The crude or " once run" naphtha obtained by this process is again distilled with steam, and the products are coilected at different temperatures, as required. A residue, amounting to about 7 per cent. of the " once run " naphtha, is left in the still ; it has a thick, tarry consistence, and is run in with the creosote oils, or sold to rosin-grease makers at 508.-65s. a ton. In distilling with steam, a large quantity of water passes. over with tha product ; as this continues during the whole opera tion, the distillate is received in a " separatory " apparatus, so as to allow the water to escape. This consists of a mag,nified "Florentine receiver," in the form of a rectangular iron tank. The " boiled tar " is distilled in vertical or horizontal stills, heated directly by a furnace. The heating should be very gradual, as this tar contains considerable quantities of water. About 1000 gall. yield 200 gall. of rough naphtha, which should give about 120 gall. of naphtha suitable for burning, .:Scc., and about 320 gall. of creosote oils. The residual pitch will amount to 35-55 per cent., according to the temperatures at which the distillation has been conducted.
Ernest Smith says that in distilling. Scotch cannel tar by the Scotch system, a little light oil, sp. gr. 0.850, first passes over, after which a rise takes place, and the whole of the light naphtha marks 0.890. The sp. gr. of the second naphtha is 0.910 ; to collect this, the receivers are changed, when the distillate rises to 0.930. The crude light naphtha from the best tars yields 10 per cent. at 120° (248° F.), with a sp. gr. of 0.880. He states that paraffin appears more copiously in tars obtained from cannel coal at a low heat, and that generally the anthracene, before purification, contains a larger quantity of paraffin. This certainly will help to account for the great differences in Scotch solvent, aud explaiva the discordant figures given by various authors.
A continuous process of distilling has been lately introduced : by running the tar over molten lead the more volatile part is driven off, and the residue is received in retorts for further treatmcnt.
The "French system " of distillation differs mainly in drawing off the aqueous portion of the tar at a much lower temperature. It is frequently effected in jacketed stills, or hy means of circulating steam pipes. The same method is obviously applicable for drawing over the lighter oila, or for carrying on the diatillation BO BB to separate the constituents of the tar at one operation. " Liquid pitch," or " thick tar," contains a little of what corresponds to the second light oils of the English method, and all the heavy oils ; " fat pitolt " is deprived of all the light oils, and partiaily of tho heavy oils ; " dry pitch " is obtained by heating until liquids are given off. The condenser uaually employed consists of a series of 4-in. flanged iron pipes, joined together by cast-iron elbow-pieces, and arranged in 11 water-tank in zigzags or ovala. The joints are best made with
sheet lead ; where cemented joints are required, slaked lime worked up with a little tar makes a lute, which, after hardening, stands for it long time. For convenience in cleaning the condensing tubes, they are joined by tee-pieces, and the ends which project through the tank are closed by iron plugs or caps. The length of worm required is a matter of some importance : the liquids which come over at the first stage of the operation should be well cooled, whilst the heavier oils, if cooled too much, would choke up the tubes. The cooling must be so regulated that the distillate shall flow freely away to the receivers. About 160 sq. ft. of condensing surface is sufficient for a still working off 2000 gall. of ordinary tar per twenty-four hours. If the product solidify on cooling, or if much naphthalene come over with the creosote oils, it may he necessary to stop the cooling altogether, or even to heat the water in the worm-tank by injecting steam. It is not easy to define precise regulations, and, in operating upon a new sample of tar, it is by far the best plan to watch the distillate itself ; for not only is it a matter of great importance to know how to adjust the condenser, but the proper time forchanging the receivers must be ascertained, especially if the preliminary distillation is used for partial fractionizing.
In treating ordinary English tar, the distillation is thus conducted :—The firing is generally got up as the filling progresses. It is important to raise the heat very cautiously, and to watch for the commencement of the distillate going over ; this can be done by feeling the condensing tube, and, as soon as it becomes warm, the fire should be checked, the object being to distil very gradually at first, and to make tolerably certain that all the watery portion has come over. If frothing occur, the still head may be cooled with water, unless the still he of cast iron. The water present in the tar makes the distillate come over more copiously. The first distillate consists of water charged with ammonia (" ammoniacal liquor ''), and the more volatile hydrocarbons. As soon as the water has stopped, the receiver is changed, and the heating is slightly increased. The product which now passes over 1.2 called " first light oils," and amounts to 8-10 per cent. of the quantity of tar originally taken. During this time, the temperature gradually rises, from the fact that the tar parts from its more volatile portion, and what remains has a much higher boiling-point. In the earlier part of the process, the refrigeration must be a.s perfect aa possible ; but as the receivers are changed for the heavier and less volatile portions of the distillate, it is not so essential to keep them very cool. This is not entirely a matter of economy of water : a little heat facilitates the flow of the condensed liquids, and hastens the distillation.