12 
ON THE SUBLIMATION OF THE ALKALOIDS. 
process was not carried further. Of these fourteen sublimates, eight showed 
crystalline forms under the microscope, and the rest were distinct and granular. 
Of this latter class I selected one, which certainly could not Lave weighed the 
fifteen-hundredth of a grain, and, having taken extreme precautions to ensure 
the absence of any trace of strychnine except in the crust, obtained from it two 
well-marked sublimates showing beautiful crystalline forms under the micro¬ 
scope, a third distinctly granular, and reacting most characteristically with 
one of the liquid reagents, and a fourth and fifth quite distinct, but neither 
crystalline nor granular. The glass from which the sublimates had been 
driven off retained a visible stain. If, then, we reject the last two crusts, we 
still have a sublimate consisting of grain, yielding three characteristic de¬ 
posits. So that a quantity as small as t^Vo g r; dn would yield abundant evi¬ 
dence of the presence of strychnine. I repeated this experiment with another 
smaller crust of the same series of fourteen, and obtained two well-marked crys¬ 
talline sublimates. So that I am able to state that a crust not weighing more 
than X sVo grain will yield other crusts, two or three in number, having the un- 
mistakeable characters of strychnine-sublimates, and one or other of which must 
weigh as little as from ^ to grain. 
Out of this conclusive experiment arose a question of great interest: Could 
such sublimates be obtained from small quantities of strychnine deposited from 
solutions of the alkaloid—from such solutions as would be formed in the course 
of medico-legal investigations? The fact stated above, that Dr. Helwig ascer¬ 
tained the quantities of the alkaloid submitted to sublimation by dilution and 
drop-measurement, goes far to answer this question ; but I sought a more com¬ 
plete answer by dealing with a small and thin deposit from a solution of strych¬ 
nine in benzole, which deposit did not present any definite crystals when examined 
by the microscope. Comparing it with the sublimates on which I had just been 
experimenting, I should estimate its weight as not exceeding grain. This 
spot gave four successive sublimates having the crystalline form of strychnine. 
In both these cases, whether operating with the sublimate or with the de¬ 
posit from solution in benzole, the procedure was perfectly simple, and the re¬ 
sult apparently certain. The alkaloid does not melt, but sublimes as a deposit 
of arsenious acid from a solution in water does. The flame of the spirit-lamp 
was continuously applied to the glass disk, a shallow glass cell with wide open¬ 
ing was superimposed so as to surround the spot, and the sublimate was re¬ 
ceived on a second glass disk, carefully cleaned and dried by being passed 
through the flame of the lamp. 
There are other preliminary questions not raised in Helwig’s work, but which 
are too important to be overlooked. Is sublimation a property of the alkaloids 
and the allied active principles as a class? is one of these. I sought for au 
answer to this question by preliminary experiments with thirty-seven sub¬ 
stances, in which were comprised-, all the active poisons and medicines ; and 
I found that, at the first experiment, no less than fifteen out of the thirty- 
seven gave distinct crystalline sublimates. So that it may be safely asserted 
that upwards of one-third of these substances respond to the test of sub¬ 
limation. The remainder, after melting, like the rest (cantharidine excepted, 
which sublimes without melting) gave off vapour which was deposited on the 
glass disk as watery patterns, generally mixed with crystalloids. Now, as even 
those alkaloids which ordinarily yield crystalline sublimates do exceptionally 
furnish these watery patterns, it is not improbable that the list of substances 
giving sublimates of more defined form would be extended by repeating the ex¬ 
periments with them, and learning by practice the temperature which suits 
them best. Indeed, I have already found that by multiplying experiments with 
the members of the opium group, 1 have transferred three of them (papaverine, 
narceine, and paramorphine) from my list of alkaloids not yielding crystalline 
