9 
the phenolphthalein in such connection would be entirely super- 
fluous, since the alkaloid could be dissolved in an excess but definite 
amount of acid in the first place, and the actual excess of acid 
remaining uncombined determmed by means of the litmus indicator. 
Further, litmus as an indicator in alkaloidal titrations does not 
appear to yield highly satisfactory results. Thus, according to 
Kebler,® litmus is not as good an indicator in alkaloidal titrations 
as is cochineal, Brazil wood, or haematoxylin, the last giving the 
best results, while the shortcomings of even haematoxylin have already 
been pointed out above. 
It thus appears from the literature that although it is known 
that the salts of most of the vegetable alkaloids can be estimated 
by means of phenolphthalein as indicator, no application of this fact 
to the estimation of the purity of the uncombined alkaloids, as found 
in commerce or as obtained in the course of the various pharma- 
copoeia! assays, has been made where the phenolphthalein is used 
as the only alkalimetric indicator, and hence as the controlling 
factor of the accuracy of the titration. It therefore occurred to the 
writer that in the course of the various pharmacopoeial assays, if 
instead of dissolving the alkaloid in an excess of sulphuric acid, ^ as 
is usually done, an excess of a volatile acid be employed, such as 
acetic or hydrochloric acid for example, a definite amount of the 
acid, proportional to the amount of alkaloid present, would become 
fixed by combining with the alkaloid, and which combination of alka- 
loid and acid might be found stable at the temperature employed 
for volatilizing the excess of acid remaining free. If this were done, 
the amount of acid thus becoming fixed. could in most cases be 
determined with the aid of phenolphthalein as indicator, and hence 
indirectly also the amount of alkaloid. And if hydrochloric acid^ 
were chosen as the volatile acid to be thus employed, we could also 
estimate the alkaloids indirectly by determining the amount of 
chlorine in the alkaloidal residue by means of the well known and 
exact Yolhard method; while the advantage thus to be gained in 
rendering it possible to estimate alkaloids indirectly by means of 
the Yolhard method would perhaps be found of even wider applica- 
tion than the adaptation of phenolphthalein to their estimation 
alkalimetrically, since although phenolphthalein would be applicable 
as indicator in the titration of most of the alkaloids, yet in the cases 
of some of them, as, for example, morphine,^^ phenolphthalein does 
not show the end reaction very sharply. Again, it is evident that 
« Log. cit. 
&U. S. Pharmacopoeia (1905), pp. 28, 67, 107, 143, 146-147, 197, 200, 300, 340, 344. 
Hydrochloric acid is actually employed by the U. S. Pharmacopoeia (1905) in a 
similar procedure for the gravimetric determination of coniine. 
^ Plugge: loc. cit. 
