¹ Albion 20, 3 (Fall 1988): 387–401.Google Scholar

² The two manuscript catalogues bear the British Library shelfmarks C.38. h.21 and C.37. h.13. The former contains Thomason's preface to the collection and other documentary materials. Fortescue thought the preface a derivative of an earlier, lost document by Thomason; see Fortescue, G. K., ed., Catalogue of the Pamphlets, Books, Newspapers, and Manuscripts Relating to the Civil War, the Commonwealth, and the Restoration, Collected by George Thomason, 1640–1661, 2 vols. (London, 1908), 1: xviGoogle Scholar. But Lois Spencer established its authenticity (“The Professional and Literary Connexions of George Thomason,” The Library, 5th ser., 13 (1958): 102–118)Google Scholar. The latter manuscript catalogue, bearing the title “Catalogue of the King's Pamphlets” (thus explaining the currency of that title among scholars before Fortescue's redaction), contains marginal notes indicating the disposition of every item in the collection according to the modern bindings and shelfmarks. One wonders if Greenberg used this catalogue in light of his remark (p. 388) about the “difficulty” of locating items listed in the catalogue by titles not identical with those on the title page.

³ Spencer, Lois, “The Professional and Literary Connexions of George Thomason,” and “The Politics of George Thomason,” The Library, 5th ser., 14 (1959): 11–27.CrossRefGoogle Scholar

⁴ When, for example, Greenberg writes (p. 395) that “only” the manuscript allows for his dating exercises he is incorrect; it is only the collection that makes it possible.

⁵ Spencer, , “The Professional and Literary Connexions of George Thomason,” p. 114.Google Scholar

⁶ Greenberg, , “Dating Civil War Pamphlets,” p. 388.Google Scholar

⁷ It is possible that some sort of acquisition register was maintained. But final cataloguing could not have begun until binding. A strong indication that no such general acquisitions list was compiled for the period of undated tracts is the fact that the place order, though an invaluable aid to dating, is far from perfect. Any sort of contemporary register most likely would have led to a tighter chronology.

⁸ Thomason's famous hand-written dates become frequent in summer 1642. Thomason's practice paralleled and may have been suggested by that of many printers and publishers at this time. Indeed for some weeks in summer 1642, most pamphlets bore day-dates in the imprint — viz., dates of purported publication, not of the events noted or commented upon. Greenberg would have done to well to analyze time lags of these dates: for London-based events, overnight publication is not uncommon, and 24- to 48-hour lags are the norm.

⁹ The earliest items in the collection were actually published in 1637 (they are not noted in Fortescue's catalogue), the latest in 1663. When Thomason actually began to make his collection is unclear — but I believe he did not begin to collect until 1641.

¹⁰ Spencer, , “The Professional and Literary Connexions of George Thomason,” p. 114.Google Scholar

¹¹ For tracts, Fortescue usually but not invariably used the earliest applicable date. Thomason's own written dates had secondary value, sometimes appearing as dates enclosed in parentheses at the end of a Fortescue entry. Fortescue or his assistants sometimes, but inconsistently, seem to have used place order in the collection as a guide to the dates of items before July 1642. For newsbooks Fortescue used the last applicable date (the claimed date of issue); Thomason's own dates are usually in agreement. This has bearing on Greenberg's argument, since as the proportion of newsbooks increases (as well as the number of short-term news separates), the two catalogues converge. In fact, if Greenberg had counted newsbooks correctly (see below), the Thomason and Fortescue curves in Figures B and C would have been quite similar.

¹² For the arrangement of the undated tracts, see n. 17 below.

¹³ The density of the collection at a given time affects the degree of confidence one can place in the place order as the order of acquisition. Nor can place order be used apart from a general knowledge of the period and a “feel” for the collection. Thomason obviously shuffled through his books, and it would be wrong to place a great deal of emphasis upon individual placements. Yet even when the pamphlets are few, it is obvious that the pamphlets in E. 203 precede those of E. 204 (Thomason's first two volumes of ordinary small quarto tracts), and so on. By April or May 1642, similarly contiguous volumes (e.g., E. 142-E. 148) contain about a fortnight's acquisitions; much closer estimates of acquisition dates are possible. It should also be noted that serious diver-gences from probable publication order occur far more commonly with earlier tracts appearing anomalously in a later volume than with later items appearing amidst earlier ones. This suggests that gross filing errors are relatively few, since the case of an early tract appearing among later items is also to be explained as a late acquisition.

¹⁴ Since the appearance of the University Microfilm edition, the once familiar bindings are now a good deal less well known to researchers, even at the British Library. For a photograph of a binding, see Williams, Franklin B. Jr., “Five Lost Thomason Tracts Come Home,” The Library, 5th ser., 19 (1964): 230–34.CrossRefGoogle Scholar

¹⁵ Readers note that the E. numbers of the quartos run in chronological batches and then shift arbitrarily. For an example, see n. 17 below. It is not hard to guess the cause of this pattern. When the volumes, once arranged in strict numerical order, were placed on the shelves that defined their shelfmarks, this was done, in effect, by picking up a batch (perhaps a shelf's worth) of contiguous (hence consecutively numbered) titles, but without any effort being made to restore exactly the original order.

¹⁶ For the pre-25 March 1647 items, Thomason originally established a single numbered series for each legal year, just as he did for the later items (these numbers are the ones in ink at the top right corner of each item; the modern numbers are in pencil, and use the 669 shelfmark). Later, these legal-year series were broken into two parts. One was for messages, orders, and other publications emanating from public bodies. The other contained items emanating from private sources, ranging from ballads to individual and group petitions to public bodies. Thus 669 f.3 contains “public” items acquired up to 25 March 1642; 669 f.4 is the parallel “private” volume. 669 f.5 and 669 f.6, 669 f.7 and 669 f.8 are similar pairs, for the legal years ending March 1643 and March 1644, respectively. 669 f.9 and 669 f.10 each contain three separate legal-year series, stretching through March 1647. In addition, it should be noted that for some reason a few items in the folio Acts series and the broadside series migrated to the other category. For the exceptional volume 669 f. 18, see n. 17 below.

¹⁷ So the reader can judge whether Thomason's practices are “too sketchy for use,” it might help simply to state the arrangement of the small quartos for the period through July 1642. Within each listing, items are in essentially in order of acquisition: E. 196-E. 200, Speeches (and a few other parliamentary items) to October 1642; E. 201 -E. 202, Newsbooks to early October 1642; E. 203-E. 208, General tracts from 1637 to May 1641; E. 156-E. 181, General tracts from May 1641 to Jan. 1642; E. 131-E. 155, General tracts (and newsbooks beginning July 1642) from Jan. 1642 to July 1642. In addition, the pertinent large quartos are in E. 238 and E. 239, some Acts in E. 1059, the folio sheets in 669 f.3–669 f.6, and one folio book in 669 f.18, a special folio volume for this rare sort of item, although some multi-sheet folios are also amidst the main series. For the octavos, see the catalogues. I do not claim that every item is exactly in its date-determined place (there is some obvious “local” shuffling). And some items are grossly out of place: in any filing system errors will be made. But the general pattern is unmistakable, and can only be ignored to knowledge's peril. (I wish to thank Lynn Johnson for her assitance with this note.)

¹⁸ P. 389, n.7.

¹⁹ The most recent count, surprisingly not employed by Greenberg, is that made by University Microfilms, which reported 23,926 items in 2142 volumes (to Fortescue's 22,225 in 2008 volumes); see The Thomason Tracts 1640–1661. an Index to the Microfilm Edition of the Thomason Collection of the British Library (Ann Arbor, 1981), Part 1, p. vGoogle Scholar. I cannot determine whether Fortescue's volume count relied upon the original or the more numerous split, modern volumes. While Fortescue undercounted the tracts, largely by simple omission and apparent loss of entries, University Microfilm's better methods are still not beyond question, and might be revised downward by 1–2%. To arrive at the original number of titles, one would have to add to the University Microfilm figure or a revision of it the lost titles. Lois Spencer thought these more numerous than the 141 allowed for by Fortescue. Thomason's apparently off-the-cuff estimate of “neere Thirty Thousand severall peeces” cannot be taken as an exact count of the items in his catalogue. See Spencer, , “The Professional and Literary Connexions of George Thomason,” pp. 104, 114.Google Scholar

²⁰ I have had to estimate Greenberg's results from his Figures A, B, and C. For Fortescue's totals, see the Catalogue, 1: xviGoogle Scholar. I cannot find any other way to generate Greenberg's monthly totals than by assuming he made the error described above, which is easy make if one works too quickly with Fortescue's listing of newsbooks in vol. 2 of the Catalogue. That Greenberg did count in this way is clear from his statement (p. 389) that Thomason collected 139 newsbooks from July 1641 to June 1643; that is exactly the number of titles in Fortescue's listing.

²¹ If Thomason's catalogue were indeed the source of Greenberg's data, it would indicate only too strongly the secondary, derivative nature of the catalogue. However, we would be forced to believe that Thomason or his clerks lost most but not all July entries from between 8–10 volumes, while evidently preserving the June or August and a few July entries in the same volumes — all while the tracts were preserved and bound in the usual manner.

²² On the other hand, the same microfilm edition has made available its own computer-generated index to the collection (see n. 11 above). To an informed user, that listing is perhaps the best available synoptic guide to the structure of the collection.

²³ The often wished-for listing of itmes Thomason supposedly missed is a partly futile and, now, a redundant labor: futile, because Thomason is usually accused to having missed itmes which he did purchase, but which have been lost; redundant, since the wing STC bibliographers' now-computerized data base contains the material for a publication-year listing, as well as the capacity (should one wish it) to “bleed off” the Thomason entries. I wish to thank Caroline Nelson for describing the data base to me.

²⁴ From British Library C.38. h.21, vol. 1.