 Solution

Keep an untouched portion, use clean apparatus and reagent blanks, separate the suspected arsenic from the organic material, and confirm it by independent chemical reactions. A coloured deposit alone would be inadequate evidence.

Record the container, seal and condition of the stomach contents before opening. Examine and describe the material, including any visible particles, without assuming that a white powder is arsenic trioxide. Retain an unaltered portion for a repeat examination. Because a solid poison may settle, examine representative liquid and solid portions separately rather than testing only the clear upper liquid. Label each portion and each subsequent extract so that a positive reaction can be traced back to its source.

A practicable historical preliminary separation is the Reinsch test. Warm a portion with hydrochloric acid and a clean strip of copper; a dark coating suggests deposition of a substance requiring further identification. Carry out a parallel blank with the same copper, chemical acid, water and apparatus. A coating may be caused by substances other than arsenic, and contaminated copper or reagents can create a false accusation. Wash and dry a coated strip, then heat it in a small tube: an arsenical deposit can give a sublimate of arsenic trioxide, whose crystals can be examined with a microscope. The deposit and sublimate, rather than their mere colour, are the material for confirmation.^[ref]

On a separate suitably prepared portion, the Marsh test provides another mode of separation. The original James Marsh apparatus collected a hydrogen mixture for deposition tests; a heated-tube implementation separates collection from the subsequent heating zone. Zinc and dilute chemical acid generate hydrogen; arsenic in a suitable chemical form is converted to volatile arsine, and heating the gas in a narrow tube can produce a metallic mirror. First run the generator with the actual reagents alone, until their freedom from arsenic has been demonstrated. Keep the sample out of the gas outlet, and distinguish chemical deposition from mechanical carry-over. The apparatus separates a volatile derivative from much of the organic matrix, which explains its analytical value. Arsine is extremely toxic: this is a description of historical analytical principles, not a procedure for an uncontained demonstration.^[ref]

A mirror is not uniquely arsenical: antimony can form stibine and give a misleading deposit. Recover or oxidize the material and use additional reactions, such as formation of yellow arsenic trisulfide under appropriate acidic conditions, together with the characteristic oxide sublimate. Separate aliquots, suitable controls and agreement between chemically distinct tests are more persuasive than repeatedly observing the same ambiguous stain. Organic material must be removed or treated by a controlled preparation that neither introduces nor loses arsenic; indiscriminate charring or heating can invalidate the examination.

Finally distinguish detection from interpretation. A demonstrated arsenical compound supports exposure, but the amount, distribution and circumstances determine its significance. A negative test on the stomach contents alone does not exclude poisoning after absorption or expulsion of the poison. Further organs or excretions could be examined if available, but no positive finding in such material is assumed here.