ON THE ASSAY OF ALKALOIDS IN EXTRACTS. 
269 
of pharmacien of the first class, was entitled “ Study of the Search for, and 
Distinctive Characters and Estimation of, the Natural Organic Alkaloids.” 
Both of the experimenters settled on the same reagent, iodohydrargyrate 
of potassium, M. Valser first reviewing the other reagents previously sug¬ 
gested for the purpose—solution of iodine, chloride of gold, and phospliomo- 
lybdic acid,—all of which are passed over as inferior to the one chosen. In 
point of sensibility the iodohydrargyrate leaves nothing to be desired. It 
lias been proved that by it the presence of txwooo part of strychnia and -^ott 
of brucia can be readily detected. But in order to effect this, soluble iodides 
must not be present, for they exert a solvent action on the most intractable of 
the alkaline iodoliydrargyrates. 
It is curious to observe that M. Valser, throughout his paper, imagines 
that the compounds he treats of are new, thus repeating three years later the 
mistake I had been guilty of in a paper inserted in the 1859 volume of the 
‘Journal of the Chemical Society.’ However, he corroborates generally the 
formula I had assigned them ; whereas the experiments of Mr. Mayer seemed 
to point to different conclusions. To his results I will now direct attention, 
it being evidently of first importance to decide on the constitution of the 
compounds on which the estimation is practically based. His manner of 
proceeding was to determine “ the equivalent quantity of a tenth normal 
solution of iodohydrargyrate of potassium, by adding it gradually from a 
burette to a measured quantity of a solution of the alkaloid in 100 parts of 
dilute acid, and determining the end of the reaction by testing on a watch- 
glass.” The iodohydrargyrate was made by dissolving in water t L- equivalent 
of chloride of mercury and At equivalent of iodide of potassium, making up 
the volume to one litre. He thus found that the number of equivalents of 
mercury required for the completion of the reaction varied from 1 to 6, accord¬ 
ing to the alkaloid. As instances, I will cite aconitia, 1 eq.; atropia, 2 eq. ; 
morphia, 11 eq. ; nicotia, 4 eq.; quinine, 3 eq., etc. He remarked, also, that 
the mercury consumed in some of these reactions remained partly in solu¬ 
tion ; but I cannot see the reason why any mercury should necessarily re¬ 
main in solution, unless an excess of the iodohydrargyrate had been added, or 
the resulting alkaline compound were to some extent soluble. Whether we 
have compounds of one equivalent of hydriodate of alkaloid and two or more 
equivalents of iodide of mercury, or simply one and one, an equation can be 
framed to meet the case, leaving no uncombined mercury in solution. 
In testing these results with the viewof ascertaining the cause of error, I, after 
a few preliminary experiments, adopted test solutions containing respectively 
one equivalent of alkaloid in 100,000 grain measures, and one equivalent of 
mercury with three equivalents of iodide of potassium in 50,000 grain mea¬ 
sures. These extremely dilute solutions are rendered advisable by the bulki¬ 
ness of the precipitates, and the necessity for frequent testing by watch-glass ; 
besides which they represent more accurately the conditions under which the 
practical testing of the extracts would probably have to be conducted. It 
will be seen that the mercurial being of twice the strength of the alkaloidal 
solution, equal measures would, on the supposition of the correctness of the 
formula AlkIH,2HgI, be required to complete the reaction. 
Strychnia was the first alkaloid I thus estimated, 100 grain measures of 
which were apparently neutralized by 110 grain measures of the precipitant. 
The experiment was repeated with the same result, which was not affected 
by the presence of a greater or less amount of acid. 
Quinine was next tried, and the end of the reaction scarcely reached after 
the addition of 230 grain measures of iodohydrargyrate. 
Cinchonine gave still more abnormal results, the liquid affording a faint 
reaction after 380 grain measures had been added. This (equal to more than 
