ing appeared to be slightly redder, especially on the parts of the salt touching the sides of the glass. No further change was noted until 245°, when the salt appeared dull and paler. The redness had not increased or spread to any portion not in contact with the glass.

At 246°, a sublimate appeared for the first time on the top of the crystals. The red patches disappeared (possibly owing to their being coated over with the white substance), and the whole of the salt lost its blue tint and became of a warm grey colour. Liquid continued to distil over, and no other change was noted until 260°, when some very small, glittering, crystalline particles appeared which disappeared again

at 273°. The top of the upper compartment of the apparatus had in the meantime become coated with sublimate, but this was washed down from time to time by condensed liquid; finally, however, the distillation of liquid ceased, and the top of the apparatus remained coated with sublimate. The whole of the substance in the lower bulb appeared of a chocolate-brown colour. After allowing the apparatus to cool, it was divided into three parts by sealing off at the side tube and in the middle. On heating the tube to seal off, the softened glass blew out, showing that a volume of fixed gas in excess of the capacity of the tube had been formed.

To investigate these products, and in particular to collect the gases formed and to prevent contact between the liquids which are condensed and the sublimate, a different form of apparatus was needed. The form which was finally adopted may be seen by reference to Fig. 3 (p. 1391).

Description of Apparatus shown in Fig. 3.

The tube in which the salt is heated is divided into two compartments, A and U, by a constriction, in which a small bulb, H, is placed, to prevent the salt falling down into the lower or sublimate chamber, U. The sublimate chamber is drawn out below into a narrow tube, bent as shown, J, and terminating in a capillary point, V. The tube is fixed by means of a rubber cork, W, into a wider tube, B. The tube B is filled with oil, and is heated by means of a ring-burner consisting of a series of hard, glass tubes, Z, seven in number, of the shape shown in the figure, which allows of their being set close to, or further from, the tube B. The burner tubes fit into a wide glass chamber, I, which is supplied with gas by an inlet tube, T. The gas, after a preliminary regulation by a gas-governor, passes through the regulator, F, before entering the inlet tube T. This regulator, F, allows of the temperature being kept constant for a long time at any desired point. For this purpose, the temperature is set approximately by taking out or putting in oil at X, and the final adjustment is made by raising or lowering the fine tube A'; a pin hole, B' in A', acts as a by-pass when the expansion of the oil in B raises the mercury in the regulator up to the end of the tube A', and so prevents the burners from being put out. By this vertical arrangement of the tube in the oil-bath, different zones of temperature are obtained, the object of which is to allow the sublimate to condense at a point where the temperature is still sufficiently high to prevent the acetic acid condensing also. Below the level of the burner-tips, the temperature of the sublimate chamber falls off rapidly, and the acetic acid and water condense at the bottom, apart from the sublimate, and flow down into the drawn out tube and thence into the washing bulbs. The tube B is supported by a clamp (not shown in the figure) at the narrow portion of the neck, below the two side tubes X and Y.

D and D' are two semicircular sheets of tin plate, partly shutting in the space enclosed by the wide, outer glass cylinder, C, which keeps up the temperature and prevents a draught from disturbing the flames. A thermometer, G, registers the temperature of the bath. E is a plate of mica to act as a shield.

The drawn out tube, J, is fitted by a gas-tight mercury joint (shown in section in Fig 4, about its actual size), so that the capillary tube, V, delivers the liquid freely from its point into the tube of the washing

bulbs, R. The washing bulbs, R, are six in number, and are so arranged that when held by one end there is a continual fall from the first to the last bulb, so that they may be conveniently washed out. The three bottom bends form three points of a triangle, and allow of the bulbs standing on the balance pan or any flat surface without support.

The bulbs are charged with water (3 or 4 c.c. in each pair), rather less being placed in the pair of bulbs nearest the capillary, as the main quantity of acetic acid and water from the decomposition of the salt is arrested there. By keeping the three portions of water separate, the washing is made much more complete, since only that portion of the acetic acid vapour which escapes from the first pair of bulbs after the water in it is nearly saturated passes into the second pair, and only the residue of this into the third pair. The result is that practically the whole of the acetic acid is stopped, whereas this is not the case if the three portions of liquid are in communication. To prevent the distillation of liquid which is apt to occur when the vacuum is good and the atmosphere moderately warm, the bulbs are enclosed in a bath of ice and water which is wrapped round in cotton wool, P. As the contact of the glass of the bulbs with the zinc of the bath, O, has sometimes caused fracture, some folds of flannel, Q, are placed on the bottom of the bath. The bulbs are connected by another mercury joint, L, with a drying tube, S, containing pumice moistened with sulphuric acid in the upper part and sulphuric acid in the lower. The drying tube, S, is connected by another mercury joint, M, with the Sprengel pump by which the gases are drawn off and collected. The washing bulbs and drying tube are of a form convenient for weighing.

The Action of Heat on Copper Acetate.

The first result of the action of heat on the salt is a forma

tion of liquid. With the apparatus described, it is somewhat difficult to determine exactly the temperature at which this occurs, but in one experiment it was observed at 115°. When the temperature has reached 150-160°, the surface of the salt appears coppered in parts. This coppering occurs also when the salt is heated in carbon dioxide or hydrogen. It appears on the parts of the salt immediately touching the glass, and spreads somewhat irregularly over the surface of the salt, but never becomes uniform over the entire surface. The coppery deposit appears a long time before any trace of sublimate is seen, and occurs whether the salt is powdered or in crystals; in the latter case, however, the smooth surfaces of the crystals give the coppering a more metallic appearance.

The next noteworthy change is the deposition of sublimate and the

evolution of gas. These two changes occur together, and the temperature at which they occur is about 230°. In some experiments, the sublimate has been observed at 215°, and in others lower than this. The formation of sublimate is, however, so extremely slow that a complete experiment at these lower temperatures is impracticable. In one instance, the experiment was continued for a week and the temperature kept below 212° the whole time; it was afterwards found that the decomposition was even then incomplete, and on reheating the residue at a higher temperature a small additional sublimation occurred. For this reason, in the later experiments, the temperature was allowed to rise fairly rapidly to 230°, at which temperature the decomposition can be completed in a reasonable time. When about 8 grams of salt are used and the gas is pumped out at a convenient rate, the experiment can be completed in about 9 hours after the appearance of the sublimate. If the pressure of the gases is allowed to rise, the decomposition is retarded.

The sublimate always appears at a point in the sublimate chamber just below the level of the tips of the burners, and spreads downwards. At this part, there is a very steep gradient of temperature. In one experiment, the temperatures of the extreme limits within which sublimate was being deposited were approximately determined by raising and lowering the thermometer. At the lower limit of the sublimate, the temperature was 103°, and at the upper limit 170°; the width of this zone was only 8 or 10 mm.

The first portions of the sublimate are almost invariably crystalline; as sublimation becomes more rapid, its appearance changes: it becomes granular and nodular, growing out from the sides of the tube until a cake is formed at the top by the sublimate meeting in the centre. After this cake is formed, the sublimate increases upwards slightly, but only for a very short distance. After a time, the bottom of the deposit gradually becomes moist through a condensation of acetic acid, and a green band is formed which spreads slowly upwards. Finally, if the experiment is prolonged, the top of the sublimate becomes of a darker colour. Condensation of liquid and evolution of gas go on more and more slowly, and finally the decomposition comes to an end, and all that remains of the salt is a reddish-brown residue in the top chamber. The drawn out tube, J (Fig. 3), is then sealed off, the burner extinguished, and the apparatus allowed to cool. After the tube has been removed from the oil-bath, the chambers A and U (Fig. 3) are separated by sealing off.

The following tables give the results of the experiments so far as the amounts of the different products obtained are concerned; afterwards the different products are dealt with separately :