Page:Encyclopædia Britannica, Ninth Edition, v. 7.djvu/280

262 DISTILLATION boiling point of that dew, i.e., of that which is just going over. The composition of the vapour as above given must not be confounded with the composition by ^ue^ght of the distillate. To obtain the latter we must multiply each of the two volumes by the density of the respective vapour, or, what comes to the same thing, by its molecular weight as expressed by the chemical formula. lu our case the vapour volume ratio water _ 55 ether "" 910 corresponds to the weight ratio 55 x H n O 55 x 18 1 901 x C 4 H ]0 ~ 910 x 74 ~ 68 neavly This consideration strips of its apparently anomalous character what we observe when vegetable substances con taining essential oils are distilled with water, when we find that these oils, although boiling far above 100 C., go over with the first fractions of the water. Take the case of lemon oil, which boils at about 174 C. The molecular weight of the oil is 136 = C 10 H 10 ; its vapour tension at 100 is 70 mm. Hence what goes over at first when lemon peel is distilled with water should contain oil and water in the proportion Oil of Lemons. Mol. wt Yap. tension. Mol. Wt. V:ip. tension. 136 x 70 : 18 x 760 = 12 : 17 (nearly). The oil, although the less volatile substance of the two, being present in small quantity, but finely diffused, is soon completely driven over. No doubt the latent heats of vaporization of the two constituents have some thing to do with the composition of the vapour formed, as the chance of every particle of the mixture to be vaporized is obviously the greater the less its latent heat of vaporization. After what has been said it will be clear that in the dis tillation of a mixture of two substances of approximately equal molecular weight and latent heats of vaporization, supposing neither to predominate overwhelmingly over the other, the one with the lower boiling point will predominate in the early, and the other will gradually accumulate in the later, fractions of the distillate. And similarly with mixtures of three or more bodies. The further the respec tive boiling points are removed from one another the more complete a separation can be effected ; but in no case is the separation perfect. It is, however, easily seen that the analytic effect of a distillation can be increased by causing the vapour, before it reaches the condenser, to undergo partial condensation, when naturally the less volatile parts chiefly will run back. This artifice is largely employed by chemists, technical as well as scientific. The simplest mode is to let the vapour ascend through a long, vertical tube before it reaches the condenser, and to distil so slowly that a sufficiently large fraction of the vapour originally formed fails to survive the ascent through the cooling influence of the atmosphere. A more effective method is to let the condensed vapour accumulate in a series of small receptacles insertedbetween flask and condenser, constructed so that the vapour cannot pass through the receptacles without bubbling through their liquid contents, and so that the liquid in the receptacles cannot rise above a certain level, the excess flowing back into the next lower receptacle or into the still. But the most effective method is to let the vapour ascend through a slanting condenser kept by means of a bath at a certain temperature, which is controlled so that while the liquid in the flask boils rapidly, the dis tillation only just progresses and no more. The general principles thus stated regarding fractional distillation are liable to not a few exceptions, of which the following may be cited as examples. A solution of one part of hydrochloric acid gas in four parts of water boils (constant) at 110 C. i.e., 10 above the boiling point of water, although the acid constituent is an almost permanent gas. This, however, is easily explained ; there can be no doubt that such an acid is a mixture of real hydrates, i.e., does not contain either free water or free hydrochloric acid. A similar explanation applies to the case of aqueous oil of vitriol, which boils the further above 100 the stronger it is, although the vapour may be, and in the case of acids contain ing less than 84 per cent, of real acid really is, pure steam. The following cases, however, can scarcely be disposed of by the assumption of the interference of chemical action. Propyl alcohol boils at 97 C., water at 1UO ; and yet a mixture of the two, as Pierre and Puchot found, when distilled always commences to boil at 88 5 with formation of a distillate of the approximate composition C 3 H S + 2 78H 2 O ; and this particular aqueous alcohol boils without apparent decomposition at 88 3. Some time later Dittmar and Steuart made a precisely analogous observation with regard to aqueous allyl alcohol. A strong temptation exists to explain these anomalies by the assumption of definite hydrates in the aqueous alcohols, and this hypothesis would serve in the meantime were it not for the curious fact, discovered by the two French chemists named, that amyl alcohol and water (two liquids which do not mix), when distilled simultaneously out of the same retort, go over at a constant temperature less than 100 D , and with formation of a distillate which, although it is not even a mixture, has a constant composition. The most natural explanation of these phenomena is to assume them to be owing, not to chemical action, but rather to an exceptional absence of chemical affinity between the two components of the mixture, which for once gives the physical forces fair play. DRY (DESTRUCTIVE) DISTILLATION. Of the great number of chemical operations falling under this head, we can notice only those which are carried out industrially for the manufacture of useful products. Of such the most important are those in which wood, coal, shale, and bones form the materials operated upon. But as these processes form so many important industries, which have all special articles devoted to them, we must confine ourselves here to summing up shortly the features common to all. In all cases the " retorts " consist of iron or fire-clay semi-cylinders placed horizontally in a furnace and con nected by iron pipes with refrigerators, and through these with gas-holders. Within these retorts the materials are brought up, more or less gradually, to a red heat, which is maintained until the formation of vapours practically ceases. Each of the materials named is a complex mixture of different chemical species. Wood consists mainly of cellulose and other carbo-hydrates, i.e., bodies composed of carbon and the elements of water ; in coal and shale the combustible part consists of compounds of carbon and hydrogen, or carbon, hydrogen, and oxygen, richer in carbon than the components of wood ; bones consist of about half of incombustible and infusible phosphate of lime (bone earth) and half of organic matter, of which the greater part is gelatine (compounds of carbon, nitrogen, hydrogen, and oxygen), and the lesser is fat (compounds of carbon, hydrogen, and oxygen). The chemical decomposi tion in each case is highly complex. An infinite variety of products is invariably formed, which, however, always readily divide into three : 1st, a non-volatile residue, con sisting of mineral matter and elementary carbon (" wood charcoal," " coke," &c.) which, in the case of animal matter, contains chemically combined nitrogen ; 2d, a part condensible at ordinary temperatures which always readily separates into two distinct layers, viz. : (a) an

aqueous portion (" tar-water "), and (b) a semifluid, viscid,