This experiment was several times repeated, but by al lowing the alcohol to remain a little longer exposed to the cold, it became so completely solid, that, on inverting the tube, not the least portion of fluid could be perceiv ed to separate from the mass.
In order to be as certain as possible of the strength of the alcohol I employed, I again took its specific gravity, and the result corresponded with what I before obtained.
These experiments, therefore, left me no room to doubt that I had frozen alcohol, which, at the temperature of 62°, is of the specific gravity 798.
Being appointed (by the Edinburgh Institute) to de liver the course of lectures on chemistry for the session I had no leisure at that time to pursue these experiments. They were resumed, however, in the au tumn of 1811. The second experiment was repeated and varied, and solid masses of alcohol of some magnitude ob tained. Some of these I soldered together, using as a hot bolt a rod of frozen mercury, and sometimes a straw cooled down to a very low temperature.
It now appeared to Inc to be an object of some impor tance, to ascertain the form of the crystals which this substance assumes. This I found attended with some difficulties which I did not anticipate, and attempts to overcome them have led to the discovery of some facts which I did not at all expect.
The common masses exhibited crystals of different forms ; two kinds appeal ed to predominate, and each was tolerably distinct in its kind, but it was not %cry easy to perceive by what increments or decrements the one could be supposed to pass into the other ; a rather casual cir cumstance, however, explained the source of this variety. Attempting to freeze alcohol by a modification of the ge neral process, which I conjectured would yield more re gular crystals than the common method, I observed, that before crystallizing, the alcohol separated into three very distinct strata ; the uppermost was of a pale yellow ish green, while the second was of a very pale yellow colour. Both these strata were very thin, the last men tioned was rather the thickest ; the lowermost stratum was nearly transparent and colourless, and very greatly exceeded the other two in quantity. After allowing a part of the lower stratum, which I conceived to be the pure alcohol, to freeze, 1 attempted to pour out the re mainder, but was prevented by the upper strata, which proved to be solidified. The lowermost of these two
strata bore some marks of crystallization, the upper had none, and proved' so firm as to resist a straw, with which I attempted to perforate it, to open a passage for the sublatent liquid. On removing part of these superior strata, and decanting the remaining fluid, the crystals of the lower stratum appeared very distinctly to be rect angular prisms of equal planes, a few of them on one side of the glass surmounted by quadrangular pyra mids, but most of them by dihedral summits. This ex periment I repeated several times, and the results coin cided.
In order to ascertain whether these phenomena arose from a decomposition of the alcohol, or from the separa tion of foreign substances previously held by it in solu tion, the products of several of these experiments were mingled together in a stoppered matrass ; the whole was then raised to the temperature of about 120", by a wa ter bath of that temperature. The substances forming the different strata united together, and formed a colour less liquor, which had the specific gravity and all the other properties of the alcohol from which it was obtain ed. This experiment was repeated several times, and the results were uniform, affording sufficient evidence that the alcohol had not been decomposed by this process, but 'hat the superior strata consisted of foreign substances ss Inch it had held in solution. The variety in the form of the Lrystals obtained by former experiments, was there ore most likely occasioned by the presence of these fo rt ign substances, a phenomenon not uncommon in che mistry.
The result of these experiments led me now to per t dive, that the assumption that alcohol, prepared by Rich ter's process, is perfectly pure, or at most contains only a very minute portion of water, is entirely gratuitous. l'hc diluted alcohol of commerce, from which the more Lonccntrated is obtained, is well known to contain differ ent volatile impurities; and since Richter's process makes no provision for the separation of these, we ought rather to expect still to meet with some portion of them in alco hol prepared in this manner.