152 THE SIMPLER NATURAL BASES 



moved by carbon dioxide; the filtrate is then treated with tannin, of 

 which the excess is removed by baryta. The choline and betaine are 

 then precipitated together as periodides from acid solution, and after 

 successive treatment of the precipitate with copper powder and with 

 cupric chloride, the mixture of chlorides is separated as described 

 above. If much betaine is present it is preferable to effect a prelimin- 

 ary separation of the dry chlorides by means of absolute alcohol, in 

 which betaine chloride is insoluble. 



Tests, Chemical Properties and Salts of Choline. 



An admirable account of choline is given by Gulewitsch [1908, i]. 

 The free base is very soluble in water, from which it cannot be ex- 

 tracted by organic solvents. (Only amyl alcohol extracts more than 

 traces from an alkaline solution.) Choline is a strong base, liberating 

 ammonia from its salts and preventing the coagulation of proteins. 



The most delicate precipitant is potassium tri-iodide (limit accord- 

 ing to Gulewitsch i : 20,000; according to Kinoshita [1910,2] the 

 limit (with Stanek's concentrated potassium tri-iodide, see above) is at 

 i : 2,000,000. The choline per-iodide on standing forms rhomboidal, 

 almost quadratic, leaflets. 



Phosphotungstic acid precipitates at I : 20,000 (Gulewitsch). Less 

 sensitive precipitants in aqueous solution are potassium bismuth iodide, 

 mercuric chloride, saturated cadmium chloride and gold chloride. 

 Tannin precipitates only in strictly neutral solution. In absolute 

 alcoholic solution mercuric chloride and platinic chloride are the most 

 delicate reagents (i : 2,000,000). 



The/*ftft&& test was used by Florence [1897] as a reaction for 

 semen; Bocarius [1901] showed that it is due to choline. The test 

 may be applied in a characteristic way to crystals of choline platinic 

 chloride. After evaporating the solution of this platinum salt in 

 1 5 per cent, alcohol on a microscope slide at 40, potassium tri-iodide 

 solution (20 grm. iodine and 60 grm. potassium iodide per litre) is 

 added ; the yellow crystals of the platinichloride disappear and are 

 replaced by dark brown doubly refractive and dichroitic prisms and 

 plates of choline periodide. When the excess of reagent evaporates, 

 the periodide dissociates and the brown crystals liquefy and disappear, 

 but they can be reformed by again adding the reagent (cf. Rosen- 

 heim [1905-6]). Joesten [1913] considers that Florence's crystals 

 are perhaps merely iodine, without any choline. His paper should be 

 consulted for an account of the literature of the reaction. 



Alloxan reaction. When a drop of choline hydrochloride is eva- 



