The Geological Survey of the Netherlands Indies unearthed 14 Homo erectus fossils in 1931-1933 from a single excavation site on Java (Excavation I Ngandong). Survey geologists attributed the hominin discoveries (along with thousands of other vertebrate remains) to a thin, gravelly volcaniclastic stratum situated near the base of a fluvialterrace remnant ~20m above the Solo River. The geologists were present 24 days during the 27-month-long operation, witnessing and documenting only a few human fossils in situ. Moreover, they published limited amounts of detail about the discoveries. Their provenience account is nonetheless substantially corroborated by surviving records. Key materials are presented here for the first time, including the geologists' photographs of two human fossils in situ and a 1:250 site map from 1934 showing individual Homo erectus discovery points.Calvarial specimens Ngandong I, II, V, VI, and VIII are the five best-documented finds. Each is securely attributable to the basal volcaniclastic fossil bed. In March 1932, geologist W.F.F. Oppenoorth photographed Ngandong V while it was still embedded in the fine volcaniclastic gravel <0.5m from the basal terrace contact. He previously had examined the contexts of Ngandong I and II when only <150m 2 of the excavation were open and all vertebrate fossils reportedly were being found in the gravely sandstone comprising the bottom ~0.7m of the sequence. In June 1932, geologists C. ter Haar and G.H.R. von Koenigswald photographed the Ngandong VI in a ~2m x 2m horizontal exposure of the basal volcaniclastic sandstone/fine conglomerate in which marl cobbles and 17 disarticulated non-hominin fossils also occurred. In August 1933, von Koenigswald removed part of Ngandong VIII from the basal bed near where six antler fragments and a Stegodon tusk were found. For the remaining nine discoveries, nothing is revealed in the available material to contradict the basal-bone-bed origin attributed to them by the discoverers. With some of the finds, such as Tibia B, there is virtually no substantiating documentation, while with others, such as Tibia A, it is unclear how the geologists knew about the stratigraphic context.The provenience detail that we present greatly improves the prospects for identifying the Homo erectus stratum at the site and collecting rock and fossil samples useful for radioisotopically dating the hominin, among other purposes. The documentation also substantially strengthens the inference that the hominin assemblage represents individuals who died at approximately the same time and whose remains were deposited at Ngandong within a few months of death. Support for this is provided by the limited amount of pre-burial weathering in evidence on the Homo erectus fossils, the highly delicate bony structures present in some calvarial specimens, and a combination of plastic deformation and sandstone-filled fractures occurring in several vaults-features which evidently represent warpage followed by breakage as the bone dried during burial.
INTROD...