nwhyte

Expand Cut Tags

No cut tags

We went to the Ulster Museum yesterday, where there is a very good and full exhibition about the Troubles. One of the exhibits is a map from the Roads Service from 1979 showing exactly which bits of Belfast were closed off for security reasons.

In the yellow zone, no unattended vehicles were allowed from 8 am to 6 pm Monday to Saturday; in the red zone, no unattended vehicles were allowed at any time; and within the barriers, on the cross-hatched streets, only “delivery vehicles etc” were allowed. There were only seven points where vehicles could enter the city centre, of the twenty-odd possibilities; the others were simply blocked off.

What I found interesting is the similarity with the wall built around Belfast in the 1640s, as mapped by Gillespie and Royle.

The gates marked on Castle Street and North Street are in exactly the same place as two of the security barriers of 330 years later. It would have been perfect if they actually coincided with vehicle access points, but they don’t quite - the North Street gate is next to (but on on) the site of the later access point at the northern end of Royal Avenue, and to reach Castle Street in the 1970s you would need to go in at the southern end of Queen Street. There was another gate in the 1640s on what was Corporation Street in the 1970s but is now the Dunbar Link.

I tried drawing the two on one map, but the 1970s security zone looked a little too phallic.

It is interesting that the commercial centre of Belfast had shifted so little in a third of a millennium. The core areas of economic activity in, say, London and Dublin would have moved a lot between the 1640s and 1970s. But I guess that is less true of some other cities I know well, like Brussels or Leuven. (Or Oxford or Cambridge, for that matter.)

There is no trace above ground now of the 1640s fortifications, and little enough from the 1970s. Long may it remain so.

This is an update from last week's post, where I considered a slightly different version of the first part of this question. I have illustrated this with some nice maps, all taken from Google, but with the latitude and longitude grid developed by Bill Chadwick of the Bracknell District Caving Club (a really neat idea which nobody else seems to have implemented).

Which line of longitude passes through the most countries?

The first part of the question is, which line of longitude, a straight line from North Pole to South Pole, passes through the most countries?

Well, my answer last week was wrong. In fact there is quite a wide band, between the westernmost point of Bulgaria at 22°31'35.2"E, and the easternmost point of Slovakia at 22°33'32.1", which passes through no less than 26 countries. This is 22 km wide at the equator, but obviously narrows as you get closer to the poles. It intersects:

1) Norway
2) Finland
3) Sweden

(Baltic Sea)

4) Estonia (the islands of Hiiumaa and Saaremaa)
5) Latvia
6) Lithuania
7) Russia (the Kaliningrad exclave)
8) Poland
9) Slovakia (the easternmost edge)
10) Ukraine (just)
11) Hungary (just)
12) Romania
13) Serbia (just)
14) Bulgaria (the westernmost edge)
15) Macedonia (just)
16) Greece

(Mediterranean Sea)

17) Libya
18) Chad
19) Sudan (just)
20) Central African Republic
21) Democratic Republic of Congo
22) Angola
23) Zambia (just)
24) Namibia (the Caprivi Strip)
25) Botswana
26) South Africa

(The line is drawn wider than it really is on the map to the left.)

The biggest settlement this passes through is Lublin, in Poland, whose population is 350,000; I'd say from a glace at the map that between a half and a third of them live on the line. Here it is passing through Europe (Norway and Sweden cut off at the top) - drawn to scale on this map:

Three of the westernmost spurs of Bulgaria are south of thirteen countries and also north of thirteen countries, Serbia counting both times.

Here's the line passing through central Africa, to scale again, showing how it intersects both Sudan and Zambia. A corner of Libya is visible at the northern edge. The two countries on the southern edge are Namibia and Botswana, with South Africa off the map.

This is a lucky confluence of frontiers at different latitudes. To the east, within less than half a degree you lose Slovakia, Serbia, Russia/Kaliningrad, Hungary and Macedonia before you gain Belarus and South Sudan, and then you lose Poland, Chad, Angola and Sweden. To the west, you lose Bulgaria, Ukraine, Zambia and Sudan before you gain Kosovo, and then you lose Estonia, Latvia, Lithuania and Macedonia.

I can't imagine that any meridian that does not pass through Europa and Africa can come close to this total. In the Americas, Asia and Oceania, the countries tend to sprawl east-west as much as north-south, and they are generally bigger anyway.

Which line of latitude passes through the most countries?

So, that's longitude sorted. What about latitude?

I found this much more tricky, but I believe I have identified two parallels of longitude which arguably pass through 27 countries. (Arguably, because one of them passes through what could be considered a 28th country, and the other passes through two territories which belong to different constituent parts of the same kingdom, so possibly should be counted together rather than separately). As far as I can tell, these are the band between 11°08'02.4"N and 11°08'25.8"N, defined by the northern frontier of Togo and the southern shore of Tobago (so I will call it the Togo-Tobago line); and the slightly wider northern band between 12°11'03.7"N and 12°11'34.4"N defined by the shores of Bikar Island in the Pacific Ocean (so I will call it the Bikar line).

As I list them below, I put the ordering of each country in the form (a/b) where it's the a'th country on the Togo-Tobago line and the b'th country on the Bikar line. If it's a country on one line but not the other, I put an asterisk for the line it isn't on.

These are actually really tight strips of the earth's surface. I reckon the Togo-Tobago line is about 700 metres from north to south, and the Bikar line just under a kilometre. Going east from the Atlantic, they hit the African coast at Guinea-Bissau (1/1) (the Bikar line just misses Senegal) and then head inland to Guinea (2/2), Mali (3/3) and Burkina Faso (4/4). The Togo/Tobago line passes through the southern suburbs of Burkina Faso's second city, Bobo-Dioulasso, but misses Côte d'Ivoire.

Now we get to a weird bit of colonial map-making. The border between Ghana and Burkina Faso plays hide and seek with the 11th parallel before Ghana (5/*) lurches north in the region of Kulungugu. The border of neighbouring Togo (6/*) also has a peculiar northern kink, just enough to define the northern edge of the Togo-Tobago line. The two lines are shown to scale below.

Farther north, the Bikar line hits Niger (*/5) before both lines pass into Benin (7/6), Nigeria (8/7), the northern spike of Cameroon (9/8), and Chad (10/9). The Togo-Tobago line just misses the Central African Republic, and both lines then pass through Sudan (11/10), South Sudan (12/11, a close shave for the Bikar line), Ethiopia (13/12) and Djibouti (14/13, a close shave for the Togo-Tobago line).

The Togo-Tobago line passes from Djibouti into the northern part of Somalia (15/*), hitting both Somaliland and Puntland (we'll get back to that later). The Bikar line just misses Eritrea to the north and Somalia to the south, but does intersect the island of Abd al Kuri and its main settlement Kilmia, part of the Socotra archipelago which belongs to Yemen (*/14). The lines on the map here are to scale.

Both lines now cross the Indian Ocean, intersecting India (16/15), Myanmar/Burma (17/16), Thailand (18/17), Cambodia (19/18), Vietnam (20/19) and the Philippines (21/20). I have not investigated the various claims on the Spratly Islands, because as far as I can tell none of them is intersected by either line. The Bikar line passes over (or at least very near) the holy hill of Arunachala in Tamil Nadu.

That takes us to the Pacific Ocean, where I have gone over the geography of the Marshall Islands (22/21) in some detail. This next map, uniquely, is not from Google but from Wikipedia, showing the various Marshallese atolls and archipelagoes. The lines are not to scale - the distances here are pretty huge.

As far as I can tell, the Togo-Tobago line just scrapes part of Ailiningae Atoll - it grazes the southern edge of Sifo Island, completely covers Manchinikon Island, and passes through the middle of Charaien Island. It also grazes Taka Atoll, but passes between the islands of Waatwerik and Lojrong without making landfall. I'm posting the satellite view of Ailiningae Atoll as well as the map view below; doesn't it look beautiful?

The tiny island of Charaien, at the eastern end of the atoll, is less than a kilometre from north to south.

A little further north and east, the Bikar line is defined by the island of Bikar, part of the much smaller Bikar Atoll, very little of which is above water level.

After a long landless passage, over 11,000 km, across the Pacific, we reach the Americas. Both lines make landfall in southern Nicaragua (23/22), and the Togo-Tobago line then dips in and out of Costa Rica (24/*). The lines cross the southern Caribbean to Colombia (25/23) and Venezuela (26/24).

The Bikar line finishes its run of countries by passing through Curaçao (*/25), Bonaire (*/26) and Grenada (*/27). The two lines are shown to scale in all three maps below.

The Togo-Tobago line, true to its name, just scrapes the southern edge of Tobago, the northern and smaller island of Trinidad and Tobago (27/*).

I think it's this map, rather than the Togo or Pacific island maps, that brings home just how narrow the Togo-Tobago line is. You can almost see the trucks parked in the plantations.

Now for the caveats. First, I don't think it's clear that you can count Curaçao and Bonaire separately - while Curaçao is a separate constituent part of the Kingdom of the Netherlands, and Bonaire is a Dutch municipality, Curaçao notably does not have foreign policy or defence powers separate from the Netherlands (and therefore Bonaire); these are surely important attributes if we are distinguishing between states. So the Bikar line ends up with a fairly impressive 26 countries for sure, the 27th being a matter for experts on Dutch constitutional law.

Second, my personal view is that Somaliland should be counted separately from the rest of Somalia. Empirically, Somaliland behaves more like and is treated more like a state than is Curaçao, so if we count the latter we should count the former as well. It may not have been formally recognised as an independent state by any other country (neither of course has Curaçao), but it does have foreign policy and defence capabilities; I have myself attended meetings of Somaliland's diplomatic representatives and ministers with European officials, and am proud to say that I had a small hand in negotiating the 2013 Special Arrangement for Somaliland, so it is clearly engaged in foreign relations; and the co-operation of Somaliland's defence forces with international anti-piracy efforts in the Gulf of Aden is not much discussed but quietly recognised by the international community. Given that the Togo-Tobago line passes through both Somaliland, at its frontier with Djibouti, and Somalia/Puntland, just south of Cape Guardafui, which are undisputedly part of Somalia (Puntland is not seeking independence), I personally will count it as the winner with 28 countries.

(Incidentally, I hadn't realised that the Doumeira Islands, off the coast of Djibouti and Eritrea, are among the rare parts of the world without any formally recognised sovereignty, though in fact Djibouti is in administrative control. They are just too far north to make a difference to my count, however.)

A final thought - it’s interesting to compare this with the Datagraver map of world population by latitude and longitude. Their biggest population spike by longitude is a little to the east of mine, around 31° East (Cairo, Kiev, St Petersburg). Their biggest population spike by latitude is around 21° North, which looks like several Mexican cities, plus Nagpur and Hanoi.

Trivia question to finish with: how many countries lie entirely south of the Tropic of Capricorn?

I’m just starting Nisi Shawl’s Everfair, which made the Nebula finalists this year (but missed the Hugos, finishing 8th on the Best Novel ballot). The frontispiece is this map, which I found a little confusing:

The book is an alternate history where the Fabians create an idealistic colony in the Congo, and the map shows the different stages of its territorial expansion.

The problem is that the map legend has four colours, but the map itself seems to have only three.

My best guess is that the south-eastern corner is the 1893 first buy, the 1897 second buy is missing (I don’t see anywhere on the map coloured that dark), the north-east is the bit conquered by 1904, and the western spur the land offered in 1914.

Maybe it will become clear as I read the book.

Second paragraph of third section:

The accompanying map (fig. 1) shows the downtown area’s original streams and shorelines, to the extent that I have been able to locate them. The original boundaries and features of figure 1 are based on the maps of A. Boschke (1856-58), James Kearney and others (1838), A. C. Harmon (1931), James R. Dermott (1797), Joseph M. Toner (E. F. M. Faehtz and F. W. Pratt, 1874), Arthur B. Cutter (1952), Campbell Graham and S. T. Albert (1849), William T. Partridge (1895?), And. Ellicott (1792; Upper Anacostia), the map accompanying the Commissioners Proceedings of 1793, on several references cited in this paper (especially Hines, 1866; Proctor, various dates; Taggart, 1908) and on anonymous articles in the Evening Star. The canal routes are based on Boschke’s 1856-59 map. Present-day features and shorelines were taken from the U.S. Geological Survey’s 1:24,000 scale topographic map “Washington, D.C., and Vicinity” (1965).

My investigation of the pre-urban hydrogeography of the capital of the United States brought me eventually to this paper published by the U.S. Geological Survey in 1977 and written by Garnett P. Williams, better known for his 1997 book Chaos Theory Tamed. It’s a brief, fact-filled yet entertaining survey of the historical evidence of D.C.’s waterways, starting with the springs and working up through the streams, creeks and canals to the Anacostia and Potomac rivers. The Franklin Square springs get half a page, and there is also a description of the stream that they fed, called Goose Creek. “At F Street, near Ninth Street [immediately north of the future site of Ford’s Theater], the ravine carrying the stream was some 14 feet deep. (This was quite hazardous at night, with no street light. On one occasion, Mr. Philip Fendall, a leading member of the bar, fell in and broke his leg in two places.)” However, it was entirely filled in during the mid-1800s. Such is the fate of watercourses which compete with property developers.

Garnett regrets that so many of the old streams have been filled in (the fate of Goose Creek) or covered over, mostly to become sewers, and points to Rock Creek as an example of how a different path could have been chosen. (My comment: of course, Rock Creek is mostly elevated above the flatter topography of the city centre, so it was bound to be more robust in the face of humanity.) He points out how central the vanished waterways were to the city until the later part of the nineteenth century - the White House and Washington Monument were both up against the river shore (the future sites of the Lincoln and Jefferson Memorials under water), and the old canal (obsolete almost as soon as it was built) cut through the heart of the city.

And he concludes that the rivers in particular filled in and narrowed because of massive deposits of sediment, “made available to the rivers over the last 200 years by man’s carelessness”, and calls for more sensitivity to the natural waters of the landscape in city planning. He has a point.

I'm in D.C. for a few days, and have been musing on the geography of the city - how the familiar grid of streets attempts to drown out the underlying landscape, and sometimes succeeds.

Googling around historical maps, I discovered Don Lockwood’s fascinating 1990 reconstruction of the topography of the Federal City as it would have been in 1791, before the building started. Here’s the central stretch:

Isn’t it interesting that the old shoreline path (the Ferry Road) was not all that far from the later route of Pennsylvania avenue?

And my attention was caught by the clustering of contour lines, with a spring emerging from them, to the northeast of the future site of the White House. The spring - two springs, in fact - fed a creek that flows southeast and then south, before joining the Tiber Creek estuary. The heads of the spring seemed roughly to be in the location of today’s Franklin Square. Was there any historical record of it, I wondered?

And actually, yes there is. This map from Thomas, Cowperthwait & Co's 1850 Atlas gives the creek a slightly different but recognisable course (and let’s bear in mind that Lockwood possibly used this map, or a version of it, as the basis for his imagined 1791 topography):

There are two springs marked in Franklin Square, with streams joining in its southeast corner; the creek is shown flowing south along 13th Street for a block, then heading east between G and H Streets, then directly south between 9th and 10th Streets, past the Patent Office (now the National Portrait Gallery) before flowing into the canal opposite the Smithsonian. A rather odd street, G Place, now runs east-west between 9th and 10th Streets, bisecting the block where the map shows the creek turning the corner.

The creek would have flowed through the backyard of Ford’s Theatre (which is labelled 33 on the map, then the First Baptist Church), where John Wilkes Booth had his getaway horse waiting after his fatal attack on President Abraham Lincoln. The last two blocks of its course are now occupied by FBI headquarters and the Department of Justice. (That first part of Louisiana Avenue, part of the infamous Murder Bay in the old days, has been obliterated by the DoJ, and the rest of that end of it is now Indiana Avenue.)

It's also marked, with much the same route, on Faehtz and Pratt's 1874 reconstruction of the pre-development landholdings of the site of the city, rising on the land of John Davidson and flowing through the land of David Burnes.

Here's my attempt to draw it onto the map of downlown DC today (click to embiggen).

There’s no trace of it in later nineteenth century maps, and I think the depiction of the creek may already have been out of date by 1850. According to this site, water from the springs in Franklin Square was piped to the White House along 14th St and F Street from 1816 until 1832, which presumably means the creek ran dry. According to this site, more underground pipes ran from the Franklin Square springs after 1832. In the 1880s the hotels around Franklin Square were built and the springs dried up entirely.

But I think you can detect its legacy in some of the building plots and shapes that survive to the present day. I find the gap to the west of the Martin Luther King Memorial Library particularly suggestive.

We're used to the mildly hidden histories of great (and also small) European cities. But the newer cities of America have their own hidden depths as well.

I posted last week about my efforts to discover whether or not the house where I lived in Germany in the summer of 1986 was, as my landlord's daughter claimed, situated astride the old frontier between the Grand Duchy of Baden and the Kingdom of Württemberg.

A bit more research using the desktop rather than relying on the iPad apps has definitively resolved the issue.

First of all, I was delighted to find the 1902 Meßtischblatt for Schwaigern, which clearly shows the boundaries of the Schluchtern enclave.

Secondly, I don't generally use OpenStreetMap but in this case I found the detail visible at different scales much more attractive than the Google equivalent.

Zooming in a bit, I was able to draw latitude and longitude lines intersecting at the front door of my old house.

Drawing these onto the 1902 map was trickly but not impossible. There is a particular kink in the river Lein just directly south of my old address; and it's exactly east of the northernmost point of the Eppingerstraße, which presumably hasn't changed its course much since 1902. So drawing those lines on the 1902 map gets me this:

That does indeed seem to show the old boundary very close to the front door of the house, and probably at least intersecting the back garden if not the building itself. So the landlord's daughter wasn't just making it up. Whew!

Incidentally the Jewish graveyard ("Iſrael. Friedh.") marked on the older map is still there, just off the Kiesbergstraße. Just two of the community survived the Holocaust.

I lived in Germany for five months in 1986, working on an archaeology site near the city of Heilbronn. Along with several of the other (paid) volunteers, I lived in a rented house in nearby Leingarten, a dormitory suburban place (not so different from where I live now) up the valley of the river Lein (or Leinbach) which joins the Neckar just north of Heilbronn.

My landlord's daughter, a student in her mid-twenties (so a few years older than me - I turned 19 while I was living there), lived downstairs, and spurred my teenage imagination by having very loud sex with her boyfriend in the room immediately beneath mine. I have completely forgotten her name, but she and her flatmate and their respective boyfriends would occasionally invite us down to the garden for a neighbourly glass of wine.

One day we got talking about local history. It turned out that Leingarten had originally been two historically distinct municipalities, Schluchtern to the west and Großgartach to the east, which had been merged in the name of administrative efficiency back in 1970. In a local microcosm of the merger between the Grand Duchy of Baden and the Kingdom of Württemberg, Schluchtern had been in Baden in the old days, and Großgartach in Württemberg.

I asked her where the boundary ran? Since 1970 there had been a lot of new build (and looking at Google Maps there seems to have been more since 1986). While the former centres of the two former towns were still fairly clear - clusters of shops around the church and the municipal buildings - it was less obvious where one stopped and the other started.

She smiled and told me that the old boundary actually ran through the house. Precisely where, she wasn't sure, but the house was on the line that had separated Schluchtern from Großgartach until 1970, and Baden from Württemberg until 1945. As a map geek since my childhood, I found this very interesting. Unfortunately the local library had mysteriously run out of historical maps, so I was never able to check it, and the question of whether I really had been living on the boundary line lingered unresolved with me for the next thirty years.

Now, thanks to the internet, you can actually pull up historical maps (I'm using the aptly named Old Maps app on the iPad, but there are many other options) and find more answers. First of all, it turns out that Schluchtern was actually an enclave, a village which was in Baden though surrounded by Württemberg. (German Wikipedia has a very long list of such cases in south-eastern Germany alone.) Here is a detail from Müller's 1812 map of Baden, just a few years after the changes wrought to the German principalities by Napoleon:

It doesn't even show Großgartach, and give the impression that the Schluchtern enclave was basically anywhere in earshot of the church's bells. The spendidly named Geognostic Travel Map of Heidelberg and Vicinity published by Groos in 1830 does give a bit more detail - and, critically, shows Großgartach - but doesn't take us a lot further.

However, an 1843 military map (unfortunately with poor definition) shows a very different boundary. Here the centre of the enclave is distinctly west of the centre of habitation; the line as it runs between the two villages goes more or less north-south, somewhat closer to the centre of Schluchtern than of Großgartach. The enclave itself is not a neat circle but an elongated shape including a couple of hills to the north and a couple of valleys to the south. (Perhaps reinforcing my church bells theory.)

Finally an undated motorists' map (I would say first half of the twentieth century) by Freytag and Berndt gives the clearest picture yet. The shape of the enclave is recognisably similar shape to that shown on the 1843 military map. But the inhabited part of Schluchtern is crammed against the eastern edge, while the western border of the enclave grazes the next town along, Schwaigern ("Schweigern" in earlier maps). Most crucially, the easternmost point of the enclave appears to be very close to the location of the place I was renting in 1986.

So, if we zoom in on Schluchtern/Leingarten as it is today, with the location of my address marked, basically my landlord's daughter's story does look plausible. You have to ignore Bundesstraße 293, the road shown in yellow passing north of the town, which was built only in the 1970s; the road shown in the earlier maps is the one shown below as Eppinger Straße, the former highway from Heilbronn to Karlsruhe in simpler times. Bearing that in mind, and the relative location of the road going north from Großgartach to Kirchhausen, I reckon that the easternmost kink of the enclave's boundary may well have been just about where I was living.