adjusting the level of the mercury until it is once more at the same height in each limb, it will be found that the mercury in the eudiometer is now standing at the second band; that is to say, the three volumes of gas originally present have now become two volumes of steam. This condensation is expressed in the molecular equation

O2+2H2=2H2O.

The Gravimetric Composition of Water.-Having learned the composition of water by volume, and knowing also that the relative weights of equal volumes of oxygen and hydrogen are as 15.88: 1, the composition by weight can readily be calculated, thus

17.88 parts by weight of water are composed of 2.00 parts by

weight of hydrogen and 15.88 parts of oxygen, or, expressed centesimally, we have

The composition of water by weight has been experimentally determined with great care by a number of chemists.

The apparatus shown in Fig. 45 represents the method employed by Dumas (1843). When copper oxide is heated in a stream of

hydrogen, the copper oxide is deprived of its oxygen, which unites with the hydrogen to form water—

CuO+H=Cu+H2O.

Dumas' method is based upon this reaction. A weighed quantity of perfectly dry copper oxide was heated in the bulb A, in a current of hydrogen generated from zinc and sulphuric acid in the bottle H, and rendered absolutely pure and dry by its passage through a series of tubes containing absorbents. The water formed by the union of the hydrogen with the oxygen of the copper oxide was collected in the second bulb, B, previously weighed ; and the uncondensed aqueous vapour which was carried forward in the stream of hydrogen was arrested in the weighed tubes which follow. The increase in weight of the bulb B and the weighed tubes gave the total weight of water produced; while the loss of weight suffered by the copper oxide gave the weight of oxygen contained in that The difference between these two weights is the weight of the hydrogen that entered into combination with the oxygen. As a mean of many experiments it was found that in the formation of 236.36 grammes of water the oxygen given up by the copper oxide was 210.06 grammes.

water.

The ratio of hydrogen to oxygen is therefore as 2:15.88.

Hydrogen prepared from zinc and sulphuric acid is liable to contain traces of (1.) Sulphuretted hydrogen. This is absorbed in the first tube containing broken glass moistened with a solution of lead nitrate.

(2.) Arsenuretted hydrogen J absorbed in the second tube, filled with glass (3.) Phosphoretted hydrogen moistened with silver sulphate.

(4.) Sulphur dioxide

(5.) Carbon dioxide

(absorbed in the third tube, containing in one limb pumice moistened with a solution of potassium hydroxide, and in the other fragments of solid potassium hydroxide.

Tubes 4, 5, 6, and 7, containing solid potassium hydroxide and phosphorus pentoxide (the two latter being placed in a freezing-mixture), are for the purpose of withdrawing every trace of aqueous vapour. Tube 8 was weighed before

and after the experiment in order to test the absolute dryness of the hydrogen that entered the bulb. In order to get rid of dissolved air, the dilute sulphuric acid used was previously boiled. Tubes 9, 10, 11 were weighed both before and after the experiment; while tube 12, which was not weighed, was placed at the end to prevent any absorption of atmospheric moisture by the weighed tubes. Since the time of Dumas this subject has been reinvestigated by other experimenters, who have introduced various modifications into the process; thus, with a view to finding the weight of hydrogen directly and of eliminating many of the possible sources of error arising from the presence of impurities in

the hydrogen, the hydrogen has been absorbed by palladium. The metal so charged with hydrogen can be weighed before and after the experiment, and the actual weight of hydrogen used directly ascertained.

Most recently the matter has been investigated by Scott and Rayleigh, and the results obtained show only the slightest departure from the numbers obtained by Dumas.

Properties of Water.-Pure water is a tasteless and odourless liquid. When seen in moderate quantities it appears to be colourless, but when viewed through a stratum of considerable thickness it presents a beautiful greenish-blue colour. This colour may be

seen by filling a horizontal tube about 15 feet long with the purest water, and passing a strong beam of light through it. It may also be perceived by directing a ray of light through a tall cylinder of water in the manner shown in the figure, and causing it to be reflected up through the water from the surface of a layer of mercury at the bottom; the immerging ray, being then reflected upon a screen, shows the characteristic colour of the water. By intercepting the ray by a hand mirror at A, the white light can be thrown upon the screen as a contrast to the greenish-blue tint.

Aitkin has recently shown that the presence of extremely finely divided suspended matters in water will give to the liquid the appearance of a blue colour. Thus, in tanks where water is being softened by the addition of milk of lime, after the bulk of the precipitated chalk has settled, and only the finest particles still remain suspended in the liquid, it is often noticed that the water appears to have a rich blue colour. The wonderful blue colour of the waters of many of the Swiss lakes is probably due in part to this optical phenomenon as well as to the intrinsic colour of the water. When a mass of pure snow, such as falls in high mountainous regions, is broken open in such a way that the light is reflected from side to side of the small crevice, the true greenish-blue colour of the water is very manifest.

Water is compressible to only a very slight extent; thus, under an additional pressure of one atmosphere, 1000 volumes of water become 999.95 volumes.

Small as this compressibility is, it exerts an important influence upon the distribution of land and water upon the earth. It has been calculated that owing to this compression, where the ocean has a depth of six miles, its surface is lower by 620 feet than it would be if water were absolutely non-compressible; and, calculated from the average depth of the sea, its average level is depressed 116 feet. The effect of this depression of the sea-level is that 2,000,000 square miles of land are now uncovered which would otherwise be submerged beneath the ocean.

Water is an extremely bad conductor of heat. A quantity of water contained in a tube held obliquely may be boiled by the application of heat to the upper layers without appreciably affecting the temperature of the water at the bottom; a fragment of weighted ice sunk to the bottom will remain for a long time unmelted, while the water a few inches above it is vigorously boiling. This low conductivity for heat is shared in common by all liquids that are not metallic. Indeed, Guthrie has shown, that water conducts heat

better than any other substance which is liquid at the ordinary temperature, with the exception of mercury.

Steam. Under a pressure of 760 mm., water boils at 100° (see p. 128), and is converted into a colourless and invisible gas or vapour. The visible effect that is observed when steam is allowed to issue into the atmosphere is due to the condensation of the steam in the form of minute drops of water. What is popularly called steam is in reality, therefore, not steam, but an aggregation of small particles of liquid water. The invisibility of steam is readily demonstrated by boiling a small quantity of water in a capacious flask; as the steam issues from the neck it condenses in contact with the cool air and presents the familiar appearance, but within the flask it will be perfectly transparent and invisible.

Ice. At a temperature of o° water solidifies to a transparent crystalline mass. In the act of solidification the water expands by nearlyth of its volume, 10 volumes of water become 10.908 volumes of ice solid water, therefore, is specifically lighter than liquid water, and floats upon its surface. Water in this respect is anomalous, for in the case of most other substances the solid form is denser than the liquid. The disruptive force exerted by water at the moment of freezing is the cause of the bursting of pipes and other vessels containing water during winter; and it is also an important factor in the economy of nature in the disintegration of rocks and of soil. Under certain conditions water may be cooled many degrees below o° without solidification taking place. Thus, if a small quantity of water contained in a vacuous tube be care fully cooled without being subjected to vibration, its temperature may be lowered to 15° without it solidifying; a slight shock, however, at once causes it to pass into the solid state, when its temperature instantly rises to o° (see p. 136). Although the exact temperature at which water freezes is liable to uncertainty from this cause, the point at which ice melts is, under ordinary cir cumstances, constant, viz., o°. Under increased pressure ice will melt at temperatures below o°; thus, Mousson found that, under a pressure of 13,000 atmospheres, ice melted at -18°. The melting-point of ice is lowered by about 0.0074° by each additional atmosphere of pressure (see p. 137).

Between the temperatures of +4° and 100°, water follows the ordinary laws that govern the expansion and contraction of liquids due to change of temperature; if water be cooled from 100', it gradually contracts until the temperature reaches 4°. Between