240 
COLOUR-TESTS FOR STRYCHNIA, ETC. 
drop of strong sulphuric acid. The acid solutions were then 
treated with equal small fragments of the colour-developing 
substances, six with the permanganate, six with the bichro¬ 
mate, and six with the ferricyanide. In one group of three 
all the reagents failed. This experiment was, therefore, ex¬ 
cluded. But the remaining five groups gave the following 
results: 
The ferricyanide of potassium gave faint indications of 
strychnia in two instances, and failed in three. 
The bichromate of potash acted characteristically once, 
gave faint indications twice, and failed twice. 
The permanganate of potash acted quite characteristically 
five times, producing a clear brilliant blue solution, changing 
to mulberry and to bright red. 
In these last experiments I estimate the quantity of 
strychnia in the stains at about the ij^th of a grain. 
The answer to the second question—of the colour-tests 
for strychnia, which is to be preferred?—is, therefore, evi¬ 
dently in favour of the permanganate of potash. The re¬ 
maining questions must be reserved for future consideration. 
I now proceed to discuss the third of the four questions 
propounded in my first communication, namely, Are the 
colour-tests, or is the selected test, open to any serious 
objection ? 
Of chemical tests in general, it may be said that they are 
more or less liable to the following objections or disparage¬ 
ments :—1. They may be wanting in delicacy; that is to say, 
they may not be able to detect minute quantities of the sub¬ 
stances to which they are applied as tests. 2. They may be 
wanting in certainty; in other words, they may fail from 
slight causes difficult to appreciate and define, such as the 
quantity or proportion in which they are applied, or the 
strength of their solutions if used in the form of liquid. 
3. They may give rise to transient effects, or to appearances 
difficult to describe or define. 4. They may be unusually 
liable to failure from impurities in themselves, or in the 
matters to which they are applied as tests. And, lastly, 
they may produce of themselves appearances nearly allied 
to those which they occasion when added to the matters under 
examination. 
From the first of these objections all the colour-tests for 
strychnia are certainly free. They are able to detect very 
minute quantities of the poison. From the fourth objection, 
too, as it will presently be shown, they are at least as free 
as the majority of chemical tests. To the second, third, 
and last objections they are certainly exposed, but under 
