^ 63. ISOLATION OF ALKALOIDS, ETC. 55 



§ 63. Isolation of Allcaloids, etc., not Removable by Agitation. — As 

 stated in § 58, there are some alkaloids which, for the reasons 

 there given, cannot be isolated by the method of agitation. *. To 

 separate these in a state of purity the extracts prepared according 

 to § 57 should be exhausted as far as possible by shaking with 

 petroleum spirit, etc., and then evaporated to dryness. The dry 

 residue should be finely powdered (if necessary, with the addition 

 of washed sand or. siliceous earth) and treated wit3i alcohol, ether, 

 chloroform, etc. The exhaustion, however., with these solvents 

 will be incomplete if the residue be not reduced to a very fine 

 powder. (See also ^ 65, 66.) 



Detection by Group-reagents. — ^Before proceeding to this extrac- 

 tion it may appear desirable to ascertain whether an alkaloid is 

 present at all To that end the liquid obtained in § 57, containing 

 sulphuric acid but no alcohol, may be tested for alkaloids by 

 precipitants which have been introduced as group-reagents for 

 that class of substances. The following may be especially recom- 

 mended : 



Tri-iodide of patasdum — that is, an aqueous solution of iodine 

 in iodide of potassium — gives, with aqueous solutions of most 

 alkaloids, amorphous precipitates of a dark-brown or kermes- 

 raineral colour, and is one of the most delicate reagents. An 

 alcoholic solution of an alkaloid frequently' remains clear on the 

 addition of the tri-iodide ; or if a precipitate is formed, it differs 

 in projierties from that produced in aqueous solutions. For 

 example, berberine and narceine would under these conditions 

 yield crystalline precipitates. 



Tribromide of potassium, prepared in a similar manner, also pre- 

 cipitates some of the alkaloids from very dilute solutions, but, ic 

 addition, forms yellowish compounds with phenol, orcin, and 

 many allied bodies (§ 168). 



Potassio-mercuric iodide, obtained by decomposing mercuric 

 chloride with an excess of iodide of potassium, yields with 

 most alkaloids, white, floccnlent precipitates,' which sometimes 

 gradually assume a crystalline character. (See also § 65.) The 

 presence of free acid may sometimes cause a difference in the 

 precipitate obtained from one and the same alkaloid. 



Fotassio-Usmuthic iodide, prepared by dissolving iodide of bis- 

 muth and iodide of potassium in water, yields, even in highly 

 dilute solutions, precipitates which are very sparingly soluble, 



