Geology and discovery record of the Trinil Pithecanthropus erectus site, Java

The scientific utility of Eugène Dubois' Pithecanthropus erectus (P.e.) Skullcap (Trinil 1), Femur I (Trinil 3) and associated paleontological specimens has been impaired for over a century by questions about their provenience. Firsthand accounts and contemporaneous field photographs, presented here, extensively document the site geology and discovery history. The P.e. specimens and numerous-other fossils were unearthed in 1891- 1893 from small excavations dug into a flat-lying bonebed exposed near the seasonal low water level of the Solo River along its incised left embankment. Dubois' on-site supervisors specified that the two P.e. fossils came from a ~0.2-m-thick bonebed subunit traced at a single elevation for ~12m from the 1891 Skullcap pit (~30m 2 ) to the 1892 Femur-discovery excavation and across an enlarged 1892-1893 trench (~170m 2 ). The depositional co-occurrence of the finds is supported by key documentation: the supervisors' letters to Dubois about Femur I; his initial reporting to the Indies government; 1892-1893 accounts about expanding excavation of the Femur I stratum; Dubois' 1891 1893 government submissions and 1894-1896 publications; confirmation by the Selenka Expedition in 1907-1908; Dubois' annotations on unpublished site photographs; and a letter he wrote the year he died. Field studies in the 1930s to 1970s confirmed the essential aspects of the site geology. The bonebed of 1891-1893 contained fossils referable to the extinct Trinil fauna species Axis lydekkeri , Duboisia santeng and Stegodon trigonocephalus . The Selenka Expedition excavations had a similar assemblage in the same stratigraphic position which they named the Hauptknochenschicht. The bonebed was thin bioclast-rich gravelly volcaniclastic sandstone with taphonomic and sedimentary features indicating an unusual origin. Bioclasts range from proboscidean craniums and logs to rat teeth, freshwater mollusc shells and leaves. The terrestrial-vertebrate skeletal elements are overwhelmingly disarticulated and frequently broken. Their surfaces are little-abraded by fluvial transport. The bone fossilization is quite uniform. More than one-hundred ungulate individuals perished. No evidence has been found of hominin- or terrestrial-carnivore involvement. The bioclasts varied in density from place-to-place and vertically, and were matrix supported in the bonebed. No substantial internal depositional hiatus was reported. In combination with Trinil's paleogeographic context, these features implicate a catastrophic mortality of ungulates in a population aggregation along the floodplain of a perennial paleo-river, followed by lahar-flood transport and deposition of gravel-size lithic- and biotic-materials. Trinil provides evidence favoring a broad archaic-hominin presence in southern Sundaland. The Trinil fauna is a lynch-pin in a long-lasting paleobiogeographic association between H. erectus and certain lineages of large bovids, cervids, proboscideans, rhinoceros, suids and tiger. The bonebed's paleogeographic setting exemplifies the stratovolcanic drainages that H. erectus occupied for >0.8 million years in Java, including the watershed of a marine delta ~150km east of Trinil, a volcanic island ~100km north of Trinil, and areas to its west for 500km (where species associated with Trinil H. erectus occur). In the Java Sea (Sunda Shelf), seismic data image immense Pleistocene river- and coastal-terranes which archaic hominins and other large-mammals, like those at Trinil, might have inhabited. ### Competing Interest Statement The authors have declared no competing interest.