The Proceedings of the Athanasius Kircher Society has posted an interesting geographical puzzle. An article, "The Story of Ink," in the 1930 issue of the American Journal of Pharmacy included the following statement:
Iron tannin inks are sometimes formed naturally; such a phenomenon has been observed in Algeria, a country in northern Africa, where there exists a "river of ink." Chemical examinations of the waters of the streams combining to form this river revealed that one of the streams is impregnated with iron from the soil through which it flows while the other stream carries tannin from a peat swamp. When the two streams joined, the chemical action between the tannic acid, the iron and the oxygen of the water caused the information of the black ferric tannate, making a natural river of ink. Does this river of ink actually exist? And if so, where is it on a map?
The earliest reference to this mysterious river I could find occurred in The Athens Messenger on May 25, 1876. The short blurb read:
"A river of ink has been discovered in Algeria. Let them find a mountain of paper, and then send for William Allen." For the next seven decades, similar passages -- almost verbatim to what ran in the Am. Jour. of Pharmacy -- appeared regularly in newspapers. They were typically thrown in as an odd bit of trivia to fill up column space. However, the name and location of the river itself (except for the fact that it was in Algeria) was never identified.
More recently, Bruce Felton and Mark Fowler included a passage about this river in their 1994 book The Best, Worst, & Most Unusual: Noteworthy Achievements, Events, Feats & Blunders of Every Conceivable Kind:
Most Unusual River: The comingling of two tributary streams in Algeria forms a river of ink: One brook contains iron; the other, which drains from a peat swamp, contains gallic acid. Swirled together, the chemicals unite to form a true black ink. (Black Brook in upstate New York is formed by a similar chemical blend.)Though the chemical composition of this "river of ink" sounds plausible, the other details about it are so vague that it sounds a bit like a geographical urban legend.