The High German four-lined dance of death / Oberdeutsche vierzeilige Totentanz is not just known from manuscripts, but also from a number of murals.
According to an old chronicle (Germania canonico-Augustiniana) this mural was created in 1440 under abbot Udalricus Strobelius (Ulrich Strobel) called Mayer in 1440: "depingi curavit anno 1440 choream defunctorum".
The mural was whitewashed over, but was rediscovered in 1923. A homeland scholar and Catholic priest named Rudolf Weser (1869-1942) described the persons in the newly found dance as "rather big, but very spoiled and hardly recognizable anymore" (»ziemlich großen, aber sehr verdorbenen und kaum mehr erkennbaren Bildern«). The mural was not a fresco but was painted on the wall with tempera, which was a contributing factor to its deterioration.
The uncovered piece was 10.8 meters — from the emperor's reply to the Physician's Death. All in all 27 verses. Above the king's dialogue was a V, which could be interpreted as the Roman numeral for 5. If the dance has started with a preacher like most other versions of the The High German four-lined dance of death, this would make the king number five. Weser therefore estimated that 2.5-3 meters were missing at the beginning and 9 meters of the end, meaning that the mural originally had measured ca. 22 meters.
Concerning the text below the painting, Weser writes a sentence that can easily be misunderstood: "I have now succeeded in reading a part of individual verses and with help of what was read and known for sure, to discover the whole content of the inscriptions" (»Es ist mir nun gelungen, einen Teil von einzelnen Versen zu lesen und mit Hilfe des Gelesenen und sicher Erkannten den ganzen Inhalt der Inschriften zu entdecken«).
Weser had found similar texts in Füssen and in Kleinbasel, and he thought that the text matched the "Ur-text", which Massmann had constructed in 1847(1) (it would appear that Weser hadn't read Fehse and didn't know the term Oberdeutsche vierzeilige Totentanz). Weser then quoted 3 verses "as a sample": »Als Probe der Ulmer Totentanzdichtung«, and this is where one can easily be fooled: These 3 × 4 lines do not originate from the church-wall in Ulm (which was badly deteriorated), but from Massmann's "Ur-text".
Weser wrote another article the same year in Ulmische Blätter für Heimatliche Geschichte, Kunst und Denkmalpflege, pages 81-83, where he quoted several verses from Massmann's ur-text and marked by boldface what he said could be read on the wall.(2)
This uncovered area perished during a fire 17/12 1944, but in return the same fire made another area appear: From noblewoman to child, although three figures had been destroyed by construction works while it was hidden.
The church was demolished in 1952/53. Before that, Hellmut Rosenfeld had studied the remains and made the drawing to the right. Wilhelm Munz has also copied the same area to a "trace" — i.e. a copy of the mural on transparent paper — along with a few letters and words. The picture to the left shows the noblewoman, in other words the same scene as the drawing to the right. The difference between the two renderings shows how difficult it has been to recognize details.
To sum up: The painting is the oldest datable from 1440, but it has only been known, while it was in a bad shape, and while it was barely possible to recognize the text as Oberdeutsche vierzeilige Totentanz.
This dance is painted on an octagonal former ossuary, but in contrast to dances in Northern Europe it was painted on the outside of the building (picture to the left), and has thus been exposed to weather and wind. When the danse was described for the first time — by Friedrich Lippmann in 1875 — it was already so damaged that Lippmann only could determine that the text was in German. Six of the participants on the west-side had then totally disappeared with only the bare wall remaining.
There is a copy in watercolour from 1885 of the surviving area. In 1891 the dance was described by Paul Gruëber along with linedrawings. The drawing of the beggar shows a graffito: "1546", which gives us a terminus ante quem for the painting. Gruëber discovered that single letters, rarely entire words, show that the dance is closely related to Heidelberg's block book.
The dance in Heidelberg has been expanded from the normal 24 dancers by inserting an apothecary between the merchant and the nun. There is also an extra person in Metnitz, but he was inserted between the canon and the physician, and it's hard to determine who he is, since the text has disappeared. In the modern reconstruction he is called "academic".
In 1968/1970 the remains were moved to a museum (picture to the right). A reconstruction was created on the outside of the chapel in 1989, based on the watercolour from 1885, drawings from 1889 and Heidelberg's block book.
The mural was uncovered in 1879 and the dance was described in 1880 by Rahn who recognized the text as Massmann's "Ur-text" (the term Oberdeutsche vierzeilige Totentanz wasn't invented before 1908).
In contrast to the two previous paintings is was possible to read entire words and understand sentences, and Rahn was able to quote from mother, child, peasant, lawyer and knight.
Seven years after the mural had been discovered, in 1886, the building was demolished. What was unknown (or at least forgotten) was that during the seven years while the painting had been uncovered, the Austrian sculptor and restaurateur Joseph Regl (1846 - 1911) had copied the paintings from the chapel on to 12 traces, i.e copies in full size on transparent paper.
These rolls were discovered in January 2009. Six of the rolls are from the dance of death, and four of these include the text. This means that scenes 1-6 no longer exist while scenes 7-11 are without text.
Most scholars agree that the dance in the Dominican nunnery Klingental was painted around 1440 and partly renovated in 1512. One scholar, Koller, instead argues that 1512 was the year the mural was created.
The painting was expanded with no less than 15 new dancers for a total of 39 dancing couples. Oddly enough, 14 of these new figures are inserted in one block at the point where the mural continued from the western wall of the cloister to the northern wall.
The dance was reproduced in watercolours by baker, illustrator, topographer and painter Emanuel Büchel in 1768. Thus we know the whole text, with the proviso that by 1768 the painting had deteriorated after more than 300 years, and at several places was broken through by new windows (and with the proviso that Swiss authorities have a century-old tradition of banking secrets, and that they store Büchel's watercolours out of the way, so that it's almost impossible to get copies).
As with the painting in Kleinbasel the row was expanded to 39, and the two paintings are probably more or less of the same age, i.e. from ca. 1440.
The mural was (and still is) very famous, and text and images have been published by Frölich and Merian among others. We have a transcription of the entire text by Ludwig Iselin from 1577, and colour gouaches from around 1600. We still have 19 fragments of the wall, although these fragments do not contain any text.
However, the painting in the big city had been subject to much more frequent renovations than in the isolated nunnery, Klingental, so already in our oldest sources there are fundamental changes like:
Footnotes: (1) (2)
Massmann . . .: Hans Ferdinand Massmann, Die Baseler todtentänze in getreuen abbildungen from 1847
At the end of his book, Massmann tried to reconstruct a "Ur-text" by combining variants from the different sources. See the page about Cgm 270 (and footnote 2 on the same page) for details about Massmann's Ur-text.
This information is from Mischa von Perger. I haven't read the article.