In February of 1935, Peter Fleming and Ella Maillart set out to travel overland from Beijing ("Peking") to India. Fleming was a correspondent for the Times of London; Maillart, a Swiss citizen, wrote for several French-language papers. In the end, they travelled for more than six months, using train, truck, camel, horse and donkey. Afterwards, each wrote a book about their adventures: Fleming wrote News From Tartary, and Maillart wrote Oases Interdit, later translated into English as Forbidden Journey.

These are delightful books, and, perhaps not surprising for two travellers who shared a tent for six months, they recount the same events, often using the same phrasing. It is all the more entertaining since both Fleming and Maillart had reputations of being forbiddingly independent, and it was for both of them a notable compromise  to travel together. They were both in their late twenties.

Their challenge was that the Chinese government would not give them permission to travel into western China. Because Xinjinag province was partially in the hands of teh Tungan rebels, adn Communist forces held Szechuan, Fleming and Maillart were lucky to get passports to go as far west as the Koko Nor, a large high-altitude lake in Qinghai province. Their plan was that once they got to the Koko Nor, there would be ways of simply keep hiring camels and guides to go west. In this way they might be able to enter Xinjinag by a little-used route,  and the Tungan rebels would likely allow them to pass through to British India.

Mapping where these two went is a classic puzzle in Central Asian map-reconstruction. Both authors include the  names of many places they passed through, but in English (or French) transliteration. Places names they mention come from one of four languages: Chinese, Mongol, Turki/Uyghur and Tibetan—each has its own alphabet/character set, and there are multiple transliteration schemes. To complicate matters, on today's maps one finds the Mongol, Turki/Uyghur and Tibetan names have often been either transliterated into pinyin (e.g., Fleming's Teijinar is now Taiji Na'er) or replaced by a Chinese translation (e.g., Koko Nor, Mongolian for 'Blue Sea' is typically now shown as "Qinghai Hu" or Blue-Sea Lake.)  In some case the Chinese name appears unrelated to the former Turki name (e.g., Guma now being called Pishan).

As always when dealing with historical narratives by Europeans in China before 1949, there is an entire set of toponyms that these writers used which are not put on maps today. Sometimes the place name was the same, but rendered in the old Wade-Giles transliteration scheme, such as Zhangzhou being called "Chengchow." Other times the connection is unclear, such as  Wuhan being called "Hankow." As far as I know no one has compiled a historical gazeteer of China where one can look these names up. Old maps seem to be the best source.

In the course of mapping their journey, I would have liked to have had the maps they carried. They do write about  looking at the map, but unfortunately they have left us no clues about what maps they carried. Two maps that I found especially useful were these, from the David Rumsey site:

• Stanford’s 1908 Atlas of the Chinese Empire. The relevant maps are Tibet, Sinkiang and Kansu. and I imagine that Fleming and Maillart had maps something like this. Quite frequently the spelling Fleming uses from places match the spelling on these maps.

• Vivien St. Martin’s 1935 Northwest India and China. In French, but once you account for the French transliterations, many of the place names agree with Stanford 1908. This map was probably compiled the year before Fleming and Maillart were on the journey, so it should represent the names Europeans had for these places at that time.

Also useful were later topographic maps from three different series:

• International Map of the World, at 1:1,000,000 scale, produced in the 1950s and 60s, and available from the Perry-Castaneda Library at the University of Texas
• The Soviet topographic mapping at 1:500,000 scale,
• US Army Map Service 1:250,000 scale mapping of northern India and Pakistan, dating from the 1950s and 60s.