In the seventeenth century, the desire to test new hypotheses of the physical universe stimulated attempts to determine accurately the dimensions and figure of the earth, and these then became possible through the invention of more precise instruments to make the necessary observations.

These included the telescope, the pendulum clock, and tables of logarithms.

The measurement of an arc on the earth’s surface was the first step, and this, though primarily a geodetic operation, contributed eventually to the advance of cartography.

The first attempt of any value to determine the length of a degree in this way had been made by Snellius in Holland in 1615, but the operation was first carried out with an approach to accuracy in France, where in the latter half of the seventeenth century there was notable scientific activity under the patronage of the “Roi Soleil”, Louis XIV, and the Académie royale des sciences, founded in 1666.

In that country, one of the first highly centralized national states in Europe, there was a growing demand for maps and charts, and a realization that they could only be based satisfactorily on a precise scientific framework.

Maps were required not only for military purposes, but for the proper organization of the extensive road system, the development of internal resources (the aim of men like Colbert), and the general promotion of commerce at home and abroad.

Stages in the creation of the new map of France

The successive stages in the making of the new map of France were (1) the measurement of an arc of the meridian of Paris by the Abbé Picard, 1669-70, by means of a chain of triangles; (2) the extension of the meridian until in 1718 it ran from the Pyrenees to Dunkirk; (3) the first attempts to produce a new map of France by adjusting existing surveys, supplemented by observations for latitude and longitude, to the Paris meridian; (4) the planned survey of the whole country, de novo, based on a complete system of triangulation, which resulted in the celebrated Cassini Survey.

This had been advocated by Picard as early as 1681.

A remarkable contribution to this work was made by the four generations of the Cassini family.

The first, Jean Dominique, who was invited to work at the Paris observatory in 1669, assisted in the determination of the meridian, but his greatest service to cartography was his perfection of a method of determining longitude by observation of the movements of Jupiter’s satellites, a great improvement on the method of lunar eclipses, though the probable error was one kilometre.

After the early work on the meridian, it was resolved to apply the new methods to rectifying the map of France, and Picard with other surveyors, including La Hire, known also for his projection, were sent to survey the coasts.

A map by La Hire, embodying the results, was presented by him to the Académie in 1684, and subsequently published as ‘Carte de France corrigée par Ordre du Roi sur les Observations de Mrs. de l’Académie des Sciences’, 1693.

This showed both the older outline of the coasts and the new, the general result being to shift the western extremity of France one and a half degrees of longitude to the east, in relation to the Paris meridian, and the southern coastline about half a degree of latitude to the north.

The sight of this map is said to have prompted Louis XIV to remark that the survey had cost him more territory than a disastrous campaign.

The second Cassini, Jacques, realizing that any attempts to fit haphazard surveys to the Paris meridian must be unsatisfactory, became the advocate of the complete triangulation of France, and with his son, César François Cassini de Thury, was engaged from 1733 on this extension. The backbone of the triangulation was the “verified” meridian of Paris.

Along this, at intervals of 60,000 toises (rather more than one degree of latitude), perpendiculars were carried geometrically to the east and west, from which the positions of towns and other points of importance were fixed.

This was the origin of the projection now known by Cassini’s name, in which the coordinates of a point are given with reference to a central meridian and the distance along the great circle through the position which cuts the meridian at right angles.

Progress was set out in 1744 on the ‘Nouvelle carte qui comprend les principaux triangles qui servent de fondement a la description géometrique de la France’.

Cassini de Thury had succeeded in obtaining government support for the proposed topographical map on this framework, and the work was begun at the national expense in 1747. Nine years later, however, this was withdrawn, owing to heavy military expenditure.

Cassini at once, without hesitation, took the bold step of assuming entire responsibility for the survey. He received authority to form an association to finance its completion, obtained the necessary support, partly from various provincial États-General who appreciated the value of accurate maps of their provinces, and carried the undertaking almost to completion.

At the time of his death in 1784, Brittany alone remained to be published.

Ultimately after suspension during the Revolutionary period, the work was taken over by the State and completed in 1818. Full details of the undertaking were given by Cassini in his “Description géometrique de la France’, published in 1783.

Instruments and techniques

Improvements in instruments had greatly contributed to an improved standard of mapping. The divided horizontal semicircles in brass were fitted with telescopic alidades, and micrometer reading allowed angles to be observed with considerable accuracy. Beacons, and sometimes lights, were used for observation marks.

The topographic detail was treated more summarily: though the plane table was commonly in use by the ‘ingenieurs géographes’, the body of military surveyors, Cassini’s men who carried out the minor triangulation sketched the details by estimation or by pacing, and worked this up in the office. Often they were content to indicate slopes by the letters D or F (‘douce’ or ‘forte’).

The Cassini map when complete comprised 182 sheets (88 x 55.5 cms.).

The scale was 1:86,400 (i.e. 1 inch to 1.36 miles). In style it is based on a map of the Paris region made in 1678, during the early days of the determination of the meridian, by Du Vivier, and engraved by F. de la Pointe.

It is carefully engraved, the general effect being clean and uncrowded; the great ‘routes’ to Paris are emphasized and named, the larger towns are shown in plan, and a variety of symbols mark smaller settlements, churches, wind and water mills, gallows and other works of man. Forests, with their walks carefully drawn, are conspicuous, as are the residences of nobility and gentry, with their owners’ names.

Only in portraying relief does the map fail notably. In areas of lesser elevation, rivers and streams are depicted as flowing in narrow valleys with the borders hatched, and isolated elevations are only occasionally shown; the general effect therefore is of a vast level plateau dissected by canyon-like valleys.

In the more rugged south and south-east, the result is even less satisfactory; the terrain is depicted in two or more tiers with the usual shading, and long crest lines show up as narrow white bands.

The landforms of any considerable area thus appear curiously unintegrated. It must be remembered, however, that it was not for many years that sufficient determinations of altitude existed for the relief to be represented with any accuracy.

Whatever the defects of the map it is a remarkable monument in the history of cartography, and one which influenced the mapping of many countries. It was not for half a century after its initiation that a comparable undertaking was begun in Britain by the Ordnance Survey.

It was incidentally on the initiative of Cassini de Thury that General Roy was commissioned to co-operate in the cross-Channel triangulation of 1787, and in so doing paved the way for the foundation of the Ordnance Survey.

General cartography in France and external observations

Having traced the history of the Cassini survey, we may now examine the progress of general cartography in France. J. D. Cassini’s method of determining longitude was employed at an early date in fixing positions outside France.

Largely with a view to improving existing marine charts, observers from the last decades of the seventeenth century onwards were sent to various countries in Europe, French Guiana, the West Indies, Africa, and southern and eastern Asia, where in course of time remarkably accurate values were obtained.

From Richer’s observations, for example, the longitude of Cayenne was determined within one degree of its true value.

The earlier results enabled J. D. Cassini in 1682 to sketch out his famous planisphere, which incorporated forty determinations, on a floor of the Paris observatory. This was later engraved with the title ‘Planispherum terrestre’, an issue of which is known from 1694.

These new observations were also the basis of a collection of sea charts covering, on Mercator’s projection, the western coasts of Europe from Norway to Spain; this was ‘Le Neptune francois, ou Atlas nouveau des cartes marines… Revue et mis en ordre par les Sieurs Pene, Cassini et autres’, Paris, 1693.

Guillaume Delisle and the “reformation of cartography”

The man who introduced this new work to the general public, and in doing so effected what has been called the ‘reformation of cartography’, was Guillaume Delisle (1675-1726).

Guillaume was the son of Claude Delisle, a celebrated teacher in his day of history and geography, to whom undoubtedly his son owed much instruction and assistance in his first ventures.

The son also had the benefit of instruction in astronomy from Cassini at the Academy, of which he became an Associate in 1718.

In 1700 Delisle began work as a compiler and publisher of maps, and for the rest of his life was the leader in cartographical progress with an international reputation. In his maps and globes he followed with understanding the progress of the Academy’s work.

Among his first productions was the ‘Mappe-Monde Dressée sur les Observations de Mrs de l’Académie Royale des Sciences’, 1700, a map in two hemispheres on the stereographic projection, carrying further the improvements of the ‘Planispherum terrestre’, and from time to time (e.g. 1724, 1745) amended versions were published.

If compared with a modern map, the outlines of the continents are seen to be extremely accurate. Africa is particularly well drawn and correctly placed in latitude and longitude. South America is also well placed, though like North America, is still given too great an extent in longitude.

The main area for which information is noticeably lacking is the northern Pacific, where Yezo (Hokkaido) is not yet clearly distinguished from the mainland, and ideas on the mythical ‘Company Land’ and the ‘Strait of Anian’ still plague the cartographer.

But if the continental outlines were now in great part known with considerable accuracy, the interiors of the continents outside Europe were still compounded of half truths, imagination and tradition.

In dealing with them Delisle made another departure, for he was prepared to admit, by ‘blanks on the map’, the limitations of contemporary knowledge.

In Africa, for example, he abandoned the system of central lakes which was an inheritance from the sixteenth century, and showed the main branch of the Nile as rising in Abyssinia, and elsewhere he displayed the same critical spirit.

Since much information, in Asia particularly, still reposed on the authority of Greek and Latin writers, he gave much time and thought to determining the equivalents of ancient measures of length.

As he lacked capital, and therefore the assistance of skilled engravers, Delisle’s maps are not outstanding in their execution, but they are free of the mythical monsters and other devices with which the older cartographers had disguised their ignorance—or attracted their customers. In this respect, again, Delisle marks the transition to the modern map.

His total output was not large, approximately 100 maps, in comparison with seventeenth-century map publishers, and much of his work was done to accompany works of travel or topography, for a map by Delisle was held to confer distinction on them.

He seems to have extended this simplicity of style to the representation of relief; he was certainly justified in objecting to some styles of mountain drawing, thought to enhance the attractiveness of a map, but on the main principle he was sound: “One of the main things demanded of a geographer is to mark clearly the rivers and mountains, because these are the natural bounds which never change, and which lead naturally to the discovery of geographical truths.”

D’Anville and the critical advance

The improvement in the map of the world initiated by Delisle was continued and greatly extended by J. B. Bourguignon d’Anville (1697-1782).

His talent lay in the critical assessment and correlation of older topographical sources, and their reconciliation with contemporary observations.

He was essentially a scholar, working mostly from written texts, which he collated with existing maps, and expressing his conclusions cartographically.

Throughout his life he never journeyed beyond the environs of Paris. His extensive collection of cartographic material (10—12,000 pieces) was famous.

Acquired by the French Government shortly before his death, it is now in the Bibliothèque Nationale, Paris.

So great was his skill and industry that he soon acquired an international reputation as a map maker in an age in which classical scholarship still dominated the world of learning. D’Anville was in fact the last, and perhaps the greatest of those, who since the Renaissance, had followed this procedure, and he probably carried it as far —as it was possible.

He was one of the first to study the works of Oriental writers for details on the countries of the East. Further approach to accuracy could be attained only from exploration and actual survey of the continental interiors.

The first mark of recognition was bestowed upon him by the Society of Jesus, when they entrusted him with the preparation for publication of the surveys of the provinces of China, upon which members of the Order had been at work since the later years of the seventeenth century.

In many instances these were based upon astronomical observations for position, but in others were simply route surveys.

From these maps, Western Europe obtained the first reasonably accurate and comprehensive conception of the geography of a large part of eastern Asia. With the aid of these sectional surveys D’Anville compiled a general map of the Empire of China.

The maps, forty-six in all on sixty-six sheets, accompanied the ‘Description géographique’ of the Chinese Empire compiled by J. B. du Halde from the Jesuit reports, and were later issued at Amsterdam with the title ‘Nouvel Atlas de la Chine’, 1737.

An English edition of Du Halde with versions of the maps appeared in 1738-41. D’Anville’s share in this Atlas was that of a compiler; but the efficiency of his general method of work was displayed by his map of Italy, 1743, based upon a critical study of Roman itineraries and measures of length.

The result was to reduce the area of the peninsula by “several thousands of square leagues’, and the accuracy of his deductions was strikingly confirmed by geodetic observations later carried out in the States of the Church by order of Pope Benedict XIV.

D’Anville’s notable maps were those of the continents, North America, 1746; South America, 1748; Africa, 1749; Asia, 1751; Europe, in three sheets, 1754-60; and a general map of the world in two hemispheres, 1761.

The outlines and positions of the continents, being based on the same data, differed little from those of Delisle; their merit is displayed in the treatment of the interiors.

On the map of Africa, for example, D’Anville went far beyond Delisle in removing the conventional and largely fictitious topography, and his representation stood until the great journeys of the nineteenth century inaugurated a new era in African cartography.

D’Anville took the correct view that the Blue Nile, rising in the Abyssinian highlands, was not the principal branch of the Nile.

Refusing to break completely with Ptolemy’s ideas, he depicted the main river issuing from two lakes in the Mountains of the Moon, in 5° N. latitude and approximately 27° 30’ E. longitude.

The northward bend of the Niger is conspicuous, but is carried over 3° too far to the north, and the river truncated in the west.

In the east it is connected with what may be intended for Lake Chad. In a note D’Anville states that there were reasons for presuming, contrary to common opinion, that the great river flowed from west to east.

Elsewhere, except in the north, the detail is almost entirely confined to the coastal areas.

Another celebrated work was his map of India published in two sheets in 1752, the best map of the sub-continent before the work of Major James Rennell and the Survey of India. D’Anville issued revised maps as the detail of contemporary exploration came to hand.

In 1761 they were published as an atlas, and amended re-issues appeared until the early years of the nineteenth century. He paid great attention to draughtsmanship and engraving—the lettering is clear and attractive—and in this respect his maps are greatly superior to those of Delisle, and to most of the products of his century.

But perhaps his greatest contribution to cartography was due to the degree to which he carried out his own precept: “Détruire de fausses opinions, sans même aller plus loin, est un des moyens qui servent au progrès de nos connaissances’.

Representation of relief: Buache, Gyger and hachures

D’Anville’s work was carried on by his son-in-law, Phillippe Buache, who had a part in developing a more satisfactory method of representing relief on topographical maps, a problem which was receiving much attention at this time.

On early engraved maps, hills and mountains, scarcely differentiated, were usually shown in profile, sometimes with shading to one side.

These symbols are often called ‘mole hills’ or ‘sugar loaves’.

The decisive step was the advance from depicting ranges of hills or mountains as separate and isolated features to the representation of the surface configuration as an integrated whole.

An interesting early example of this is the map of the upper Rhine valley in the Strasburg Ptolemy of 1513.

On this the escarpments of the valleys are shaded, and the tributary valleys are incised in the uplands, which, however, are shown with a uniformly level surface.

In countries such as Switzerland, the first attempts were more in the nature of oblique perspective drawings; as the science of surveying developed, efforts were made to represent the actual area occupied by a range.

This, combined with the profile, produced a three-dimensional effect.

One of the more successful methods of rendering relief was evolved by the Swiss cartographer, Hans Konrad Gyger (1599-1674).

In his maps of the Swiss cantons he attempted to show the land surface as though viewed from above, working in the folds and hollows by careful shading and leaving the higher areas untouched.

His skilful workmanship, combined with his extensive personal knowledge of the country, produced a remarkably plastic effect, though he could convey only relative, not absolute, differences in altitude.

That his method does not appear to have been generally followed was no doubt due to the lack of adequate data.

For the next century at least the representation of relief was generally confined to shading valley slopes at a more or less uniform distance from the rivers.

This style is employed, for instance, in the map of the environs of Paris made by members of the Academy of Sciences and engraved by La Pointe in 1678.

Even on the sheets of the Cassini survey, seventy years later, no essential advance had been made, and the effect is much inferior to that achieved by Gyger.

The method of hachuring, by which relief is indicated by lines (hachures) running down the direction of greatest slope, may have been a development of this practice.

The principle was fully worked out in the course of the eighteenth century to meet the requirements of military commanders. J. G. Lehman, on the analogy of the shadows thrown by an overhead light, propounded the theory that the greater the inclination of the surface to the horizon the heavier should be the hachuring, and he worked out a systematic scale for the thickness of the strokes.

Hachuring, however, has several defects; if carried out elaborately, the heavy shading obscures much of the other detail on the map, and by itself can give no absolute value for the difference in elevation between one point and another.

Moreover, without reference to other features, it is difficult to distinguish elevations from depressions.

Contours (contouring) and subsequent development

The solution of the problem now generally employed is the contour line, i.e. a line running through all points at a given elevation. Unlike the hachure, it runs along, and not down, the slope.

The origin of contouring is still somewhat obscure.

An obvious contour is the line of high or low water, and it is not surprising therefore that it appears to have been developed in the Netherlands, at first to show the configuration of the sea bottom.

Soundings off coasts and in estuaries are common on charts of the sixteenth century, on which banks are also enclosed by broken lines.

It would not be a great step in advance to run these lines through soundings indicating a given depth of water.

This appears to have been the practice by the beginning of the eighteenth century, by which time the number of soundings had greatly increased.

On a map of the Merwede estuary (1729), N. S. Criscnstal a Dutch engineer, showed depths by lines of equal soundings, referred to a common datum. Soon after, Philippe Buache drew a bathymetric map of the English Channel, with underwater contours at intervals of ten fathoms, but this map was not published in the Mémoires of the Paris Academy until 1752.

In 1737 he had submitted to the Academy a chart of Fernao da Noronha with submarine contours, accompanied by a vertical section across an off-lying bank.

Since he was also engaged in levelling operations in Paris, he must have recognized the applicability of the contour method to land surfaces.

Its first use on land however is usually credited to Milet de Mureau, who about 1749 used lines of equal altitude on his plans of fortifications.

The eighteenth century was a period of great activity in the construction of canals, and it is therefore quite probable that the engineers responsible for them discovered the principle independently, just as Charles Hutton did in 1777 when seeking a method to determine the mass of Schiehallion, a mountain in Scotland.

The general use in maps of large areas was delayed by the lack of sufficient data, though Cassini and others in France had calculated some heights by triangulation and by the barometer.

The earliest British map to include spot heights appears to be Christopher Packe’s ‘Physico-chorographical chart’ of Kent, 1743.

Packe obtained his altitudes by the comparison of barometric readings. Spot heights were frequently used before the end of the eighteenth century, e.g. on Mayer’s ‘Atlas de la Suisse’, 1796-1802.

One of the earliest examples of the use of contours for a considerable area was Dupain-Triel’s map, ‘La France considerée dans les differentes hauteurs de ses plaines’.

This purports to show France contoured at intervals of ten toises (about sixty feet), but the representation is largely influenced by his ideas on the orderly relations of mountains and plains.

No general levelling had been carried out at this date, so that his contours were largely theoretical, but a number of summit heights are given, some of considerable accuracy, especially that of Mt. Blanc, and he added a vertical section across France.

Dupain-Triel elaborated his methods and advocated their adoption in education in his ‘Méthodes nouvelles de nivellement’, 1802.

Thus by the beginning of the nineteenth century the method was becoming known, and with the initiation of the great national surveys in the following decades it passed into general use.

A further step, the colouring of areas between successive contours by a given scale of tints, was taken in Stieler’s ‘Handatlas’ of 1820.

This hypsometric layering allows a general idea of the relief of a wide area to be formed rapidly. The value of contouring lies in the fact that, unlike hachuring, it enables the altitude of a particular point to be determined with considerable accuracy, for heights between contours can be estimated with practice.

It does not, however, always permit an idea of the relief to be formed rapidly, and minor topographical features between contours are unrecorded.

Consequently it is frequently combined with hachuring or hill-shading. In 1931, contours, hachuring, and layer colouring were all employed by the Ordnance Survey in the Fifth (Physical relief) edition of the One Inch map.

This post is also on: