upon. It acts also by extracting the acid from salts, and thus leaving the metallic oxide more exposed to the action of reducing agents, and sometimes serves to protect the surface of reduced metals, and to prevent their reoxidation.

Carbonate of soda employed as a blowpipe reagent, should, strictly speaking, be perfectly pure; but the common washing-soda of commerce, after one or two recrystallizations, is sufficiently so for ordinary purposes; the crystals should be thoroughly dried in a porcelain dish on a sand-bath, and afterwards rather coarsely powdered.

Black flux, which has been noticed when speaking of fusion, is also sometimes employed in blowpipe experiments.

Cyanide of potassium (KCy=KNC,, prepared by Liebig's process) is very valuable as a reducing agent; most metallic oxides, when fused with cyanide of potassium, part with their oxygen to this salt, converting it into cyanate of potassa (KO.CyO); the great fusibility of the cyanide, however, prevents its application in many cases. This reagent should be used in the form of a coarse powder.

A rather strong solution of nitrate of cobalt (CoO.NO,) is sometimes useful in experiments on charcoal.

Borax (biborate of soda, NaO.2BO,+10Aq) is a most important blowpipe reagent. It should be finely powdered. The chief value of borax in blowpipe analysis depends upon its property of dissolving certain metallic oxides, forming glasses of peculiar colors. When heated in the blowpipe-flame, borax first swells up (intumesces), evolves steam, and then fuses to a perfectly colorless glass, which remains transparent on cooling.

Phosphorus-salt (microcosmic salt, NaO.NH,O.HO.PO,+8Aq) is sometimes. used instead of borax for producing colored glasses. When heated, this salt loses its ammonia and water, and is converted into the metaphosphate of soda (NaO.PO,), which fuses into a perfectly clear glass. Phosphorus-salt is used in small crystals.

Nitrate of potassa (nitre, KO.NO) is occasionally employed as an oxidizing agent; it should be kept in powder.

§ 64. It may be useful, in this place, to explain a few terms which are frequently used in describing the behavior of substances under the blowpipe.

INTUMESCENCE is the swelling up of the fused salt, in consequence of the rapid expulsion of its water of crystallization; borax affords a very good example of this (§ 63).

DECREPITATION has been already defined as the splitting up of the crystals of a salt, by the expansion of the mechanically-inclosed water.

DEFLAGRATION is the vivid combustion observed when powerful oxidizing agents are heated in contact with oxidizable substances (e. g. nitre upon charcoal), or when such agents are exposed to the inner blowpipe-flame.

DETONATION takes place when one or all of the substances acting upon each other are suddenly converted into the gaseous form, so as to give rise to a sharp

noise.

INCANDESCENCE is the production of a bright light, when solid infusible substances are introduced into the hottest part of the blowpipe-flame.

§ 65. Before proceeding to describe the manipulations with the blowpipe, it will not be out of place to say a few words with regard to the nature of the blowpipe-flame.

In a coal-gas flame (which is quite similar to the flames of lamps and candles), the carbon and hydrogen which constitute the fuel, enter into combination with oxygen (forming respectively carbonic acid and water), but not simultaneously. The affinity of hydrogen being greater than that of carbon, for oxygen, under the present conditions, the former undergoes combustion first, leaving the carbon free in the midst of the flame, by the heat of which it is raised to so high a

temperature, that it emits a white light; this carbon afterwards undergoes combustion on coming in contact with the external air.

Fig. 62.

In a common flame, we may perceive three cones, the innermost of which (Fig. 62 A) is a hollow space, filled with combustible gas, whilst the second or luminous cone, B, is that in which the hydrogen undergoes combustion and the carbon is raised to a white heat, to be subsequently burnt in the cone, c, which emits very little light.

-A

In examining the action of the blowpipe upon such a flame, it must be observed that the air projected into the latter does not proceed from the lungs, but is simply conveyed through the pas sages of the nose into the mouth, and thence ejected by a muscular effort of the cheeks, so that a stream of nearly pure atmospheric air (oxygen and nitrogen) is forced into the flame, to which enough oxygen is thus supplied to enable the carbon and hydrogen to burn simultaneously, in consequence of which no carbon separates, and the flame emits no white light. The blowpipe-flame (Fig. 63), like that of the

Fig. 63.

ordinary candle or gas-jet, consists of three parts; the inner hollow cone, where the cold air first passes into the flame; the inner cone of partial combustion; and the outermost cone, where the combustion is completed. In that part of the inner blue cone which is nearest to the blowpipe-jet, there must be an excess of oxygen, and in this place the combustion is perfect; beyond this, there is a point where neither the oxygen nor the combustible gas is in excess, and this is consequently the hottest part of the blowpipe-flame; this point is found near the extremity of the blue flame; a little within the point of the blue flame (unless a very considerable amount of air be forced into it from a large blowpipe-jet), there will be an excess of combustible gas (carbon and hydrogen), which, at the high temperature to which it is here raised, is capable of abstracting oxygen from most metallic oxides, thust reducing them either to metals or to a lower state of oxidation. This part of the flame is termed the reducing or deoxidizing flame.

When the heated gas has passed the point of the blue flame, it is oxidized at the expense of the surrounding air, and gives rise to the very slightly luminous cone of complete combustion. Around this cone, then, there is an excess of oxygen, and if any substance be introduced into it which has any considerable affinity for this element, it will be at once oxidized; hence this outer cone has been named the oxidizing flame. The capabilities of the different blowpipeflames may be readily tested by introducing a little oxide of lead, in a small iron spoon, into the inner flame, which will at once reduce it to the metallic state, and the metal thus obtained may be reoxidized by transferring it to the outer flame.

In order to obtain a well-defined blowpipe-flame with the gas-jet above described, the aperture of the blowpipe-jet should be placed just within the flame, immediately above the edge of the aperture from which the gas issues; if a reducing flame is required, a blowpipe with a small jet should be used, a larger orifice being better suited for oxidation. A broad scattered flame, which is very useful for heating crucibles, drawing glass, &c., is obtained when the blowpipejet is withdrawn to the distance of or inch from the margin of the gasflame. A good blowpipe-flame should be free from white-light, and the two cones should be very well defined. The stream of air must not be intermitted, the operator acquiring, by continued practice, the habit of breathing through the nose without relaxing the muscles of the mouth and cheeks.

In examining the action of the blowpipe-flame upon test-specimens placed on charcoal supports, it is generally desirable to ascertain if any substance is reduced to the metallic state, and whether the surrounding portion of charcoal is covered with an incrustation of oxide resulting from the reoxidation of the metallic vapor in passing through the outer flame. To determine these points, a piece of charcoal having been selected and prepared according to the directions given above, a small shallow cavity (of about inch in diameter) is scooped with a penknife at one end of the smooth surface, within about half an inch of the edge, and in this the test-specimen is placed, and covered with the reagent to be employed; the charcoal is now held in the blowpipe-flame in such a manner that the reducing (inner) flame may be directed into the cavity, and the oxidizing (outer) flame allowed to flow over the surface of the charcoal, upon which it should extend itself in the form of a cone, within the limits of which we afterwards look for the incrustation; the stream of air must not be violent at first, or the substance will be blown away, but should be increased as the operation proceeds. The specimen to be examined should be first powdered.

In some cases, especially in the analysis of minerals, it is necessary to roast the specimen, by exposing it for some time to the outer flame, to oxidize and remove sulphur, &c., before attempting to reduce it; this should always be attended to in the case of metallic sulphides, the carbonate of soda, or other reagent, not being added till the roasting is completed, which will be the case when the odor of sulphurous acid is no longer perceptible; the test-specimen should be turned about once or twice during the operation, to expose fresh surfaces to the oxidizing action.

The reduced metal is generally seen either in one pretty large globule or in a multitude of smaller particles, which may generally be induced to unite by judiciously directing the blowpipe-flame upon them; cyanide of potassium is very useful in favoring the union of such globules, since it becomes very liquid at a comparatively low temperature; the globules are best seen when the mass is redhot. A globule having been obtained, it may be desirable to ascertain whether it is malleable or brittle; for this purpose, it is allowed to cool perfectly, and carefully removed with a pair of tweezers; having been placed upon the bottom of a strong inverted mortar, it is now struck sharply with the pestle, when, if brittle, it of course falls to powder (as in the case of antimony), if semi-malleable, it flattens out, at the same time breaking into several pieces (as with bismuth), and, if fully malleable, flattens out without breaking (like lead).

In some cases, and particularly where large quantities of earthy matters are present, small portions of reduced metal are disseminated throughout the mass after exposure to the inner blowpipe-flame, but will not join into globules; to detect these, the test-specimen, together with the surrounding portions of charcoal, may be scraped into an agate mortar, and reduced to a very fine powder; if this is submitted to repeated levigation, all the metallic particles will be left behind. When looking for an incrustation upon the surface of the charcoal, it must be remembered that the latter is generally covered with a thin film of bluish-white ash after having been exposed to the oxidizing flame.

§ 66. A few metallic oxides are recognized by the color of their compounds with oxide of cobalt; in order to subject them to this test, the specimen is very strongly heated, on charcoal, in the hottest part of the blowpipe flame; it is then removed from the flame, moistened with a drop or two of the solution of nitrate of cobalt, and again very strongly heated; the color of the resulting compound should be observed when it has cooled, and by daylight.

The formation of colored glasses by dissolving certain metallic oxides in fused borax or phosphorus-salt, is often had recourse to as a means of recognizing

This may be prevented by slightly moistening the substance.

them, and as such glasses usually present different appearances in both flames, every inference obtained in this way is supported by two indications. The colorless glass to be employed must first be prepared, and the oxide added to this by degrees, till a distinct color is produced; a piece of the thicker variety of platinum wire above mentioned is selected, and its extremity (previously well washed) bent round in the form of a loop, which should be somewhat smaller than the section of the reducing flame; this loop is now heated to redness in the flame, and plunged into the flux to be employed (borax or phosphorus-salt), when a sufficient quantity will adhere to it to form a small bead in the loop when fused in the blowpipe flame;1 in fusing the bead, the wire must be dexterously turned with the hand, to prevent the fused flux from dropping off, and the fusion is discontinued when the effervescence and boiling have ceased. The bead thus obtained, which must be perfectly transparent and colorless, and not larger than the section of the reducing flame, is now again heated to redness, and a very small particle of the substance to be examined made to adhere to it; the glass is fused in the outer flame (near the point) for some time, the bead allowed to cool, in order that its color may be observed, a fresh quantity of the substance added as before, and this operation repeated (always fusing in the outer flame) until either a distinct color is obtained, or a considerable amount of the substance has been added without affecting the color of the glass. The color (by transmitted daylight) of the hot and cold bead, should be carefully observed, and the latter then exposed to the reducing flame for some seconds, the color of the glass being afterwards again noted. It is obvious that the smaller the bead, consistently with distinct perception, the better; and that very little of the coloring matter should be added at once, since some metallic oxides impart such intense colors as, when in considerable quantity, to cause the bead to appear black.

Other metallic oxides are known by their imparting particular tints to the outer blowpipe-flame, in consequence of the reduction and volatilization of the metal in the inner flame, and its subsequent burning with the color in question on arriving in the oxidizing cone. In order to test substances in this manner, a very small loop is made at the extremity of the thinner platinum wire, and well washed with distilled water; it is now introduced into the inner flame, and if it impart any tint to the outer flame, it is removed, after a few seconds, and again washed; this process must be repeated until the wire ceases to tinge the flame ;* the loop is now moistened with pure water, and a little of the powder under examination is taken upon it and introduced into the point of the inner flame, where it should be held for two or three minutes before we conclude that it imparts no tint to the outer flame.

When substances are heated on platinum foil, in iron spoons, or glass reductiontubes, it is usually with the intention of raising them to a high temperature without subjecting them to any chemical action of the flame; a broad-scattered blowpipe-flame is generally used for this purpose, and is directed on to the bottom of the support.

GLASSBLOWING.

§ 67. It is an important qualification of the practical chemist to be able to fashion the simpler kinds of apparatus without the aid of the glassblower, since

1 When carbonate of soda is employed as a flux, the loop should be wetted, in order that the carbonate may adhere, since it does not readily attach itself to the redhot wire. When phosphorus-salt is used, it is advisable to give two turns to the wire in making the loop, and to allow the loops thus made to cross, so as to form a sort of grating on which the very fusible glass may be retained.

2 This (yellow) tint is generally imparted to the flame by the soda derived from the fingers of the operator, who should not touch the loop when once cleansed.

it not only effects a considerable saving of expense, but enables him to give to his instruments that form which suits his own taste; we shall here give a few brief directions, which may be useful in guiding the practice of the novice in this department.

Considerable difficulty is experienced in drawing and blowing glass before the mouth-blowpipe, and hence certain blowpipes are provided especially for this purpose. The chief of these are known as the table-blowpipe, Herapath's blowpipe, and the spirit-blowpipe, or, as it is commonly termed in the laboratory, the

roarer.

The table-blowpipe is simply a table furnished with a lamp and blowpipe-jet, to which air is supplied from a pair of double-action bellows, worked by a treadle and weights beneath the table; the lamp is generally supplied with oil, and should have a good broad wick, which is kept well trimmed.

The Herapath's blowpipe consists of two brass tubes, one within the other, so contrived that when screwed on to the gas-pipe, a jet of gas may issue from the outer tube, and a stream of air may be forced from the mouth through a tube of vulcanized Indian-rubber, into the inner brass tube, which is terminated by a blowpipe-jet; the air is thus projected into the very centre of the gas-flame, and, the inner tube being made to slide up and down in the outer, the jet may be approached to, or withdrawn from the flame, so as to furnish a blowpipe-flame of any dimensions.

The spirit-blowpipe-lamp (Rose's lamp) is a sort of brass pot with double walls, into the interval between which a small brass tube penetrates nearly to the top, and enters the pot at the bottom, an inch above which it terminates in a pretty large blowpipe-jet; the space between the walls is about three-parts filled (through an aperture made for the purpose, and stopped either with the handle of the pot or with a good cork) with wood-naphtha, a small quantity of which is poured into the inside of the pot, so as to reach within about a quarter of an inch of the blowpipe aperture; if this be kindled, its flame heats the naphtha between the two walls, and converts it into vapor, which rushes out with a roaring noise through the jet, where it takes fire, thus producing a broad column of flame very well adapted for heating crucibles, drawing thick glass tubes, &c. It is scarcely necessary to observe that the naphtha poured into the space between the walls must be perfectly clear, for if any fragments of cork, &c. get into the blowpipeaperture, the lamp may burst with considerable violence, and hence the dangerous reputation which these lamps have acquired; with a little care, however, they may be used with perfect safety, and are very valuable instruments, especially in laboratories where gas cannot be procured. If the jet of vapor should suddenly cease, the lamp must be immediately extinguished with the cover provided for this purpose.

The ordinary cases of working in glass which come under our notice in the laboratory, and which have not yet been referred to, are, the simple closure of tubes so as to preserve a uniform thickness, the sealing of tubes required to stand considerable internal pressure, the expansion into bulbs, the drawing out of tubes to a long open point, and the manufacture of the combustion-tubes used in organic analyses.1

It is not difficult to close a tube so as to preserve a uniform thickness; a piece of tube is selected, about three inches longer than the required closed tube, and, having been first heated in the common flame (which precaution must be attended

In manipulating with glass before the blowpipe, it should be observed, that the English glass is very liable to blacken, from the reduction of lead, and should therefore be heated only in the oxidizing flame, whilst the German glass may be exposed to the hottest part of the flame; in fact, for most purposes, the German glass is much superior to the English, and is always used when the tubes are required to bear a high temperature.