The First Unorthodox, Modernist Mapping of the World?
/Before Bernard Cahill and Buckminster Fuller and their fragmented and angular world maps, there was Richard A. Proctor and his Star Atlas (1870), New Star Atlas (1874), and Student’s Atlas (1889)
update 4 November 2024: some typos fixed and phrases reworded; image of the 1866 maps added.
In working on a section of an essay on the difference of some world maps in the twentieth century, which is due soon(!), I decided to review the chapter on equal-area and compromise projections for world maps used in the USA and UK between 1850 and 1950 in F. V. Botley’s 1952 MA thesis. After quickly establishing the basic pattern of world maps in atlases, with general relief and some other environmental information on double-hemisphere maps on Arrowsmith’s “globular” projection and with political-economic and statistical-climatic data mapped on Mercator’s projection (p. 81), Botley turned to the increasing dissatisfaction after 1880 with Mercator’s projection for failing to give the reader a clear sense of geographical relationships. Botley specifically cited (p. 82), as an early innovation in this process, an atlas that was quite unknown to me:
Proctor, Richard Anthony. 1889. The Student’s Atlas in Twelve Circular Maps (on a Uniform Projection and One Scale), with Two Index Maps, Intended as a Vade-Mecum for the Student of History, Travel, Geography, Geology, and Political Economy, with a Letterpress Introduction, Illustrated by Several Cuts. London and New York: Longmans, Green, and Co.
Worldcat records only a few cataloged examples. Fortunately, the copy in Memorial Library at the University of Wisconsin–Madison (shelved down in the basement within the old Cutter classification books and no-one has bothered to recatalog it according to the Library of Congress system) has been scanned by Google and is also available at the Internet Archive.
Proctor’s small atlas is a very interesting work: he prefigured the later work by Bernard Cahill and Buckminster Fuller of mapping the spherical earth as a faceted polyhedron, specifically as a dodecahedron, while his use of the azimuthal equidistant projection might also have prefigured the world maps of Richard Edes Harrison; his complaints about the inconsistencies of contemporary atlases echo comments by Élisée Reclus and, much later, those of Arno Peters. I’m not saying that Proctor directly influenced these twentieth-century map makers, but his work does point to the dissatisfaction with the established conventions of world mapping and the desire for new and thoroughly modernist solution to the problem of imaging the three-dimensional globe in a two-dimensional map.
A further intriguing element about Proctor’s atlas was his decision to frame the oceanic basins in their own maps, being “almost as important and interesting a region of our earth as any continent” but for which “the ordinary atlas gives no map at all” (Proctor 1889, 9):
Map 2, showing the northern Atlantic Ocean, from Proctor, A Student’s Atlas
The Student’s Atlas was published posthumously, as its creator had died of yellow fever in New York in September 1888. Born in 1837, Richard Anthony Proctor was a British astronomer and writer of popular science books, many of which seem to have been self-financed. He is perhaps best remembered for his soon-superseded mapping of Mars in the later 1860s (Lane 2011, 26–28, 40–41, 44). He was sufficiently popular as a writer and public lecturer to be depicted in Vanity Fair in March 1833 (as imaged in the Wikipedia entry on him). After his second marriage, in 1881, he moved to the USA. Proctor signed the undated introduction to The Student’s Atlas as “Corona Lodge, Orange Lake, Fla.” (Orange Lake is just south of Gainesville; I have to wonder if Proctor was involved in some land development scheme.) The name of “Corona Lodge” seems to refer to astronomical work, including an 1871 essay on “theoretical considerations” about the Corona in the monthly notices of the Royal Astronomical Society.
— — — —
Before mapping the globe, Proctor had been mapping the heavens in sets of twelve maps. The first attempt, with steel-engraved maps, was a set of maps of twelve segments of the night sky on the azimuthal gnomonic projection: two pentagons for the celestial poles and then five quadrilaterals for the remainder of each hemisphere (Proctor 1866). The following image is from Barry Lawrence Ruderman Antique Maps and shows the fundamental structure that Proctor would follow for his later celestial and terrestrial atlases:
(Note the shapes of each of the ten non-polar star charts, and how they relate to Proctor’s schematic explanation of the arrangement of maps, below.)
Proctor then undertook a set of twelve circular maps of the heavens, defined as pentagons in the above manner but each drawn as circles on an azimuthal equidistant projection. He actually issued these maps in parallel works, each of which would go through multiple editions. (It took me much longer to figure out there similarities and differences than I had planned to spend on this blog post, although the examination of the star atlases turned out to be very useful in understanding the terrestrial atlas!)
First, in his A Star Atlas, Proctor (1870) reused the celestial hemispheres on the gnomonic projection as the index maps to twelve circular maps on the azimuthal equidistant projection. With the fourth, 1877 edition, Proctor replaced the two gnomonic index maps with two new maps on the azimuthal stereographic projection (that appear structurally identical to the index maps to the Student’s Atlas). According to library catalog records that I’ve seen online, the Star Atlas was about 24cm tall. He later referred to this book as his “Library Star Atlas” (Proctor 1889, 10).
Second, Proctor created A New Star Atlas (Proctor 1872), which was slightly smaller, being 21cm tall. This further work was intended to accompany a general text on observational techniques for astronomers (Webb 1859) and possessed a substantial explanatory introduction before offering twelve slightly smaller and redrawn circular maps of the heavens, prefaced by two index maps on the stereographic projection. The New Star Atlas—or his “Student Star Atlas” (Proctor 1889, 10)—went on to be reissued in many editions and revisions (as Proctor 1908).
The introductions to the two celestial atlases encompass the same arguments and spatial structure that Proctor would present in the introduction to his posthumous Student’s Atlas. In his Star Atlas, he explained that an atlas should comprise maps all “of equal size” and all “alike in shape,” that they should not cover too large an area else their distortions would be too great, that the distortion on each map should be uniformly centered on the map’s center point. To do this, he proposed to map the sphere of the celestial vault onto one of the regular solids: in particular, he proposed the use of the dodecahedron with its twelve pentagon-shaped faces as the most efficient in covering the heavens with the least accrual of distortion; moreover, the azimuthal equidistant projection offered the least appreciable distortions over such relatively small areas. The final step was to actually map each pentagonal facet by its enclosing circle, which therefore overlapped with the neighboring maps (Proctor 1870, 7–10, quotations from 7). (I’m skipping over the rest of his discussion of how to orient the facets, then how to construct each map, and finally how a photolithographic process was used to reduce his large drafts to the printing surface.)
In the New Star Atlas, Proctor referred the reader to the slightly larger atlas for a full explanation of the design of the atlas. He was more interested in selling the smaller work, beginning by noting that its reduction in size made “a little book which can be carried in the pocket” (Proctor 1872, vi). It was therefore far more useful, as well as more accurate, than the star maps just published by the Society for the Diffusion of Useful Knowledge (see Cain 1994). In the main work, however, Proctor returned to the matter of how his maps were better than others and how they were designed, now introducing some graphics to explain the process—these are interesting because he reused the images almost without any change for the Student’s Atlas, to which we (finally) turn…
— — — —
The title to the Student’s Atlas (above) lays out its similarity to the New Star Atlas: it was intended as a vade mecum for students, a book that they might easily carry around as they researched geography, politics, and history. Whereas the celestial atlases long remained in print, along with many of Proctor’s other astronomical books, the Student’s Atlas is known in just the one edition.
Frustrated by how geographical atlases commonly show different regions of the world on different projections and at different scales, so that it is impossible “of readily forming clear and exact ideas of the relations of different parts of the earth’s surface to each other” (Proctor 1889, v). The full text of the first paragraph of introduction is worth quoting at length:
THERE are some purposes which ordinary atlases fail to fulfil, and could not well fulfil even if they were free from the faults which characterise all the atlases in use. Some of these faults are serious; and when we consider the length of time during which geographers have had the problem of atlas-making before them, it seems singular that they have not been long since corrected. The ill-chosen methods of projection which were naturally enough employed when geography was in its infancy remain in vogue still. An even worse fault is, that even in the use of unsatisfactory methods there is no congruity. Some maps of large tracts, as Africa and South America, are on old Flamsteed’s projection, others, as the maps of Europe and North America (generally), are on the conical projection, others, as the maps of Asia and North America (occasionally), on a curious mixture of Flamsteed’s and the conical projections, which is perhaps the worst ever devised for mapping. The addition of a scale of miles to most of these maps is misleading—the distances measured in different directions being in many cases on quite different scales. Maps of smaller tracts are drawn usually on the conical projection, and this is well enough, though not the best that might be; but they differ too much in scale to give correct ideas. The maps of hemispheres could not perhaps be on a much better projection than that usually employed—a modification of the equidistant projection—but such maps are quite insufficient to bring the teachings of the atlas as a whole into unison. And when, as usual, a map on Mercator’s projection is added, the ideas suggested in regard to the position and relative extent of the different countries and tracts of the earth become for the most part erroneous, while they are often very confusing even when approaching in some degree to correctness. (Proctor 1889, 7–8)
After continuing in this vein, Proctor then explains the construction of the geographical maps, using the same images that he had used in his New School Atlas. First, the mapping of the earth to a dodecahedron:
The Earth as a dodecahedron, from page 10 of Proctor, A Student’s Atlas
Then the division of the northern and southern hemispheres into twelve pentagons, each one replicating the lay out of his gnomonic maps of the heavens from 1866:
Arrangement of the 12 pentagons, from page 11 of Proctor, A Student’s Atlas
And finally a rather abstract diagram of how the twelve pentagons might be arranged into a single spread
How the 12 pentagons assemble into a globe, from page 12 of Proctor, A StuDent’s Atlas
Proctor followed the last diagram with the following suggestion:
The student will find it a useful exercise to construct a dodecahedron by cutting a piece of card into the form shown in fig. 5 (cutting half through along the dotted lines), and folding it up into a dodecahedron such as is pictured in figs. 1and 2. (Proctor 1889, 12)
Significantly, Proctor did not make the further step that would characterize the similar map work by Bernard Cahill (1909) and Buckminster Fuller (1943). For those later map makers, the goal was to make a world map that showed the interconnections of the earth’s parts with minimal geometrical distortion. But Proctor seems not to have made that leap. For him, the atlas as a whole was the global map. To show the world as a whole still required two index maps. Drawn on the stereographic projection, which is conformal (shape-preserving) so the individual maps appear properly as circles, the index maps to the two hemispheres show how the more detailed maps fit together and overlap:
Index Map for the Northern Hemisphere, from Proctor, A Student’s Atlas
The index map reveals the markedly unorthodox way in which Proctor divided up the earth’s surface. His goal was to show the connectivity of regions. So, rather than focusing on particular regions of great importance to him—i.e., Great Britain—he placed those regions on the edges of each detailed map, blistering the edge of the maps as necessary, so that their interconnections might be readily traced. Britain, for example, sat on the edge of the maps of the Arctic (map 1), the northern Atlantic (map 2, above), and western Eurasia (map 10):
Map 10, showing western Eurasia, from Proctor, A Student’s Atlas
Overall, Proctor’s little atlas was amazingly unconventional for its time. The entire atlas was a world map, showing the entire globe in equal-sized chunks, all at the same scale. It did not focus on any particular state or empire in any way, it seemed to lack political purpose. Even when it depicted Great Britain in triplicate, it placed Britain on the edges of the maps, traditionally a place of lesser importance in regional mapping. It offered a new intellectual arrangement of global space, one that emphasized interconnectivity. Proctor did not go all the way to create a single map to show global interconnectivity; he seems unable to step beyond the conventional practices of mapping the world as a whole. And he seems unwilling to use any shape other than the circle, already widely used in terrestrial mapping, not even to map the earth in the same, angular way that he had mapped the celestial hemispheres in 1866 (above). Ultimately, Proctor could not break from the idea of the atlas as a work of reference, a function that dragged him back to conventionality.
An amazing work!
References
Cahill, Bernard J. S. 1909. “An Account of a New Land Map of the World.” Scottish Geographical Magazine 25, no. 9: 449–69.
Cain, Mead T. 1994. “The Maps of the Society for the Diffusion of Useful Knowledge: A Publishing History.” Imago Mundi 46: 151–67.
Fuller, R. Buckminster. 1943. “Life Presents R. Buckminster Fuller’s Dymaxion World.” Life 14, no. 9: 41–55.
Lane, K. Maria D. 2011. Geographies of Mars: Seeing and Knowing the Red Planet. Chicago: University of Chicago Press.
Proctor, Richard Anthony. 1866. The Stars, in Twelve Maps on the Gnomonic Projection, Collected, in Duplicate, in Four Plates. London: Longmans, Green & Co. Images available from Barry Lawrence Ruderman Antique Maps.
———. 1870. A Star Atlas for the Library, the School, and the Observatory, Showing All the Stars Visible to the Naked Eye, and Fifteen Hundred Objects of Interest, in Twelve Circular Maps on the Equidistant Projection; and Picturing the Heavens for the First Time on a Systematic Plan, without Appreciable Distortion, on an Adequate Scale, and within a Convenient Volume. With Two Index Plates, in which the Six Northern and the Six Southern Maps Are Exhibited in Their Proper Relative Positions, All the Stars to the Fifth Magnitude Being Shown, with Coloured Constellation Figures. London: Longmans, Green, & Co. Scan at Hathi Trust.
———. 1872. A New Star Atlas for the Library, the School, and the Observatory, in Twelve Circular Maps (with Two Index Plates), Intended as a Companion to ‘Webb’s Celestial Objects for Common Telescopes,’ with a Letterpress Introduction on the Study of the Stars, Illustrated by Several Woodcuts. London: Longmans, Green and Co. Scan at Internet Archive. 3rd edition 1873 (scan).
———. 1874. A Star Atlas for the Library, the School, and the Observatory, Showing 6,000 Stars and 1,500 Objects of Interest, in Twelve Circular Maps on the Equidistant Projection; with Two Coloured Index Plates, in Their Proper Relative Positions, Including All the Stars to the Fifth Magnitude, and the Constellation Figures. 3rd ed. London: Longmans, Green and Co. Scan at Internet Archive.
———. 1877. A Star Atlas for the Library, the School, and the Observatory, Showing 6000 Stars and 1500 Double Stars, Nebulæ, &c., in Twelve Maps on the Equidistant Projection; with Index Maps on the Stereographic Projection. 4th ed. London: Longmans, Green and Co. Svan at Google Books.
———. 1889. The Student’s Atlas in Twelve Circular Maps (on a Uniform Projection and One Scale), with Two Index Maps, Intended as a Vade-Mecum for the Student of History, Travel, Geography, Geology, and Political Economy, with a Letterpress Introduction, Illustrated by Several Cuts. London and New York: Longmans, Green, and Co.
———. 1908. A New Star Atlas for the Library, the School, and the Observatory, in Twelve Circular Maps (with Two Index Plates), Intended as a Companion to ‘Webb’s Celestial Objects for Common Telescopes,’ with a Letterpress Introduction on the Study of the Stars, Illustrated by Several Woodcuts. Revised and corrected by T. E. Espin (1895); 22nd impression. London: Longmans, Green and Co. Scan at Internet Archive.
Webb, T. W. 1859. Celestial Objects for Common Telescopes. London. 2nd ed., 1868, 3rd ed. 1873.
