[Arsenic drugs in the four medical translations in the late Qing Dynasty].

This paper attempts to recreate and reconstruct the process of 'translation-transformation-application' of arsenical knowledge in the late Qing Dynasty by examining the four representative medical translations: An Outline of Western Medicine (Xi Yi Lue Lun), Explanations of Western Medicine (Xi Yi Lue Shi), Universal Prescriptions (Wan Guo Yao Fang), and A Compendium of Western Drugs (Xi Yao Da Cheng). It was found that in terms of nomenclature, the early coexistence of heterogeneous vernacular names such as Xin () and Xin Shi () underwent evolution and adjustment within medical texts, eventually coexisting with chemical nomenclature such as Shen (、). This resulted in a hybrid terminology system that integrated traditional experiences with modern chemical elements. In terms of dosage forms and quantity, arsenicals expanded from a single medicinal liquor into five major categories - solution, ointments, solids, pastes, and pills. This evolution presented a trajectory from an early phase of 'extensive collection' to a later phase of 'simplified consolidation' based on chemical classification. In terms of chemistry and toxicology, the translated works helped to establish a quantifiable and verifiable framework of modern pharmacology for substances long regarded as 'poisons' initially by successively introduced experimental detection techniques (such as the Marsh test), valence-based classifications, and semi-sinicized chemical formulas. In terms of authoritative citation, translators attempted to establish a chemical basis for arsenicals, facilitating their transformation from empirical poisons into 'scientific drugs' by drawing extensively on experimental data from European and American chemists and physicians. This study argues that the transfer of Western arsenical knowledge in the late Qing Dynasty was a process of continuous communication between translators and the medical community by introducing the knowledge and local practice rather than a simple linear process of scientization.