70-71; "International Standard Paper Sizes — Pipe Dream or Tangible Reality?", British Printer, LXXV (June 1962), 107-111; John Tomkins, "DIN — A New, Old Cause," Typographica, n.s., no. 5 (June 1962); Labarre, Dictionary, pp. 286-287; W. C. Kenneison and A. J. B. Spilman, Dictionary of Printing, Papermaking and Bookbinding (1963), pp. 211-213. In the DIN system, the basic sheet has an area of one square meter, and its dimensions are in the ratio of the side to the diagonal of a square (1:√2) — i.e., 1189 x 841 mm. Such a sheet is referred to as "A0"; "A1" is the sheet resulting from a halving of the longer dimension (841 x 594); "A2" from another halving (594 x 420), and so on. A "B" series (based on a 1414 x 1000 sheet) establishes intermediate sizes between those of the "A" series; long sizes may be specified in terms of a fraction of a standard size, as "¼ A4" (210 x 74).

[41]

The printed catalogue cards prepared by the Library of Congress express the height of books and the dimensions of broadsides in centimeters, and these practices are recommended in the Anglo-American Cataloging Rules (North American Text, 1967), pp. 210-211.

[42]

The direction of the chainlines in the sheet need not be mentioned unless it is unusual; the direction of the chainlines in the leaf — like other facts about the folded form of the sheets — should be recorded later, at the end of the paragraph on paper.

[43]

See A. T. Hazen, "Baskerville and James Whatman," SB, V (1952-53), 187-189; and Thomas Balston, William Balston, Paper Maker, 1759-1849 (1954), James Whatman, Father and Son (1957), and "Whatman Paper in a Book Dated 1757," Book Collector, VIII (1959), 306-308.

[44]

The importance of chainlines for bibliographical analysis in tests for cancels and conjugacy — not simply their distance apart but also the leaf-patterns of mould- and felt-sides as revealed by the indentations of chainlines — is discussed by Allan Stevenson in "Chain-Indentations in Paper as Evidence," SB, VI (1954), 181-195.

[45]

See G. T. Tanselle, "The Bibliographical Description of Patterns," SB, XXIII (1970), 71-102.

[46]

"Height" and "width" here refer to the mark itself; for some marks, therefore, the larger figure will appear second.

[47]

"Paper as Bibliographical Evidence," Library, 5th ser., XVII (1962), 200. An example of the use of chainspace measurements is Stevenson's "Tudor Roses from John Tate," SB, XX (1967), 15-34.

[48]

Since moulds were regularly used in pairs (see below, note 51), the presence of companion watermarks need not be specifically mentioned; however, a difference between the two can be conveniently recorded in this fashion: "bull's head tau mark, (I) 49 x 5[28]5, (II) 49 x 4[28]6."

[49]

See especially Observations on Paper as Evidence (1961), and "Paper as Bibliographical Evidence," Library, 5th ser., XVII (1962), 197-212; for a brief statement, see "The Natural History of Watermarks," in C. M. Briquet, Les Filigranes, ed. Stevenson (1968), pp. *20-23. In one of his earlier articles, he demonstrated how watermarks can be useful in the detection and analysis of press variants, the sequence of formes through the press, and related problems: "New Uses of Watermarks as Bibliographical Evidence," SB, I (1948-49), 151-182. For his discussion of chainlines, see above, note 44.

[50]

A condensed version of the argument appears in his "Paper Evidence and the Missale speciale," Gutenberg Jahrbuch 1962, pp. 93-105.

[51]

Stevenson, "Watermarks Are Twins," SB, IV (1951-52), 57-91.

[52]

See T. Gerardy, "Die Fotografische Registrierung von Wasserzeichen," Papier-geschichte, XVI (December 1966), 22-25. The use of sensitized paper to make direct photographs of watermarks was suggested at least as early as 1904 by Gilbert R. Redgrave, in "The Water-Marks in Paper," Library, 2nd ser., V (1904), 91-92.

[53]

See J. S. G. Simmons, "The Leningrad Method of Watermark Reproduction," Book Collector, X (1961), 329-330, which describes the method first announced in 1960 by D. P. Erastov and also used in J. L. Putman's Isotopes (1960); O. K. Nordstrand, "Beta-Radiographie von Wasserzeichen," Papiergeschichte, XVII (1967), 25-28; Stevenson, Problem, pp. 66-68; Stevenson, "Beta-Radiography and Paper Research," in VII International Congress of Paper Historians Communications (1967), pp. 159-168; Stevenson, "Watermark Beta-Radiographs," in his edition of C. M. Briquet's Les Filigranes (1968), following p. *36; Papermaking: Art and Craft (Library of Congress, 1968), pp. 72-77; and J. S. G. Simmons, "The Delft Method of Watermark Reproduction," Book Collector, XVIII (1969), 514-515.

[54]

Stevenson, Problem, pp. 248-252.

[55]

See also his essay on "Briquet and the Future of Paper Studies," in Briquet's Opuscula (1955), pp. xv-l.

[56]

Heawood's article on "Watermarks," in Labarre's Dictionary, pp. 328-360, is a useful introductory survey, but its tracings are less appropriate for bibliographical citation.

[57]

In using this work, one should consult Allan Stevenson, "A Critical Study of Heawood's Watermarks, Mainly of the 17th and 18th Centuries," PBSA, XLV (1951), 23-36.

[58]

At the back of Dard Hunter's Papermaking in Pioneer America (1952) are some good photographs of early American marks.

[59]

Reprinted in The Nostitz Papers (1956), pp. xxxvii-xlii, and in Philobiblon, I (1957), 237-251. Another important list of material, limited to the period before 1600, is in C. M. Briquet, Les Filigranes, ed. Stevenson (1968), pp. *37-53; an earlier list is Dard Hunter, Handmade Paper and Its Watermarks: A Bibliography (1916). See also E. J. Labarre, "The Study of Watermarks in Great Britain," in The Briquet Album (1952), pp. 97-106.

[60]

Stevenson, in the Hunt Catalogue, uses "cf." and also explains concisely its significance: "The references to Heawood, Churchill, Voorn, Nicolaï, and others are not to marks from the same moulds (which are difficult to be sure of from tracings) but to sufficiently similar marks for the reader's understanding" (p. clxxix).

[61]

Gasparinetti, "On the Adoption of a Universal Terminology for Watermarks," in The Briquet Album, pp. 122-124.

[62]

The Briquet Album, pp. 125-154 (English, German, Italian); Les Filigranes, pp. *109-131 (French, English, German).

[63]

Gaskell, in the Baskerville bibliography, uses an oblique line for this purpose; oblique lines have been suggested above, however, as separators for variant measurements.

[64]

The basic tests for paper are outlined in any of the standard works on paper manufacture. See, for example, Julius Grant, A Laboratory Handbook of Pulp and Paper Manufacture (1942), pp. 179-297 (which includes a good analysis of the appearance of fibers, pp. 251-258); Robert H. Clapperton, Modern Paper-Making (3rd ed., 1952), pp. 450-464; J. Newell Stephenson (ed.), Manufacture and Testing of Paper and Board (1953); C. Earl Libby (ed.), Pulp and Paper Science and Technology (1962), II, 373-398; Robert R. A. Higham, A Handbook of Papermaking (1963), pp. 72-86; and Victor Strauss, The Printing Industry (1967), pp. 577-580. An important introduction to the subject of paper testing is Paper and Paperboard: Characteristics, Nomenclature and Significance of Tests (ASTM Special Technical Publication No. 60-B; 3rd ed., 1963); it provides, at the beginning, a list showing the correspondences between the two sets of test standards, those of the American Society for Testing and Materials (ASTM) and those of the Technical Association of the Pulp and Paper Industry (TAPPI). The ASTM standards are set forth in Part 15 of the Book of ASTM Standards, issued annually; the TAPPI Standard Testing Procedures are available in two loose-leaf volumes, revised continually. A Bibliography of Paper Testing appears in the Institute of Paper Chemistry Bibliographic Series as Nos. 154-156 (2nd ed., 1954, with supplements in 1960 and 1965) and Nos. 157-159 (2nd ed., 1960).

[65]

Cf. Tanselle, "Tolerances," pp. 5-6.

[66]

See TAPPI Method T410-os and ASTM Method D646.

[67]

Tables of equivalences, for converting pounds per ream to grams per square meter, are available; see, for example, Clapperton, Modern Paper-Making, pp. 496-497.

[68]

The term "bulk" is also used to mean the total thickness of a given number of sheets, and it is in this sense that the word may appear later in a bibliographical description of paper to indicate the combined thickness of the folded sheets in a finished book.

[69]

Micrometers for official paper testing must conform with TAPPI Method 7411-m44 and ASTM Method D645-64T and have .0001” graduation; but most dial micrometers with .001” graduation are adequate for bibliographical work.

[70]

A convenient pocket micrometer, satisfactory for bibliographical purposes, is the Ames Thickness Measure No. 25 (with .001” graduation) or No. 25M (with .01 mm. graduation), manufactured by the B. C. Ames Co., Waltham, Massachusetts 02154; or the Cady Pocket Micrometer, Model CPM (with .001” graduation), manufactured by E. J. Cady & Co., Chicago 60635.

[71]

Few bibliographers in the past have utilized micrometer measurements of thickness to distinguish impressions. For an exception, see Matthew J. Bruccoli and Joseph Katz, "Scholarship and Mere Artifacts: The British and Empire Publications of Stephen Crane," SB, XXII (1969), 277-287 (esp. p. 278).

[72]

A micron (μ) is equal to one thousandth of a millimeter and is an appropriate unit for expressing the thickness of paper.

[73]

See Bowers, Principles, p. 446.

[74]

Obviously leaf-cancels can often be located by micrometer measurements also, and one can argue for this reason that micrometer readings should be taken on every leaf of a book, not just on every sheet. One cannot rely on the detection of stubs as the sole method for finding cancels, and chainlines and watermarks are not always present; micrometer measurements are therefore an invaluable aid in approaching the problem of cancels in machine-produced books printed on wove paper — a problem set forth by John Carter in "Some Bibliographical Agenda," in Nineteenth-Century English Books (1952), pp. 68-70.

[75]

See G. T. Tanselle, "A System of Color Identification for Bibliographical Description," SB, XX (1967), 203-234.

[76]

Judd, "Systematic Color Designations for Paper," Paper Trade Journal, CXI (17 October 1940), TS201-TS206.

[77]

For a brief list of works on this subject, see SB, XX (1967), 233.

[78]

Definitions of these terms can be found in The Dictionary of Paper (3rd ed., 1965).

[79]

This series is suggested by William Bond Wheelwright in "Identification of Paper Samples," Paper & Printing Digest, V (September 1939), 3-9.

[80]

Among the twentieth-century specimen books which offer samples of various finishes, two comprehensive ones are A Book of Samples of Paper (Champion Paper Co., 1922) and, especially, Paper for Books: A Comprehensive Survey of the Various Types of Paper Used in Book Production (Robert Horne & Co., 1953, 1961).

[81]

Of course, if the format of a book cannot be determined and the dimensions of the leaf are given in the collation paragraph, they need not be repeated here.

[82]

The treatment of the edges (stained, rough-trimmed, and so on) is more appropriately specified in the paragraph on binding, since it is not a characteristic of the paper but represents something done to the paper in the process of binding.

[83]

Bulk measurements which include the covers and endpapers are cited for this book in Merle Johnson, American First Editions, ed. Jacob Blanck (4th ed., 1942), p. 401. Higginson's Graves bibliography reports two figures for bulk (as "1.8/2.3 cm."), one for the sheets and endpapers and another for the sheets, endpapers, and covers. However, bulk measurements should properly include only the sheets; if in rare instances a variation in the thickness of the covers is significant, the information can be recorded in the paragraph on binding.

[84]

The idea of a series of levels to represent the spectrum of possible details for investigation was suggested by Kenneth L. Kelly's use of this device in "A Universal Color Language," Color Engineering, III (March-April 1965), 2-7; further comments on its usefulness in descriptive bibliography are found in the essay on "Tolerances" cited above (note 37).

[85]

The degree of accuracy of measurements is a separate question from the quantity of detail. Generally speaking, dimensions of sheets and leaves, as well as total bulk, should be reported to the nearest millimeter (or the nearest thirty-second of an inch) and thicknesses of paper to the nearest thousandth of a millimeter (or of an inch).

[86]

The first of these descriptions is from Philip Gaskell, John Baskerville: A Bibliography (1959), entry 37; the second is from Allan Stevenson, Catalogue of Botanical Books in the Collection of Rachel McMasters Miller Hunt, II (1961), entry 466.

[87]

See, for example, Strength and Other Characteristics of Book Papers, 1800-1899 (1967), Publication No. 5 in the series on Permanence/Durability of the Book (1963- ), issued by the W. J. Barrow Research Laboratory of Richmond, Va.

[88]

A recent example occurs in The National Union Catalog: Pre-1956 Imprints (1968- ), where the paper on which the work is printed is described in this fashion: "Substance 89 gsm / pH cold extract 9.4 / Fold endurance (MIT ½ kg. tension) 1200 / Tear resistance (Elmendorf) 73 (or 67 x 3) / Opacity 90.3%."

[89]

I am grateful to the late Allan Stevenson and to Dr. Philip Gaskell for their generosity in taking time to read a manuscript version of this article.