Maps bound in the Atlas
The maps bound in the atlas are listed in the collation in the order that they appear. Approximately half of the maps — nineteen out of thirty-seven — are by John Speed. They include county maps from his Theatre and foreign maps from his Prospect; in all cases they lack text on the reverse. The county maps are in their second state with the original imprints, usually of Sudbury and Humble, but with the insertion of the British tribal names and other corrections made after 1614. This state was maintained throughout the next fifty years, up to and including the last (William) Humble edition in the mid 1650s.


The foreign Speed maps are identical with those in his Prospect, with borders of town views and costumed figures. As the map of Ireland is dated 1651 with the rare imprint of William Humble this suggests that the maps come from that issue of the Prospect or the following edition of 1652. For the next two editions the maps of France and the Low Countries have added to them the dates of 1653 and 1654 whereas the maps in the atlas retain their original dates of 1626. The county maps could well be remainder maps from the 1646 printing or later issues by William Humble when (on both occasions) county maps (but not, as far as is known, the foreign maps) were printed without text.

The latest date on any map bound in the atlas is 1656 — on Visscher's map of Prussia — indicating a date no earlier than this for the atlas' compilation. Other dated maps cover the period 1640-1647, such as Cornelis Danckerts' Italy 1640, Edward Wright's Azores printed by Peter Stent 1645, three Web-Saxtons 1642-45, the Siege of Newark also printed by Stent 1646, and France 1647; these dates however cannot be evidence of earlier compilation. Several rare foreign maps such as Jodocus Hondius' Germany 1607, Willem Lodewijcksz' East Indies 1599, the Petrus Plancius - Paul de la Houve Italy circa 1610, Claes Visscher's Rhine Palatinate 1621, and William Baffin's Moghul Empire 1632 were presumably occasional loose sheets or taken from old plates, perhaps in anticipation of re-issue.


The Loose Maps
The eighty-five loose maps total more than those bound in the atlas but are not quite so heterogeneous in origin. There are a further thirty-three Speed county maps making virtually a full set of all English counties. One map, the Isle of Man, is dated 1605 and is one of the rare proofs issued without text prior to the first 1611/12 edition. Others are representative of various states: from the first edition (one map, Dorset), state 2 (fourteen maps), state 3 (fifteen maps with the imprint of the Roger Reas with some maps re-dated 1662), state 4 (two maps, Derbyshire and Sussex with the date amended to 1666). Berkshire is by Wenzel Hollar also dated 1666 and Norfolk is the later plate by John Goddard.

There are six maps from the William Smith series of 1602-03, all in their second state with the imprint 'Printed and sold by Peter Stent'. The British Library and Royal Geographical Society each possess some but not all of the maps in their first state. The imprint of Peter Stent is noted by Skelton as being 'extremely scarce' and he ascribes their date to circa 1650. John Overton later acquired the plates from Stent and reprinted the maps with his own imprint; they are much more frequently found in this third state, either loose or bound in one of the several Overton atlases. Stent's imprint is also to be found on John Norden's small town plan of Elizabethan London, originally published in 1593. Two post-Eire prints of London bear John Overton's imprint.

Another work by Norden in the collection is his original map of Surrey, 1594. This map is now in the British Library. It might be expected that this would be a re-issue by Peter Stent made in the 1640s in common with others of the period such as Norden's Hampshire, his London (above) or Symonson's Kent. However, armorial evidence shows that Norden's original plate of Surrey was altered shortly after 1603, and broadside lists of Stent's stock made no mention of any plates or maps of this county by Norden.

The thirty-four foreign maps are mostly by Blaeu, Jansson or Visscher. Several of the maps by the Visscher family have characteristic costumed figures and decorative borders. Some, but not all, of the loose foreign maps have an extended left hand margin and are numbered in manuscript as if it was intended that they should be systematically bound as part of the atlas.


Questions and tentative conclusions
It is difficult to establish a rationale for such a wide variety of mapmakers, locations and issues as found in this atlas. It would appear to have been compiled in the 1650s, and some of the loose maps also belong to this period. But others clearly date from just pre- and post- the great Fire of London in 1665. One or two (such as Overton's London of 1676 and the Greene-Berry map of Scotland of 1679) may have joined the collection later.

The sequence of maps bound in the atlas, as listed in the Appendix, seems to be part of no logical plan. Is there any significance in the selection of Speed maps from the Home Counties only? Why do all of the maps lack text, and have any other Prospect maps without text been recorded? Was the atlas put together to form a book of proofs or samples, or was it some sort of international 'campaigning' atlas? The example of a Speed atlas of circa 1646 (also without text) bound in a vellum roll for case of travel and at one time in the Gardner collection suggests a precedent.

The only contemporary publishers' imprints to occur are those of Peter Stent, C.J. Visscher, and (on certain maps) Stent's successor John Overton. Stent was active from about 1642 until his death from plague in 1665, and he was in the possession of George Humble's stock of plates as well as a number of plates which originated at the turn of the century. His published catalogues of stock (c1655 and 1662) include several of the maps found in the atlas or loose, among them 'Mr. Wright's ... Sea Chart for the Hand Azores', prints by Hollar who worked for him, and twenty plates of Wilton Gardens. Stent's address is usually recorded as the White Horse in Giltspur Street, although it is his earlier shop (pre-1646) that is recorded on his siege plan of Newark. On the Azores map, and the Smith maps it is just 'London'. The likely date of the atlas and certain of the loose maps tie in with Stent's years of activity throughout the late 1650's and it is a reasonable hypothesis that they were collected by him. He may also have been responsible for binding some of the maps into an atlas, or alternatively, this may have been done according to the wish of a purchaser.

In the collection are sixteen maps by Claes Janszoon Visscher, the Amsterdam-based mapmaker and publisher. The senior member of the family was active from 1620 until his death in 1652 when his son continued the family business under the same name. The Visschcrs' maps are more often found in composite atlases and although no formal connections have been recorded between the Visscher family and Stent or Overton there seems to have been a close trading relationship and, perhaps, an exchange of plates or surplus maps in the expectation that Stent or his successors might develop a new type of composite or international atlas. Alternatively the Visscher and other foreign maps may merely have been collected to reproduce or plagiarise, as was only too common a practice among rival publishers of all countries in the seventeenth century.

John Overton took over Stent's stock and shop after the hitter's death in 1665. A number of the loose maps come from the rather confused period just before and after the Fire of 1666; for instance the Visscher and other foreign maps first published 1662-64, the Rea Speeds of 1665, the two amended Speeds of 1666, and a Dunstall print of London of circa 1661. A few years later, in about 1670, Overton produced the first of his composite atlases including a number of maps paralleling those in this collection but now with his own imprint. Presumably Overton would have preferred to present a consistent set of Speed county maps, had because of the Fire only relatively few sheets of the last (Rea) Speed issue were left and there was no alternative but to use Blaeu or Smith maps, and subsequently those of Jansson, in lieu.


Conclusions
The evidence suggests that:

1 The main atlas was compiled some time in the late 1650s, possibly by Peter Stent, using sheets from earlier plates. Whatever the actual date, the origins of several of the maps remain uncertain.

2 Some of the loose maps may also come from the period when Stent was active but others date from the time after the transfer of his stock to Overton in 1665.

3 A few loose maps are a decade or so later and these, with some of the views, were evidently added subsequently.

The most likely possibility is that the collection initially represented a sample or proof atlas which was assembled by Stent but which for some reason was discarded. Then, when Overton took over, the atlas was built up as an essay for the type of composite atlas which Overton finally succeeded in bringing out in limited form around 1670. Some maps and views may have been added later but the collection was put aside, remaining more or less intact. It passed out of Overton's hands undisturbed by subsequent ownership until it was found at Jaggards in Stratford-upon-Avon and re-studied nearly three centuries later.

There still remain a number of queries about individual maps and the collection as a whole. It also seems possible that there are in existence other similar collections which have not been recorded and readers comments will be welcomed by the author.

The collation is as follows:

MAPS BOUND IN THE ATLAS

1 John Speed, '...The World...' 395 x 510mm.
A world map presumed once present by evidence of a remaining shred of margin.
[several blank pages]

2 John Speed, 'Europe...' 395 x 510mm.
This map (and items 3, 4, 6, 10, 29, 30, 33, 34 and 36) is from Speed's Prospect first published in 1627. None of these maps has text on the reverse.

3 John Speed, 'Africae...' 395 x 510mm.

4 John Speed, 'America...' 395 x 515mm.

5 Edward Wright, 'A Particular Sea Chart for The Hands Azores' [Imprint] 'London printed and sold by Peter Stent 1645.' 400 x 520mm. Wright's chart was originally prepared for his book Certaine Errors in Navigation, an important landmark in the history of theoretical navigation which was first published in 1599. The only other copy of this state with Stent's imprint is (as far as is known) in the Bodleian Library. Oxford. A later version of the chart with a different title and re-drawn cartouche appeared in 1657. See E.J.S. Parsons and W.F. Morris, Edward Wright and his Work, Imago Mundi III, 1939.
[two blank pages]