III. Cartographic Notes

There are several summaries of the different editions and versions of Mitchell's Map, all of which are different! This page explains how the different listings relate to each other. While I have seen many different Mitchell Maps, I have not examined instances of all the variations listed here; the identification of editions and issues is based on the published data.

This listing includes only the London editions of Mitchell's Map. For information concerning the Dutch, French, and Italian editions see Martin (1972, 110-13).

This listing has been superceded by the appendix to Matthew H. Edney, “A Publishing History of John Mitchell’s 1755 Map of North America,” Cartographic Perspectives, no. 58 (2007): 4-27 and 71-75, esp. 18-21.

The Physical Artifact: Size, Storage, and Color

The basic character of Mitchell's map is readily apparent. It was engraved on eight sheets of copper, measuring 136 cm (4'5") high by 195 cm (6'5") wide when assembled. As with other wall maps, there are relatively few surviving copies. As a large, expensive product, relatively few copies would have been made. Furthermore, the act of hanging is seriously destructive: maps fade with long exposure to light; they tear under their own weight.

Surviving maps fall into two categories. There are those which were never assembled into the whole but which were left as single sheets. Many of this group were bound into atlases factices; we might point for example to the two copies of the French editions of Mitchell's map in the Osher Map Library's Smith Collection, both of which are bound into atlases. The second group of surviving maps are those which were dissected into smaller pieces, mounted on coarse cloth, and folded for storage in a case. The National Archives has a third edition, second issue of the map (Record Group 76, Cartographic Series 27, Map 3; see Goggin 1968, item 18) each of whose sheets was dissected into 20 sections and mounted separately; small leather tags on the back of the sheets indicate that they were all folded and placed into a small box, the tags being used to pull out particular sheets from the tight mass. The Osher Map Library's fourth edition was dissected into 32 sections (each sheet being quartered), mounted on a single piece of cloth, and was stored in a large case (separately conserved).

Like any map from the period, Mitchell's map was printed in black ink on white, hand-laid paper. Any color that appears on the map was applied with water-colors by hand. The surviving copies seem to be equally divided between uncolored, colored in outline (the political boundaries highlighted with thin ribbons of color), and washed color (each political unit entirely filled with color). Colors were generally applied at the time of printing -- publishers sold their maps both colored and uncolored -- but many maps have been given their color in the nineteenth century.

The copy of the fourth edition of Mitchell's map on display at the Osher Map Library received a wash of water-colors probably between 1773/74 and the end of the eighteenth century (discussed elsewhere). The color wash was applied to each sheet separately, as the colors are not all continuous across the edges of the sheets; the sheets were dissected after coloring.

Printing from Copper: "States" and "Editions"

The technology of printing from copper plates into which an image has been engraved was in many respects rather limiting. It did however permit the easy alteration of the images. An engraver needed only to place the plate face-down and hammer out the section of the plate to be changed; the plate could then be re-engraved. The importance of this fact for the production of maps was profound. It allowed maps to be updated with new data, and sometimes extensive areas of a plate would be re-engraved. It also allowed plates to be traded between map publishers; the publishers would add their own name and then re-issue the plate. Plates could thus enjoy a long life -- sometimes more than a century -- with both small and large changes to the images they bore.

Cartographic historians have employed various terms to help them in their discussions of the changes which particular maps have undergone.

When dealing with a map engraved on a single piece of copper, the terminology is straightforward. We refer then to "states" of "plates." Each set of changes made to a plate prior to a new printing constitutes a new state; each fresh engraving of the map image is a new plate.

Multi-sheet maps, such as Mitchell's, present a far more complex situation. To be most precise, we should consider an entire map as being an assemblage of different plates, each of which has a specific history of states. The work necessary to track state changes in large maps means that such precision is difficult to achieve. Certainly it has not been achieved with respect to the Mitchell Map. On the other hand, multi-sheet maps have tended to be issued as a set, and changes in one plate have tended to be matched with changes on others. Cartobibliographers therefore tend to refer to "issues" or "editions" of the whole map. A greater sense of precision is attained by distinguishing between "editions" and "impressions," an "edition" being defined as a change in publisher or title, an "impression" being any other change. It is important to realize that the changes entailed in a new "edition" might easily be far less than those in a new "impression."

The cartobibliographic listings for Mitchell's map have used this terminology in somewhat different ways. Mitchell himself distinguished a "second edition" of his map when he had made some alterations to the coastline of New England and Nova Scotia (in 1755-57). In contrast, Fite & Freeman (1926) used "second edition" to describe the 1775 change in the map's title. Stevens & Tree (1980, 86-87) in part use any change to the publisher's imprint, no matter how minor, to signify a new edition.

This summary follows the numbering of "editions" defined by Martin (1972, 109-10). My reasons are: (a) I am of the opinion that Stevens & Tree (1980, 86-87) are just a bit too precise; and (b) I suspect that there might be an error in their listing (see below under Third Edition). I prefer the term "issue" to "impression" to refer to varients within an edition.

1st Edition 1st Issue 1755

[title] "A Map of the British and French Dominions in North America with the Roads, Distances, Limits and Extent of the Settlements, Humbly Inscribed to the Right Honourable The Earl of Halifax, And the other Right Honourable The Lords Commissioners for Trade and Plantations, by their Lordships Most Obliged and very humble Servant Jno. Mitchell."

[inside bottom margin] "Tho: Kitchen Sculp. Clerkenwell Green."

[outside bottom margin] "Publish'd by the Author Febry 13th 1755 according to Act of Parliament, and Sold by And: Miller opposite Katherine Street in the Strand."

Both "Miller" and "Katherine" in the imprint are misspelled; Worcester, Mass., is improperly labelled "Leicester," in addition to the town of that name.

Reproduced in Sellers & Van Ee (1981, 12).

NOTE: Schwartz & Ehrenberg (1980, 164) reproduce a very handsome example -- in color -- of the Mitchell Map. They state that the reproduction is of the map's "third edition" although the map's lack of the large text blocks added with the second edition (below) would indicate that this is incorrect. It is possible that they intended to mean the third issue (Martin's "impression") of the first edition. Unfortunately, the reproduction is too small for the issue (one to three) to be discerned. Some of the confusion might have derived from the existence in the Public Record Office, Kew, of a first edition that has been called "the red line map (so called)" (Stevens 1897) and "the English 'Red-Line' Map" (Martin nd); this has a red chalk line marking a boundary.

1st Edition 2nd Issue 1755

[inside bottom margin] as 1st edition, 1st issue

[outside bottom margin] "Publish'd by the Author Febry 13th 1755 according to Act of Parliament, and Sold by And: Millar opposite Katharine Street in the Strand."

The "Millar" and "Katharine" in the imprint are now correctly spelled.

No changes are recorded in the geographical detail. Worcester, Mass., is still improperly labelled "Leicester," in addition to the town of that name. Such minor changes probably indicate tinkering to the eighth plate during 1755.

1st Edition 3rd Issue 1755

[title] as 1st edition, 1st issue

[inside bottom margin] as 1st edition, 1st issue

[outside bottom margin] as 1st edition, 2nd issue

Again, only one minor change, indicating continued tinkering in 1755. Specifically, Worcester is now correctly named in Massachusetts.

NOTE 1: Stevens & Tree (1980) make no mention of this issue, presumably because they paid attention only to map titles, imprints, and gross geographical changes.

2nd Edition 1755-57

[title] as 1st edition, 1st issue

[inside bottom margin] as 1st edition, 1st issue

[outside bottom margin] as 1st edition, 2nd issue

The title and copyright dates have not been changed, but Mitchell added two large text blocks in the middle of the Atlantic Ocean, in the seventh sheet. (The two scale bars in the Atlantic on the first edition were re-engraved -- as four bars -- above the cartouche on the eighth sheet.) The upper text block, in three columns, is a statement about the changes which Mitchell made to the coastal areas of New England, Nova Scotia, and Newfoundland as a result of receiving new data. In the first column, Mitchell explicitly refers to this map as the "second edition." The lower text block, in seven columns, is a listing of the data Mitchell used to structure -- and to restructure -- the map.

Other changes include the addition of observations of magnetic variations off the Atlantic coast, labeled with Roman numerals. More importantly, Mitchell redrew the northeastern coast because he redefined the positions of two key headlands: Cape Race was shifted in latitude from 46°55' to 46°35'; Cape Sable was shifted in longitude from 66°35' to 65°35'. A Maine-related detail: "Sagadahook," as on the first edition, was respelled "Sagadahock."

NOTE 1: with respect to dating, the copyright date remains unchanged. Stevens & Tree (1980) give a date of 1757, citing the British Library's catalogue as evidence; that catalogue is however uncertain, and dates the map to "1755 or 57?" The same catalogue's suggestions for dates of later editions are incorrect. I tend to favor 1757 as a publication date for reasons discussed in the narrative account. I nonetheless go with a date range of 1755-57 to highlight the uncertainty.

NOTE 2: Fite & Freeman (1926, 180) originally reproduced one of the third editions, as referred to in the text (p.182); production problems meant that the 1969 Dover reprint (see p. vi) actually reproduces the second edition, although the text has not been changed to reflect this.

Transcription of the upper text block

[column 1]

After the first Drawing of this Map in 1750, it was again corrected and improved before it was published, and I have since taken Care to procure & examine all the Information I could get, in order to render it as correct & usefull as possible; which has given occasion to this Second Edition of it; in which I have likewise inserted all the Observations I beleive we have for the Geography of N. America, since I find them grossly misrepresented by others.

The Foundation of this Map is the several manuscript Maps, Charts, & Surveys that have been lately made of our Colonies, which represent most Places from the Ocean to the Missisipi. But in order to know the true Situation of those Places, we must have their Latitudes & Longitudes, which are of much more Consequence in a general Map, than their bare Shape or Figure, which we only find represented in our Draughts & Surveys. But after having consulted all the Observations I beleive that we have, I found the true Situation, or Latitude & Longitude, of many Places was undetermined or uncertain; & that in the principal Parts on the Coast, and that we had no Accounts of them, but what might be found in the Journals of our Ships of War kept in the Admiralty Office; which I had recourse to for that reason, and have extracted from them whatever relates to our purpose, which are the chief Source of the Corrections & observations here inserted.

Since the Publication of this Map likewise I have examined & compared with other Accounts, the Observations of Mr. Chabert; which were not made when our Map was first drawn, nor known in England till after it was published; so that we neither followed nor rejected them.

From these authorities we find but two Alterations necessary in our Map: 1. In the Latitude of Cape Race: 2. In the Longitude of Cape Sable.

I Cape Race was laid down in Lat. 46°.55' from the Surveys of Capt. Gaudy & Durrell, who make it in Lat. 47°.2' & 46°.57'; and the Observations of Bellin, which make it in 46°.50'. But by al the accounts in our Ships of War, confirmed by those of Chabert, it is in Lat. 46°.30', or 35' at most: from which and the other Observations here inserted, we have corrected that and other Parts of Newfoundland.

[column 2]

In the Longitude of Cape Race, different Reckonings in or Ships from Cape Breton, Canso, & Boston, differ 2 Degrees (on Accunt of the Currents), and more than Reckonings from Europe. At a Medium they make it in Longt. 53°.57'. Durrell’s Survey makes it in 64°.9' & Chabert’s in 53°.2' by its Distance from C. Raye: the Mean between them is 53°.35'.30": which is exactly the Mean o the Reckonings of our Ships of War from England.

C. St. Mary’s they make 1°.20' West of C. Race; and St. John’s Harbour 35' or 40' East of it.

II For the Longitude of Cape Sable we had no Certain Authorities, before the Observations of Chabert, and the Reckonings of our Ships here inserted, which make it in 65°.35' or 33'; agreeing within 2 min.; by which we have altered that and consequently the adjacent Parts of Nova Scotia and New England.

Capt. Durrell in his Survey of this Coast makes C. Sable a Degree farther West from Canso, from which it was before laid down: and Chabert observes, that the French manuscript Charts & others, do the same. Our new Maps, blindly copyed from Chabert’s, that make C. Sable in Longt. 65°.25' (according to Chabert’s Reckonings, instead of his Observations), and Boston in 71°, make their Diff. Longt. 23' greater, we see, than our Ships make it.

III The rest of the Coast agrees with the Accounts of our Mariners, being laid down from Charts made from their Observations, as the Coasts of most Parts of the World are: which are fixt Points, from whence the Situation of the Inland Parts is chiefly deduced.

The Map of New England & Nova Scotia requires a farther Consideration, as we find them very erroneously laid down in all our Maps & Charts, especially our many New Maps, copyed from a New Map of Nova Scotia, copyed from Popple and d’Anville: and those Errors are maintained by Arguments & pretended authorities, which seem to have confirmed them.

The only Authority they have for all this is a feigned Survey by a pretended Surveyor General Blackmore in 1711-12: who appears by his Journals to have been Lieutenant of the Dragon Man of War 1711, and made a rude Draught of this Coast (as

[third column]

well as he remembered it perhaps) in 1715, with a Petition to the Board of Trade to enable him to Survey it at that time, which he never did as we can learn. But this draught falling into the Hands of some Workmen, Mr. Moll published it as an Actual Survey “made by her Majesties especial Command,” from which this Coast has been thus erroneously laid down ever since.

It is thus laid down from the Supposition that Mt. Desart Rock lies from C. Anne N.N.E. 30 Leags. and W. by S. 3 Leagues from Annapolis: from whence they make Mt. Desart I. in Lat. 44°.45' and the Rock in 44°.25'. But Surely it is well known that the Coast of New England runs rather E.N.E. than N.N.E., and that Mt. Desart is much farther from C. Anne than from Annapolis. Our Ships of War make Mt. Desart I. about N.E. 6 E. 50 Leags. from C. Anne. These observations then appear to have been made, as well as published, by Mr. Moll. There are no such Observations in Mr. Blackmore’s original Chart in the Plantation office. Mr. Blackmore observed in the Dragon Holt I. in Lat. 44°.1', the Leostaff in Company makes it the same, and the S.W. End of Gr. Manan in 44°.39'; Govr. Dudley makes Pemmaquid Ft. in Lat. 43°.55'; Mr. Richer observed Pentagouet Ft. in 44°.22'; and Bellin laies down Mt. Desart Rock from observation in 44°.5'; all which observations agree with one another, and with the Journals & Reckonings of our Ships of War; whence Mt. Desart I. can never be in Lat. 44°.45'; nor the rest of that Coast so far North as this makes it.

By this they make the whole Coast of New England from 10 to 25 Min. of Latitude too far North, which leads Ships upon that dangerous Coast of Mt. Desart. Thence to Long I. they make the Course E.S.E. nigh 30 Leags. which appears to be about due East 20 Leags. Long I. & Gr. Manan, the two big Land-marks by which Ships steer, they make 50 & 60 Miles asunder instead of 29 or 30. By placing Quebec at the same time in Longt. 70°.35', they make the whole Country nigh a degree of Longitude & half a degree of Latitude out of its trrue Situation, which strangely misrepresents the whole.

In short, we do not find a single spot, hardly, justly laid down in these our New Maps of Nova Scotia, altho they were rightly laid down formerly by de L’Isle & others. We are so far from improving then in the Geography of America that we see it made worse & worse for want of certain Observations, which we have endeavoured to collect & thus to represent in one view.

3rd Edition ? Issue 1773-75

[title] as 1st edition, 1st issue

[inside bottom margin 1] "Tho: Kitchen Sculp."

[inside bottom margin 2] "Printed by Jefferys and Faden, St. Martins Lane, Charing Cross, London."

[outside bottom margin] "Publish'd by the Author Febry 13th 1755."

This version of the map is recorded only by Stevens & Tree (1980, no.54, 86-87). My presumption is that this entry is the same as the following (first issue), but with an incorrectly transcribed imprint for Jefferys and Faden. The data recorded in this work began with notes which Henry N. Stevens began to collect in 1880; it is possible that an incomplete reference by him was later taken as accurate by Stevens & Tree.

Indeed, Stevens & Tree are not as precise as they should have been: they fail to record the words "according to Act of Parliament" in the remains of the first imprint (outside the lower margin) that persist on the later versions of the map. Furthermore, I am very much taken by the failure of Lawrence Martin to record such a version in his extensive researches of the map in the early twentieth century (Martin 1972).

Lacking any census of surviving copies of the Mitchell Map, I posted a request in April 1997 to maphist (the cartographic history listserv), asking historians and librarians to bring any copies of this issue to my attention. But none have; on the other hand, overall response to the question was quite slim. This question will have to remain unresolved until a comprehensive census can be undertaken.

3rd Edition 1st Issue 1773-75

[title] as 1st edition, 1st issue

[inside bottom margin 1] "Tho: Kitchen Sculp."

[inside bottom margin 2] "Printed for Jefferys and Faden Geographers to the King at the Corner of St. Martins Lane Charing Cross London"

[outside bottom margin] "Publish'd by the Author Febry 13th 1755 according to Act of Parliament"

The contraction of both the engraver's identification and the copyright statement, and the addition of a new publisher's identification, indicate several factors. In the first place, Thomas Kitchen had moved from Clerkenwell Green to Holburn; furthermore, the plates had been acquired by the firm of Jefferys and Faden, and had been updated, before this publication. Stevens & Tree (1980) note that the firm was first entered into the Rate Book in 1773. This suggests that the change in ownership of the plates probably occured after Mitchell's death in 1768.

In addition to the changes in the imprints, further changes have been made to the geographical details. Some boundaries have been altered and many new place-names added.

NOTE: with respect to dating, the map's copyright date of 1755 remains unchanged. The map would however have appeared between 1773, as just noted, and 1775 (see fourth edition, below). There is no real reason why Stevens & Tree (1980) dated this map to 1773 and the second issue of the edition to 1774. A possible reason for this new edition of Mitchell's map -- the passing of the Quebec Act of 1774, with its redefinition of several borders -- is discussed in the narrative account; if this was indeed the case, it would restrict the dates of both issues to 1774-75.

3rd Edition 2nd Issue 1773-75

[title] as 1st edition, 1st issue

[inside bottom margin 1] as 3rd edition, 1st issue

[inside bottom margin 2] as 3rd edition, 1st issue

[outside bottom margin] as 3rd edition, 1st issue

There are numerous changes, particularly in the realm of boundaries. For example, the straight line marked on earlier maps as the boundary between Canada and the Iroquois (running north of Lake Ontario) has been deleted; a new boundary has been engraved running through Lake Ontario, which for Martin (1972, 109-10) is a key to identifying this issue. Stevens & Tree (1980) also identify lines delimiting territorial claims by both New York and New Jersey as being deleted.

NOTE: The imprecision of Stevens' & Tree's (1980) recording of the third edition, first issue means that they interpreted this issue as representing a new edition, because of the change in the imprint of Jefferys and Faden.

4th Edition 1775

[title] "A Map of the British Colonies in North America ..." remainder as 1st edition, 1st issue

[inside bottom margin 1] as 3rd edition, 1st issue

[inside bottom margin 2] as 3rd edition, 2nd issue

[outside bottom margin] as 3rd edition, 1st issue

Stevens & Tree (1980) are of the opinion that the only change since the third edition, second issue is the new title.

NOTE 1: given that the copyright statement still reads 1755, the dating of this edition to 1775 is based on a 1778 catalogue of Faden's maps which includes "A Map of the British Colonies in North America ... on 8 sheets, 1775, Mitchell." Goss (1990, 130) mistakenly dates this edition to 1785, an impossibility as the map was used in the Treaty of Paris negotiations in 1782; he gives a good color reproduction of George III's "red line" map.

NOTE 2: Fite & Freeman (1926)were rather confused as to the meaning of "second edition": at one place (p.182) they stated that the new edition was marked by the change in imprint to Jefferys and Faden; but elsewhere (290-91) they use the retitling of the map as an indication of a "second edition."

An uncolored copy of this edition is reproduced at full-size in North America (1974).