270 
BRITISH PHARMACEUTICAL CONFERENCE. 
seven equivalents of mercury) being evidently wrong, the filtered liquid was 
tested with a solution of cinchonine, and the marked reaction that followed 
unmasked the fallacy. I found, in fact, that a solution of alkaloid undergo¬ 
ing estimation by iodokydrargyrate, gives at the end of the reaction, an am¬ 
biguous response to reagents,—a decided precipitation being caused by either 
alkaloid or iodokydrargyrate. This state of things continues for a greater 
or less time, according to the nature of the alkaloid. Thus 100 grain mea¬ 
sures of quinine and 130 of iodokydrargyrate gave a filtrate reacting still 
with iodokydrargyrate, but not with quinine; but with 140 grain measures 
it reacted decidedly with both reagents. This was at 9 p.m. At 9.30 p.m. 
the same liquid reacted decidedly with quinine, but scarcely perceptibly 
with iodokydrargyrate. The same results were observed after twenty-four 
hours’ additional delay. Supposing 135 grain measures to be the correct 
number, the result tends to confirm my original opinion of the abnormal cha¬ 
racter of the quinine compound, suggestive of an incorrectness in its received 
formula. I may say, that I have recently repeated the analysis of this salt 
in the old way, and with the same results as previously. M. Valser’s figures 
were not abnormal beyond this, that they pointed to the old equivalent of 
quinine, —162, being the correct one. Strychnia possesses the same pecu¬ 
liarity as regards ambiguous reacting, but retains it only for a short time. 
Thus 100 grain measures, added to 100 grain measures of iodokydrargyrate, 
gave a filtrate that reacted strongly with both reagents. After the lapse 
of two hours, it had entirely lost its reaction with the alkaloid, and retained 
that with the iodokydrargyrate in a very small degree only. These pheno¬ 
mena were quite independent of acidity and neutrality. 
With morphia the results were more singular. 100 grain measures of this 
base were estimated in the usual way. After 70 grain measures of iodoky¬ 
drargyrate had been added, the filtrate proved neutral to both reagents, 
though hydrosulphate of ammonia indicated the presence of mercury in solo 
tion. No alteration was observed after the lapse of twenty-four hours. The 
70 were then increased to 100, still no precipitation by ether reagent; but 
when a solution of strychnia was added, an immediate and abundant precipi¬ 
tate was the result. A series of experiments showed the cause of this to be 
the ready solubility of iodokydrargyrate of morphia in the iodide of potassium, 
of which the reagent contains an excess. It is also soluble in hot water,— 
sufficiently so in either case to precipitate the more sparingly soluble strychnia 
compound. 
Nicotina again presents some peculiarities. 100 grain measures, represent¬ 
ing of an equivalent, were treated with 95 grain measures of iodokydrar¬ 
gyrate. The filtrate was clouded decidedly by a further addition, so also 
when the amount was increased to 100. At 200 grain measures (— 4 equiv. 
of mercury) the reaction was not stopped, but a distinct cloud obtained on 
adding the iodokydrargyrate guttatim. When added in larger quantity, the 
effect was to redissolve the precipitate. At all stages of the reaction, cin- 
chonia or strychnia, being less soluble, produced a precipitate in the filtrate. 
Conia does not behave so similarly to nicotina as would be anticipated 
from their near resemblance. The reaction stops promptly after the addi¬ 
tion of 100 grain measures, —2 eq. of mercury, the filtrate not being clouded 
by an excess of iodokydrargyrate, and only slightly by more alkaloid. 
I conclude, then, that Professor Mayer’s volumetric assays do not in¬ 
validate the formula Aik,HI,2 Hgl, independently arrived at by M. Yalser 
and myself, and that the simple volumetric estimation of the alkaloids can¬ 
not safely be attempted, except in the case of some few, such as strychnia 
and conia, whose reactions are in a great measure devoid of the ambiguities I 
have pointed out. 
