§ 65. ESTIMATION OF ALKALOIDS. 59 



3. The alkaloid may be precipitated from its aqueous solution 

 by certain reagents, and estimated gravimetrically. 



Chloride of gold, or sometimes perchloride of platinum (§ 173), 

 may be advantageously used as precipitant in the last case, as the 

 amount of alkaloid and chlorine present may be approximately 

 calculated from the amount of gold or platinum contained in the 

 precipitate. Alkaloids may often be estimated gravimetrically 

 aud volumetrically by precipitation with potassio-mercuric iodide 

 (§ 174). 



I have discussed this subject fully in my ' Chemische Werth- 

 bestimmung starkwirfcender Droguen,'' where I have shown that 

 many alkaloids may thus be accurately estimated. I found, how- 

 ever, that the precipitates produced were not always analogous in 

 composition!, and that therefore the precipitating power or value 

 of the unit-quantity of reagent must be determined for each single 

 alkaloid. The composition of the precipitate yielded by one and 

 the same alkaloid may vary with the concentration of the solution, 

 and a difference in the amount of sulphuric acid present may 

 sometimes influence the result. A large excess of acid is incom- 

 patible with the accurate estimation of certain alkaloids, such as 

 brucine and coniine, whilst in other cases (nicotine, colchicine) it 

 is necessary. The latter alkaloid, together with atropine and 

 others, requires a considerable excess of the reagent for complete 

 precipitation, and in its gravimetric estimation therefore this 

 condition must obtain j on the other hand, the precipitate first 

 produced is sometimes redissolved on the addition of an excess 

 of the precipitant. With regard to the reagent itself, I may 

 observe that, accordingto Mayer, it is not advisable to prepare it 

 by dissolving iodide of mercury in iodide of potassium, the best 

 method being to mix 13 "546 gram of perchloride of mercury 

 ■with 49 "8 gram of iodide of potassium and water to make one 

 litre. 



For details of experiments I refer to the work already men- 

 ployed the method of agitation for the estimation of atropine. Compare 

 Phann. Zeitsohr. f. Ruasland, 1869, p. 89 (Year-book Pharm. 1872, 236). In 

 cotchicnm also it would be more advisable to determine the alkaloid by shaking 

 with chloroform than by precipitating with potassio-mercuric iodide. The. 

 material should be extracted with pare water, and the solution made acid 

 rather tlian alkaline before shaking with chlorofqrm. 



' St. Petersburg, 1874. Schmitzdorff. 



