388 Prof. Curistison on the Poisonous Properties of Hemlock, 
have been given to a dog without marked effect of any kind, 
and even an ounce has acted only as a feeble poison ;* while it 
will presently be found that the same preparation is sometimes 
a poison of exceeding energy. 
These discrepancies are easily understood now, on consider- 
ing the extreme proneness of conia to decomposition, provided 
the conig of Grrcer be the true active principle of hemlock. 
From what has come under my observation, the extracts of hem- 
lock may become feeble, if not even inert, in one of two ways,— 
either by the heat being continued after the concentration has 
been carried to a certain extent, or by long keeping. On the one 
hand, I have always observed, that from the point at which the 
extract attains the consistence of thin syrup, ammonia begins to 
be given off in abundance, together with a modified odour of 
conia. And on the other hand, I have found extracts which 
were unquestionably well prepared at first, entirely destitute of 
conia in the course of a few years,—a remark which applies even 
to the superior extract prepared by M7 Barry of London by 
evaporation im vacuo. The mode of ascertaining the presence of 
conia is simply to triturate the extract or other preparation with 
solution of potassa, upon which an odour of conia is given off. 
I have no doubt that potassa is in this way a test of very great 
delicacy. 
Of the various extracts I have examined, that which has 
yielded the largest proportion of conia is one prepared by alcohol 
from the ripe seeds. Two hundred and twenty grains gave to- 
wards five grains of colourless hydrate of conia. From this pre- 
paration, indeed, it is probably to be obtained with more ease 
than from any other. I have prepared it quite colourless and 
free of all impurity but water, by exhausting ground hem- 
lock-seeds with cold rectified spirit in a percolator ; distilling off 
the spirit and concentrating in an open vessel over the vapour- 
* Orrita, Toxicologie générale, ii. 305. 
