in alcohol deposits, upon cooling, in the crystalline state that portion of the solid body which the liquid is not capable of retaining in solution at the lower temperature.

A remarkable anomaly is often observed in the cooling of such solutions, viz., that when kept perfectly still and in covered or stoppered vessels, they may, without losing their liquid form, be cooled much below the temperature at which they usually crystallize; under these circumstances, a sudden crystallization may be brought about by agitation, the introduction of a solid body, &c., &c. Of this the following are examples. Water, when cooled in open vessels and not at absolute rest, below 0° C., solidifies in a crystalline mass: but in stoppered bottles or thermometer bulbs, it may often be cooled as low as 6° without freezing, and when the vessel is shaken or opened, or a piece of ice put in, crystallization suddenly takes place, the crystals beginning to form at several points at once and quickly radiating through the liquid; the temperature at the same time rises to 0°. The formation of mists over the surface of lakes sometimes observed at the moment of freezing is perhaps dependent upon this action. Pure phosphorus melted in a capsule under warm water will remain all night in the liquid state at a temperature of 4.5° C., and solidify when poured out upon the hand. (Clark, Ed. Journ. of Sc. 7, 281.) Phosphorus boiled with caustic potash retains its fluid state at the ordinary temperature for months, when kept under the solution, and solidifies when touched with a dry solid body. (Poggendorff.) Sulphur sublimed in small melted drops retains its fluid state over night at the ordinary temperature, and solidifies when touched with any solid body. (Faraday, Qu. Jour. of Sc. 21, 392; also Pogg. 7, 240.) Sulphur precipitated by water from chloride of sulphur remains fluid at the bottom of the liquid, but solidifies immediately on exposure to the air. (Poggendorff, Pogg. 7,241.) Glacial acetic acid, which in open vessels crystallizes at +16° C., may be cooled to 12° even with agitation in stoppered bottles without solidifying: if, however, the vessel be opened and shaken, it crystallises at +15°, even when the outward air is warmer than the acid, the crystallisation beginning at the top and quickly shooting through the whole mass. (Löwitz, Crell. Ann. 1790, 1, 209; Geiger, Schw. 15, 231.) Oil of anise, when cooled at rest in a closed vessel, often remains liquid till it is shaken, crystallization then taking place instantaneously. (Buchner, Repert. 15, 64.) Scheererite after fusion may be kept in the liquid state at ordinary temperatures for several days, and crystallizes instantly on being touched with a platinum wire or glass rod. (Stromeyer, Kastn. Archiv. 10, 114.)

Many salts dissolved in hot water exhibit this anomaly, especially sulphate of soda. A hot solution consisting of equal parts of water and crystallized Glauber's salt (sulphate of soda with 10 atoms of water of crystallization) does not crystallize either on slow cooling, or when quickly cooled by immersion in cold water,-whether it be contained in a barometer-tube freed from air by boiling; or in an exhausted well closed vessel; or in an open vessel with a layer of oil of turpentine upon its surface, (Gay-Lussac); or in a vessel containing air, either well stopped or merely furnished with a loose cover, (Schweigger); or in an open vessel under a bell jar full of air and closed at the bottom with a waterjoint; or in open bottles placed in a quiet situation; or in an open glass enclosed in a stoppered vessel containing air and some potash to dry it, in which Glauber's salt effloresces and when washed down again does not cause instant crystallization but dissolves. (Ziz.) The crystallization of

a solution cooled in this manner is often rought about instantaneously, often again after a short time: (1.) By agitation, viz., when the solution has been cooled in an open vessel. (2.) By access of air caused by opening the vessel, the crystallization taking place the more quickly in proportion to the size of the opening; some degree of motion appears also to be necessary. In this case, the crystallization begins at the top where the solution, the vessel, and the air come in contact with each other; it is only when a particle of dust falls in on opening the vessel that the crystallization begins a little under the surface. When the solution has been cooled in vacuo, a bubble of air, hydrogen, carbonic acid, or nitrous gas is sufficient to set up the crystallization. (Gay-Lussac.) (3.) By contact of the solution with a solid body (a glass rod, flint, iron wire, crystal of Glauber's salt, or a grain of dust floating in the air.) These bodies do not bring about the crystallization when they have cooled in contact with the hot solution, nor (excepting Glauber's salt) when they are wetted or warmed before contact with the solution. (Ziz.) In these cases crystallization is effected by the action of foreign bodies. If a solution of 8 parts of Glauber's salt in 9 parts of water be left to crystallize, the whole then warmed in a flask to between 50° and 55° Č., till only about of the crystals remain undissolved, and the flask corked up and cooled, it often happens that the remaining crystals, instead of causing the rest to crystallize, are themselves completely dissolved,-slowly when the flask is inclined in such a manner as to bring them in contact with the upper strata of the liquid, more quickly on agitation, which however is very likely to cause crystallization. If, on the other hand, the solution formed between 50° and 55° be poured off from the crystals into a basin and allowed to crystallize, the mother-liquid thus obtained will not dissolve the of the crystals above mentioned. There are therefore two solutions to be distinguished, (1.) The saturated solution, i. e., the liquid which remains after crystallization of the superabundant quantity of salt, from a hot solution in an open vessel, and, (2.) The supersaturated solution, i. e., the solution saturated at a high temperature and cooled in a close vessel; this latter can even dissolve an additional quantity of salt but deposits at a lower temperature crystals of sulphate of soda containing 1 atom of the salt and 8 atoms of water. (H. Ogden.) A solution of 2 parts of Glauber's salt in 1 part of hot water yields on cooling in close vessels hard transparent crystals of sulphate of soda with 8 atoms of water, which, when the supernatant liquid is made to crystallize by any of the preceding methods immediately become opaque. (Coxe, Ziz.) When 51 parts of crystallized Glauber's salt are dissolved in 49 of water, and the solution after cooling below 10' C. made to crystallize suddenly by any of the preceding methods, nearly of the Glauber's salt is deposited, and the temperature rises to 13° C. This is attributed by Thomson to the conversion of liquid water into solid water of crystallization, a supposition agreeing pretty well with calculation (the development of heat consequent on the passage of the salt from the liquid to the solid state must however be included in the calculation. Gm.) The assertion of Thénard (Schw. 15, 257), that after this crystallization there remains a motherliquid which is no longer saturated with salt at the existing temperature seems to be erroneous. Thomson, on the contrary, finds that the motherliquor, from its rise of temperature, holds in solution a corresponding quantity of salt, a great part of which crystallizes out when the temperature is brought back to 10°.

A hot concentrated solution of chloride of calcium cooled in a stop

pered bottle crystallizes when agitated, without requiring the bottle to be opened, the temperature at the same time rising very considerably. (Coxe.) The addition of a few drops of oil of vitriol or oxalate of ammonia does not cause it to crystallize; but crystallization takes place when cold water is poured upon the liquid or a stream of dry air directed upon it. (Ogden.) A solution of 1 part of crystallized carbonate of soda in 4 parts of warm water, when cooled below + 10° C. crystallizes on opening and shaking the bottle; the temperature rising about 8°. A highly concentrated solution of acetate of soda remains liquid when cooled to +10° in a loosely stopped vessel: but on pouring it into another vessel it solidifies to a fibrous mass, the temperature rising to 52·5°. (Gm.) Under similar circumstances Hashoff (Br. Archiv. 38, 326) obtained, upon stirring, a rise of temperature of 59° C. The introduction of a crystal of acetate of soda produces crystallization, but not so quickly as in the case of sulphate of soda. (Ogden.) A hot solution of sulphate of magnesia often remains fluid when cooled in close vessels, and yields granular crystals when shaken. (Coxe.) A drop of alcohol produces in the solution a nucleus from which the crystallization radiates. (Ogden.) A mixture of nitre and sulphuric acid yielded, after heating for some time, a clear liquid which did not crystallize on cooling till a crystal of nitre was thrown in, a rise of temperature then taking place. (Green, Gilb. 70, 320.) The following salts exhibit similar phenomena to the above:Hyposulphate of soda, (Heeren); carbonate, phosphate and borate of soda, (Gay-Lussac); nitrate of lime, sulphate of magnesia, sulphate of copper and nitrate of silver, acetate of lead, (Fischer, Schw. 12, 187); nitrate of ammonia, bisulphate of potash, bichromate of potash, chloride of barium, sulphate of magnesia and ammonia, sulphate of zinc, ferrocyanide of potassium, oxalate of ammonia, tartrate of potash and soda, and tartrate of antimony and potash. (Ogden.) Warm solutions of alum, protosulphate of iron and sulphate of copper deposit considerable quantities of crystals upon cooling; but opening of the vessel and shaking bring out a fresh quantity. (Coxe.) Nitro-saccharic acid also exhibits this property. (Mulder, J. pr. Ch. 16, 293.) The following salts, on the contrary, crystallize out from a warm solution on the slightest lowering of temperature:-Sulphate and hydrochlorate of ammonia; sulphate, chlorate, nitrate and neutral chromate of potash; chloride of potassium, chloride of sodium; baryta, strontia, nitrate of baryta, sulphate of magnesia and potash, nitrate of lead, corrosive sublimate and oxalic acid. (Gay-Lussac, Ogden.) Generally speaking, those salts which exhibit the anomaly crystallize in combination with a large quantity of water of crystallization (bichromate of potash and nitrate of silver excepted), and those which do not exhibit it, with little or no water of crystallization. This anomaly is explained by Berthollet (Statique Chim. 1, 32,) and by Gay-Lussac, on the supposition of a sluggishness of the ultimate particles of the salt. So also Thénard supposes that the particles are brought by agitation into different relative positions. (It must be observed however that shaking in closed vessels often fails to produce crystallization.) At all events it may be admitted that in those cases in which the cohesive force has gained the preponderance over other forces, such as the affinity of a body for heat or for ponderable solvents, mechanical disturbance is often required to bring it into active operation. Compare Gay-Lussac (Ann. Chim. 87, 225, also Schw. 9, 70; likewise Ann. Chim. Phys. 11, 301); Schweigger (Schw. 9, 79), Ziz. (Schw. 15, 160), Thomson (Ann. Phil. 19, 169); H. Ogden (N. Ed. Phil. J. 13, 309).

This is

II. If the crystallizable body has been liquefied by combination with another ponderable body, the latter must be separated from it. effected either by driving off the latter in the form of vapour, at a high or a low temperature, in the open air or in vacuo, as the case may be; or else by removing the solvent by virtue of its affinity for another ponderable body. Thus nitre crystallizes from an aqueous solution on the addition of alcohol, camphor from an alcoholic solution on the addition of water, iodine from its solution in hydriodic acid gas on the introduction of a small quantity of chlorine.

During crystallization the following phenomena are observable:-1. The more slowly the liquefied body is brought back to the solid state, and the more the liquid is kept at rest, the smaller is the number and the greater the size and regularity of the crystals; but if the solvent be cooled or separated quickly, the crystals are numerous but small and ill defined. For in the former case, the particles of the solidifying body have time to unite themselves regularly with those which separate first from the fluid and form nuclei of crystallization; if on the contrary the crystallization takes place rapidly, a great number of particles solidify at the same time, each forming a nucleus to which other portions may attach themselves, and thus we obtain a number of crystals irregularly formed and interlacing each other in all directions. In this consists the difference between sugar-candy and loaf-sugar; similarly, all granular and fibrous bodies, such as statuary marble and fibrous gypsum, must be regarded as collections of imperfectly formed crystals equal in number to that of the grains or fibres. To obtain crystals as large and regular as possible, Leblanc recommends (J. Phys. 55, 300) to allow a solution not quite saturated to cool slowly, so that none but distinct crystals may be formed, then to pick out the best formed of these and lay them separate from one another in a solution of the same salt, which by gentle warming in contact with the salt has been made to hold in solution a quantity of it just a little greater than that which it can retain at the ordinary temperature, so that it may deposit this excess on the crystals laid in it. This treatment is repeated till the crystals have obtained the desired magnitude, care being taken to turn them frequently, because the surfaces resting on the bottom are in a less favourable position than the rest for taking up fresh particles. The trouble of repeatedly preparing a slightly supersaturated solution may be saved by suspending in the upper part of the liquid a quantity of the salt contained in a bag of muslin or a funnel. For whenever the temperature rises, the warmer part of the liquid will come to the surface, where it will dissolve a portion of the suspended salt, and then becoming specifically heavier, it will sink to the bottom of the vessel where the crystals are placed: thus the crystals will continue to grow, and nothing further will be necessary than to turn them frequently.

2. Generally speaking, crystals exhibit, so far as can be observed, the same external form on their first formation, that they have at a subsequent stage. Probably the primitive form is first developed and passes into the secondary by the subsequent annexation of matter according to fixed laws. Thus the octohedron of alum exhibits, as lately observed by F. Richter (Zeitschr. Ph. v. W. 3, 348), the same truncations of the edges and summits at its first formation as after its complete development. When however the formation of crystals is suffered to go on at intervals and in liquids of different nature, the external form may be greatly modified. This may be observed in certain crystallized minerals which often possess a nucleus and an envelop of different colours. Fluor-spar exhi

bits, according to F. Richter (Zeitschr. Ph. v. W. 2, 111), the following forms: : a rose-coloured octohedron within a green cube; a yellow cubooctohedron (fig. 4) whose octobedral faces alone are covered with a violet layer; a blue dodecahedron (fig. 3) enveloped in a green cube; a cubooctohedron (fig. 4) within a pyramidal cube (fig. 9), &c. In such cases the inner and outer mass always exhibit parallel cleavage. Similar phenomena are also seen according to Richter in calcspar and other minerals.

3. Crystals are formed in situations where the principle of fluidity is removed from them, or where they are led to attach themselves by adhesion. Hence they form on the surface of the liquid in so far as evaporation and cooling by the influence of the air, or adhesion of the air to the crystals can give rise to their production; also on the bottoms and sides of the containing vessels, inasmuch as these abstract heat, and exert adhesive power on the crystals; lastly, on solid bodies immersed in the liquid, such bodies acting by adhesion. For the most part, crystals deposit themselves more easily on wood and string than on porcelain, glass, and metal; more easily on porcelain than on glass; on rough than on smooth glass-compare Griffiths (Ann. Pharm. 22, 210). Hence it is easy to explain the fact that when a glass containing a crystallizable liquid is scratched with a glass rod, the crystals deposit themselves in preference on the scratches. The first-formed mass of crystals serves for the rest to attach themselves to, and attracts them more strongly than foreign bodies would. Thus, according to Löwitz, the introduction of a crystal of nitre into a solution of nitre and Glauber's salt, prepared hot and subsequently cooled, causes the nitre to separate alone: a crystal of Glauber's salt removes only the Glauber's salt; whereas if the solution be left to itself, both salts crystallize out together, the crystals interlacing each other. If a drop of a solution of gypsum be left to evaporate on a freshly cleft surface of a crystal of gypsum, the microscope will show an innumerable collection of crystals of gypsum formed on the surface, all parallel to one another and to the original crystal. If the evaporation takes place on the surface of a foreign body this parallelism is not observed (Frankenheim, Pogg. 27, 516.) When a solution evaporates below its boiling-point, the first crystals are usually deposited on the sides of the vessel at the uppermost surface of the liquid: another portion of the liquid often rises through these and yields by evaporation new crystals which ultimately make their way over the edge of the vessel. This is Efflorescence. When crystals form at the bottom of a liquid, a current is produced, because the individual crystals take from that part of the solution with which they are immediately in contact as much of the salt as is possible under the existing circumstances; consequently this part of the liquid becomes lighter and rises to the surface, its place being supplied by a more saturated portion of the liquid.

4. When a body crystallizes from solution in a liquid, and the latter is not completely removed by evaporation, there remains a portion called the Mother-liquor (Eau mère, Mutterlauge). This liquid holds in solution. as much of the crystallizing body as is consistent with its quantity and temperature. It often happens, especially when crystallization proceeds rapidly, and the crystalline laminae in the act of uniting, leave small spaces between them, that small and (even with regard to the same substance) very variable quantities of the mother-liquid remain enclosed in the crystalline mass, forming the Water of Decrepitation (Zerknisterungswasser). Crystals which contain liquids thus enclosed and do not melt below the boiling point of the mother-liquid, exhibit, when heated, the