372 
ON THE MELTING AND SUBLIMING TEMPERATURES 
shown in the annexed woodcut. It consists of a plate of copper, with a hollow 
nipple of the same metal worked into the centre of 
it, without the use of solder. This nipple receives 
the narrow bulb of a thermometer indicating all 
temperatures up to 700° F. The thermometer is 
supported by a perforated cork fitting a ring of a 
retort-stand. The apparatus is used as follows :— 
A minute quantity of the poison or other substance to 
be examined (say the grain) is placed on a small 
fragment of microscopic glass, and this on a part of 
the copper plate, between the centre and circum¬ 
ference. It is then surrounded with a glass ring, 
on which a clean disk, previously dried in the 
flame of the spirit-lamp, is placed. The flame of 
the spirit-lamp is then steadily applied to some 
point of the copper plate equidistant from the ob¬ 
ject and the thermometer. The instrument must be 
placed in a good side-light, and the eye of the ob¬ 
server should be at such a level as to see with equal 
ease the object itself, the surface of the glass disk, 
and the scale of the thermometer. To get good re¬ 
sults with this simple instrument, the flame of the 
lamp should be so applied that the temperature may 
rise steadily and rather slowly, so that each obser¬ 
vation made on the substance under examination or 
on the superimposed disk, may coincide very nearly 
with the temperature indicated. But as the heat 
imparted to the mercury in the bulb of the thermo¬ 
meter does not immediately travel to the metal in 
the upper part of the tube, it is necessary to with¬ 
draw the lamp directly the change which we are 
looking for occurs, either in the substance under 
examination or on the disk of glass. The thermo¬ 
meter will then be oj^erved to rise a few decrees, 
and the highest point to which it attains should be 
set down as the temperature (approximatively) at 
which the change in question was perceived. 
From this description it will be seen at once that this method is not suscep¬ 
tible of extreme accuracy. When we speak of any particular temperature, as 
that at which a certain change (as melting or subliming) takes place, we mean 
the temperature at which, using care and caution, we first perceived a change 
of form in the substance, or a mist upon the disk. We may miss the precise 
moment of the change, and make too high an estimate of the temperature. 
Such errors of observation have been guarded against, in the case of the results 
presently to be detailed, by repeating the experiment several times with the 
same substance. Proceeding thus, I have rarely found occasion to correct my 
first entries of results; and I believe that the method will be found to yield 
approximations sufficiently close for every practical purpose, if we only adopt 
the simple and obvious precaution of not attaching importance to differences of 
a few degrees between the melting or subliming points of two poisonous sub¬ 
stances. If, for instance, we experiment with a substance which melts at 220°, 
and assume it to be codeine, it would be a valid objection that paramorphia and 
papaverine melt at 210°. So also with aconitine and atropine. A temperature 
of 140° would not justify us in assuming that we are dealing with aconitine, 
seeing that atropine melts at 150°. But if we suppose that a minute quantity 
