been attracted, the card is withdrawn and a clean card or paper substituted; the current is then interrupted, and the particles fall off and are collected. M. Fouqué has separated by this process microscopic prisms of felspar, the presence of which was not revealed in the rock even when a strong lens was used; but a small quantity of glassy matrix always remained associated with them. In this case the experiment was made upon an impalpable powder.* The possession of the heavy liquids described on pp. 29 to 31, the earliest of which was introduced by Sonstadt for the determination of specific gravities and the separation of gems from sand, has given to geologists a most valuable method for the a Fig. 13. isolation of the constituents of rocks. It is clear that if we prepare a solution of density intermediate between the densities of any two constituents, one of these will float up to the surface while the other will sink. If the lighter mineral is the only one to be collected and examined, the operation may be performed in an ordinary beaker and the surface - material skimmed off with a spatula. For economy of the liquid, the beaker should be fairly narrow, since some depth of liquid must be used to allow of perfect separation. If Klein's convenient borotungstate of cadmium solution is used, the powdered rock must be treated beforehand with dilute acid to ensure the removal of carbonates. The material must be well stirred on immersion, and both top and bottom layers stirred later to prevent entangling of inappropriate constituents d in either. The particles when removed must be well washed with distilled water, or with benzole if methylene iodide is used in the separation; the washings are collected in a dish and evaporated down until a concentrated liquid is again obtained for future use. The material separated, when washed and dried, should be carefully searched over with a lens or low microscopic power, since some composite grains are sure to be included. Any doubtful object must be rejected if a quantitative analysis is contemplated; or, for ordinary qualitative tests, only the purest grains must be selected. *M. Fouqué also notes that, contrary to expectation, chlorite is not picked out by the electro-maguet. But in the majority of cases more delicate means of separation will be required, and a number of instruments with taps and closed chambers, and with means of adequately mixing the material with the liquid, have been brought forward in the last few years. Probably one of the neatest is the apparatus of M. Thoulet* (fig. 13). A glass tube, a, 15 mm. in diameter and about 30 cm. long, is graduated in cubic centimetres. At its base it is prolonged by a narrower tube, b, in which are two taps. Between the taps a tube, c, enters, bearing above a rubber prolongation, which is closed at will by a piece of glass rod thrust into the end. The tube a can be closed by a rubber cork through which a tube passes which may be connected with an air-pump. To perform a separation, a quantity of the dense liquid is poured into a. The powdered rock is added, the air-pump is applied so that the particles may be freed from bubbles, and the minerals of greater specific gravity than the liquid will fall to the bottom. These are drawn off through the two taps. After each drawing off of the heavier particles, the tap d is closed, and the liquid that has run out is drawn up once or twice into b by sucking some of the air out through c. In this way the last particles are washed down out of b into the receiving vessel. A fine tube bent upwards at the end, through which water is allowed to run, also serves to wash out b; but the liquid becomes thus further diluted and requires a longer process of concentration before it can again be used. e is then also closed; and water can be added from above until the liquid is sufficiently diluted for a fresh mineral to descend. To procure a solution of particular density, the amount in the tube a is read off, and water is added according to the following formula: Volume of Vol. of liquid × (Density of liquid - Density of required mixture) = The two liquids are mingled by opening the tap d and blowing lightly through c. For ordinary purposes simpler apparatus works extremely well. Thus Herr T. Harada, about 1881, used what is practically an ordinary separating-funnel of the shape of a pear rather than a globe. The stopper at the top and the tap below permit of the thorough mixing of the liquid and the powder by agitation, 'Séparation mécanique des divers éléments minéralogiques des roches." Bull. Soc. Min. de France, t. ii. (1879), p. 17. The apparatus is sold by Dagincourt, 15 Rue de Tournon, Paris, for 35 francs. though care must be taken lest particles remain clinging to the sides of the vessel above the surface of the liquid. The objection to the use of taps is, however, obvious, and becomes more and more forced on one in practice. It is difficult, moreover, to get a tap of sufficiently large bore unless the instrument has been specially made. The tap of Harada's apparatus may be got rid of by substituting a piece of rubber tubing and a spring-clip, such as are often used with burettes. The portion below the tap or clip must be carefully washed out, to prevent the accumulation there of crystallised products from Klein's solution, which will check free action when the liquid above has to be drawn off. The power of closing the vessel by a stopper above is useful in preventing the too rapid flow of the solution, since the outfall of the heavy particles on the opening of the lower orifice can only occur as air rises through the liquid into the upper part of the vessel. But very many useful separations can be performed in the simplest possible manner in an ordinary open chemical funnel about 8 cm. across. A rubber tube and clip are fixed at the outlet of the funnel, and, in place of shaking, the powder and the liquid are mixed by stirring with a rod. The clinging of particles to the sides of the funnel and glass tube canses occasional errors, and all such sluggishly ascending or descending grains must be touched and kept moving with the glass rod. The rubber tube must be removed and thoroughly cleaned before putting the instrument away; a bent tube, like that used with Thoulet's apparatus, serves well to wash out the part below the clip during each successive separation. Dr. J. W. Evans (Geol. Mag., 1891, p. 67) has found the following a safe and thorough method of removing the heavy minerals without drawing off any of the upper material :-Take a thistle-headed funnel or pipette, the tube of which is fairly long and will fit into that of the separating funnel from above. Surround the end of this with a piece of rubber tubing, so that it can be thrust down into the upper part of the neck of the separating-funnel and will there act as a stopper. When the separation has taken place in the liquid, and the heavier minerals are all resting in the tube of the separating funnel, just above the clip, insert this stopper, which should be ciosed during its descent by a piece of glass rod thrust down into it from above. Then remove the rod, open the clip, and the heavier materials will come out as usual, but without any necessity for precaution in their release. Pour distilled water down the tube of the stopper, and the tube of the separating-funnel will be thus efficiently * Except when liquids diluted with benzene are employed. washed out. Close the clip, remove the stopper, and a second separation can be made by further dilution of the liquid. Mr. W. F. Smeeth* has devised a strong and simple separating. instrument which dispenses with the use of clips or taps. This consists of an urn-shaped vessel with sloping sides, closed above by a large stopper and open below, the contracted base being ground so as to fit into a tubular bottle, which has an expanded lip and which acts as a support. An extra glass-stopper with a long handle is made so that it can be passed through the upper opening of the urn and thrust down so as to close the passage into the bottle. The dense liquid and the powder are shaken together in the instrument, the ordinary stopper being in its place and the urn and bottle remaining connected. The heavier particles will thus descend into the bottle. Now pass in the long stopper, moving it in the liquid as it descends so as to free it from any adhering particles of the lighter material; close the base of the urn with it and lift off the urn from the bottle. The two classes of materials are thus efficiently separated from one another. If the urn is now fitted into a second and similar bottle, a further separation can be proceeded with by removing the long stopper and diluting the liquid to the requisite amount. Or the original bottle can be at once cleaned out and some other separation carried on in it without additional apparatus. Mr. Smeeth informs us that the instruments were made for him by Messrs. Becker, of London. He suggests as the simplest type of this apparatus an ordinary funnel cut down so as to leave a stem half-an-inch in length. This is connected by a piece of rubber tubing with a glass tube which serves as the receiving bottle. The materials are mixed with the liquid in the funnel; its lower orifice is closed by a glass rod carrying an india-rubber cork at the end; and the rubber tubing can then be pulled away safely from the shortened neck of the funnel, the two parts becoming thus disconnected. Herr A. Hauenschild has recently invented a very neat and compact apparatus, consisting of a large glass tap, with two lateral notches instead of a central opening, whereby the passage of particles is facilitated. Each end of the short wide tube in which this tap is placed is ground, and over each a tubular vessel with a foot is fitted. The upper and lower ends of the instrument are thus identical, except that one is inverted, its foot remaining upwards. The liquid and powder are placed in Scientific Proceedings of the Royal Dublin Society, vol. vi. (new series; 1888), p. 58. Sollas, Trans. R. Irish Acad., vol. xxix., p. 430. the lower vessel, into which the central tube, containing the closed tap and supporting the upper inverted vessel, is now fitted; the instrument is then inverted, and, in due course, the heavier particles can be drawn off into the second vessel, which is now below. Hauenschild's separation-apparatus is sold by Muencke, 58 Luisenstrasse, Berlin, for 12 marks. Prof. Sollas (Nature, vol. xlix., p. 211) has shown how particles may be removed from any zone of his diffusion-column (see p. 31), by inserting a pipette of 1.5 mm. bore, plugged at its base by a stem of Esparto grass, round the end of which a little unravelled cotton thread is wound. This stem is inserted from above; the pipette is thrust down until its end lies against the grains to be extracted; the plug is then pulled back slowly, and the fluid and grains follow it into the tube. A very thin glass rod, bent up at one end into a crook, is then thrust down parallel to the pipette, and the crook is pulled up so as to enter and plug the bottom of the pipette. Now remove both together, clean off foreign grains from the outside of the pipette with blottingpaper, and wash out the collected and isolated grains. Where a liquid of high density is required, Dr. J. W. Retgers has recommended the use of certain easily fusible salts, on which he has made a careful series of experiments. In 1893 (Neues Jahrb. für Min., 1893, Bd. i., p. 90) he introduced thallium silver nitrate, which melts at 75° C., giving a density of nearly 50. It can thus be manipulated on a water-bath, and can be diluted by water, added carefully drop by drop. Sulphides, however, are attacked by it. The joint thallium nitrate and acetate, melting at 65° O., with G. = 4.5, gave also good results (Neues Jahrb., 1896, Bd. i., p. 212). In a later paper (Ibid., 1896, Bd. ii., p. 183), Retgers recommends mercurous nitrate as cheap, melting at 70° C., with G. 4.3. Quickness is necessary during a separation, as some decomposition of the salt occurs. The most satisfactory salt is now reported to be thallium mercurous nitrate, melting at 76° C., with G. 5.3. This can be diluted with water, and has no effect on sulphides. = The procedure is to effect the separation in the melted salt in a test-tube, with a fair depth of liquid. When the heaviest minerals have sunk, remove from the water-bath, and dip the outside of the test-tube into a glass of water, moving it about, so as to cool it quickly. This prevents the minerals from floating up again during the rise of density on cooling. Then break out the bottom of the tube, remove the glass-fragments, and melt off the lower layers into a beaker by means of an obliquely |