Fig. 19.

jacket of clay, or metal, surrounding the burner, and so preventing draughts, as shown in fig. 19. Iron triangles also to rest on the top of the jacket, tripod, or retort-ring, are in constant requisition. A triangle of iron wire, sheathed with three pieces of tobacco-pipe, is useful for supporting small porcelain crucibles that have to be made red-hot, but the mass of the tobacco-pipe interferes with the attainment of a very high temperature. Small triangles of platinum wire are far more convenient.

Flasks and retorts may be heated at some little distance over an argand flame without the interposition of any medium. But with the gauze

burner, and more particularly with the Bunsen burner, it is advisable to protect the bottom of the vessel by a piece of stout wire gauze. Beakers should be placed either on a flat iron plate or sand-bath, and the heat be raised cautiously. Porcelain evaporating dishes may be heated almost anyhow, according to circumstances; but when the flame touches the bottom of the dish some little care is necessary, both at the commencement of the operation, and when the liquid is evaporated nearly to dryness. Watch-glasses require very careful heating. They are best held by the thumb and finger over a small flame, but may be supported by forceps, wire-triangle, or special watchglass holder. When containing liquid, to be heated by means of a sand-bath, they should not be depressed into, but just rest on, the top of the hot sand. Test tubes may be readily heated in the flame of a spirit lamp or gauze-burner. They should be held between the thumb and fingers, and be constantly shaken from side to side, especially during boiling, to prevent any

D

sudden expulsion of the contained liquid, with which they should not be more than half filled. When the boiling has to be long continued, a piece of paper or cloth may be bound or twisted round the upper part of the tube, so as to protect the fingers. Test tube holders are rarely of much use.

Fig. 20.

Small capsules and crucibles, of platinum or Berlin ware, may be heated to redness over an argand (fig. 20), or to full redness over a Bunsen burner, or by the blowpipe flame. They may be held either with forceps or on triangles of wire or tobacco-pipe. When a strong heat is required, they should be surrounded with a small jacket of metal or clay.

Sand-baths are usually formed of sheet iron. Some are made rather deep to receive flasks and retorts, others nearly flat for

Fig. 21.

beakers. The sand, which must not be heaped above the level of the liquid to be heated, should be of uniform size and rather coarse. An iron plate, or trellis of thick iron wire, forms an excellent substitute for a flat sand-bath.

The water-bath is used for heating substances to a temperature not exceeding 100° C. A small saucepan, with an evaporating dish for a cover, forms a capital makeshift. The saucepan lid may be replaced by a series of broad rings of tin plate, having

apertures of different sizes to support small evaporating dishes (fig. 21), capsules, watch-glasses, necks of flasks, &c. Beakers, flasks, and retorts, to be heated in a water-bath, should not be allowed to touch the bottom of the bath, but should rest on a piece of tow or folded cloth, as well to avoid danger from bumping as to prevent the communication of an increased temperature by contact with the heated metal.

(20.) Heat is applied to liquids in order to warm, evaporate, boil, or distil them. Evaporations are performed on slips of flat glass, or on watch-glasses, or evaporating dishes. The crystalline forms of various salts may often be recognised by dissolving a grain or so of the salt in a drop or two of water on a glass slip, evaporating until a solid margin appears, and setting aside to crystallise. The residue may then be examined by a lens, or under the low power of a microscope. The evaporation of a small quantity of liquid, contained in a watch-glass for instance, may often be promoted by gently blowing on its surface for a little while. A dish in which evaporation is taking place (fig. 19) may be loosely covered with a piece of filtering paper,, stiffened by a glass strip passed through and across it; and care should be taken not to allow the liquid to boil. In evaporating to dryness, it is well to have the heat lowered as the process approaches completion, and the residue, if considerable, kept constantly stirred.

Ebullition is conducted on a small scale in test tubes, and on a larger scale in flasks,-Florence oil flasks being among the best and cheapest that can be employed. A flask of cold liquid, held over a naked flame, quickly becomes covered with a deposit of moisture, which it is advisable to wipe off once or twice. Ebullition sometimes takes place intermittingly, and with considerable jerking or bumping. This may often be prevented by introducing a coil of platinum wire or piece of tobacco pipe into the liquid, either before boiling or after cooling down a degree or two. If introduced during boiling, it is apt to produce a violent rush of vapour. It is occasionally useful to adapt a long upright tube to

D 2

the mouth of a flask (fig. 22) in which a liquid has been heated,

Fig. 22.

so that any vapour given off may be condensed in the tube, and flow back again into the flask.

Distillation is usually conducted in an apparatus composed of a boiler, condenser, and receiver. On a small scale the boiler is represented by a glass retort, or flask with its bent tube, the receiver by a test tube, flask or bottle, and the condenser by a long glass tube placed between the retort and receiving flask, and surrounded either by a constantly changing layer of water, as in Liebig's condenser, or by a piece of blotting-paper kept moist by the constant dripping of water. Very often a separate condensing tube is dispensed with, and the long neck either of the retort or receiving flask alone employed. Fig. 23 shows a very simple arrange

ment of this description. The liquid is boiled in a long-necked retort, and the distillate conducted by means of a bent adapter into an upright bottle. A twist of thoroughly wetted tow, or lamp-cotton, is tied somewhat tightly round the retort-neck, at about an inch above the point where it enters the perforated cork of the adapter, and the ends of the twist allowed to hang down for two or three inches. A piece of filtering paper of suitable size and shape to embrace almost the entire circumference of the neck, and reach from just above the twist of tow almost to the curve of the retort, is moistened with water, laid upon the retort-neck and bent round it so as to adhere closely. A second piece of filtering paper is at first folded by means of

three creases into four strips and then placed over the other, so that the flaps being bent down, its middle portion will form a

small channel along the top of the retort-neck. It should be about two-thirds the length of the other, and not reach so high up. By this arrangement the water constantly dropping out of a partly plugged funnel from the height of half an inch or so on to the surface of the inner paper, at a little distance from its upper extremity, is conveyed along the channel, spread uniformly over the retort-neck and drained away by the twist of tow. If the short length of neck intervening between the twist and adapter be dry at the beginning, it will continue so throughout the experiment. For further security, however, it may be marked with a ring of grease. Adapters are of all shapes and sizes. Such a one as that shown in the woodcut is easily made out of a piece of tubing or broken retort neck. Others, intended to act more or less as condensers, are represented in fig. 24. Another convenient form of distilling apparatus is shown in fig. 40. The boiler consists of a round flask, from which a long wide bent tube dips through a perforated cork into a Florence flask resting in a basin of water.