One kilogram (2.5 moles) of iodoform (Note 1) is placed in a 5-l. round-bottomed flask fitted with a rubber stopper holding a mechanical stirrer (Note 2), a reflux condenser, a separatory funnel, and a thermometer with bulb immersed in the liquid. To this is added 500 cc. of the sodium arsenite solution [prepared from 274 g. (1.38 moles) of c. p. arsenious oxide, 532 g. (13.3 moles) of sodium hydroxide, and 2600 cc. of water] (Note 3). Stirring is started, and heat is applied until the temperature of the thermometer reaches 60–65°, where it is held constant during the whole reaction (Note 4). If the temperature is allowed to go much higher, especially at the beginning of the experiment, the reaction becomes too vigorous, and at the same time a little methyl iodide is formed. The remainder of the sodium arsenite solution is added gradually during the course of one-half hour. The whole mixture is then kept one hour at 60–65° to make sure that the reaction is completed.

The reaction mixture is allowed to cool somewhat (to about 40°) and is filtered with suction in order to remove the mechanical impurities which have come from the technical arsenious oxide and sodium hydroxide. The filtrate consists of a clear water solution with a pale yellow oil underneath. It is better not to allow the reaction mixture to cool completely before filtering or else a precipitate of sodium arsenate will gradually separate and this will tend to make an emulsion of the methylene iodide and hinder filtration and separation. The methylene iodide is now separated from the filtrate, dried with a little calcium chloride, and then distilled under reduced pressure. Practically all of it (Note 5) comes over as a light straw-colored oil between 106° and 107°/70 mm. (Note 6), the bath being kept at 130–135°. The yield of crude product is 610–660 g. (90–97 per cent of the theoretical amount). On redistillation not more than 10–20 g. is lost. The crude product is quite satisfactory for most purposes.