The celebrated poison-tree of Java was said to blight the country around it. In the account reproduced by Thomas Stamford Raffles, however, Thomas Horsfield found the large anchar tree surrounded by vigorous shrubs, climbing plants and other forest trees. At the largest specimen he visited in Balambangan, dense vegetation made it difficult to approach the trunk. The obstacle was living forest, rather than a bare circle created by a tree’s deadly atmosphere.
This case concerns the historical upas-tree legend and the investigation that separated its central environmental claim from a real poisonous plant. The outcome is Closed (Explained) for the supposed vegetation-killing surroundings: Horsfield’s reported direct observations contradict that feature of the story. It does not follow that the tree’s sap was harmless, or that every tradition using the word upas describes the same plant, place or event.
A sensational account and a botanical response
In a long note to The History of Java, Raffles identifies Foersch’s account as an extravagant misrepresentation. He distinguishes the fabulous tree’s setting, effects on neighbouring country and alleged use in executions from the demonstrable existence of a tree whose sap could be made into a poison. The distinction is already present in the historical source; it is not a modern attempt to explain away an account that Horsfield otherwise accepted in full.
Raffles also credits poetry with helping the story circulate. He points to Erasmus Darwin’s treatment of the tree as a malevolent presence in The Botanic Garden. The passage is valuable evidence of how this author understood the legend’s appeal. It does not establish a complete publication history of every version, and this entry does not assign an exact first-publication date to Foersch’s story from information absent from the reviewed passage.
Raffles says that Horsfield prepared his account at Raffles’s request, and that it appeared in volume seven of the Batavian Transactions. The source used here is the account as reproduced in the 1830 second edition of The History of Java, volume I, note 33, pages 51–62. Raffles’s editorial introduction, Horsfield’s first-person field observations and earlier information cited by Horsfield are different layers of evidence, rather than three independent investigations of the same specimen.
What Horsfield reported seeing
Horsfield describes the anchar as a very large forest tree in eastern Java, with a tall trunk, broad buttress-like extensions near its base and a spreading crown. His botanical account discusses bark, leaves, male and female flowers and fruit. These are the observations of a plant treated as part of a forest community, rather than as a solitary supernatural object outside ordinary natural history.
The decisive passage concerns its immediate surroundings. Horsfield says that he did not find bare ground around the trees he observed. The largest tree in Balambangan had healthy vines and climbing shrubs ascending its trunk, while young trees had grown from seeds nearby. That description directly conflicts with a claim that proximity to the tree necessarily destroys neighbouring vegetation.
He also reports visiting the tree and obtaining sap with local assistance. Some prospective helpers feared a skin eruption, which he distinguishes from a landscape-wide lethal influence. Their reluctance is evidence of a reported local concern, not proof of the much larger European story. His own account of standing near the tree provides a further practical contrast with the imagined impossibility of approaching it.
The young trees growing around the parent specimen prompt Horsfield to recall Darwin’s poetic image. In the field account, the comparison highlights the gap between an arresting literary picture and a growing plant population. The important observation is the presence of vigorous life around the trunk; no dramatic new incident needs to be invented to make that contrast intelligible.
A real poison did not validate the barren-zone story
The same account discusses the milky material obtained from the tree and describes historical experiments on its effects. These sections are not evidence that merely walking nearby was fatal. They concern substances and particular kinds of exposure under experimental conditions. This article records the historical distinction without reproducing the source’s preparation instructions, quantities or experimental procedures.
Horsfield differentiates the anchar from another plant called chetik. He describes the latter as a winding shrub and says that he had not obtained its flowers, leaving its classification uncertain in his account. Combining the two plants would erase a distinction he considered important. Nor should a modern botanical name be attached to the chetik merely because both appeared in a discussion of poison.
His references to effects on people also require care. He cites the earlier writer Rumphius and information supplied by a Banyuwangi resident. Those passages do not become first-hand human experiments by Horsfield when quoted in his paper. They illustrate the mixture of personal observation, experimentation and reported information that must be kept separate when assessing an early natural-history investigation.
The tree in modern botanical reference
The accepted name Antiaris toxicaria is recorded by the Royal Botanic Gardens, Kew’s Plants of the World Online, which includes Java in its distribution. Its general-information record includes upas tree among the common names. These modern botanical references support the identification of a real plant behind the familiar name; they do not turn every historical use of upas into a precise species identification.
The illustration above is a later botanical plate of Antiaris toxicaria from the work of Karl Schumann and Ernst Gilg, dated to the 1890s in its source record. It is included to show the plant, with its public-domain credit. It was not made by Horsfield during the visit described in Raffles’s book and is not an image of the particular Balambangan specimen.
Outcome and geographical limits
The original investigation as reproduced by Raffles supports a narrow, concrete explanation. A real tree and historically investigated toxic material lay behind a fabulous account of a deadly setting. Healthy surrounding vegetation and the investigator’s visit contradict the claim of a necessarily barren, lethal zone around the anchar. The Closed (Explained) label applies to that claim, rather than asserting that every historical detail or anecdote in the chapter has been independently verified.
The text places the field observation in Balambangan in eastern Java and includes information from Banyuwangi. The map uses a rounded Banyuwangi regional reference, at −8.23, 114.36. It is not the position of Horsfield’s tree, a surviving specimen, a dangerous valley or a recommended collecting location. No such precise site is identified in the passage.
No single incident year is assigned to the legend or the undated field visit in this extract. The 1830 date belongs to the reviewed edition; the first edition of Raffles’s book appeared earlier. Keeping those dates separate avoids presenting the book’s publication as a newly witnessed environmental event.