and had contracted to at least one half of its former volume. For these reasons, I had recourse to the balls proposed by Professor Dæbereiner. I mixed together spongy platinum and pipe-clay, in different proportions, and, by the addition of a little water, made a set of balls, about the size of a pea. As the activity of spongy platinum is in some way connected with its porous texture, I endeavoured to bring the balls as nearly into a similar condition as possible. With this intention, a little powdered sal-ammoniac was mixed with the platinum and pipe-clay, and when the balls had become dry, they were cautiously ignited at the flame of a spirit-lamp. The sal-ammoniac escaping from all parts of the ball, gave it a degree of porosity which was very favourable to its action. In this way twelve different balls were formed, the exact composition of which I subjoin.

Siliceous earth was added to the four last, since the pipeclay alone made the mass too tenacious. I found that any of the four first balls, the weakest of which contained 3ds its weight of platinum, might be very well substituted for the pure spongy metal. Freshly ignited, and allowed to cool, they became red when a jet of hydrogen was thrown upon them, and the gas itself then took fire. Put in contact with a mixture of two measures of hydrogen to one of oxygen, they became red, and an explosion followed, though not quite so rapidly as when spongy platinum was used. For exam ple, when I put the ball No. 2, in 1.5 cubic inch of the explosive mixture, a dense dew formed on the mercury; rapid diminution succeeded, and in one or two seconds a dull light appeared, which passed rapidly to vivid redness; but by this time all the gas was consumed. With ball No. 3., the same results were obtained. Had I operated in these cases on a larger quantity of gas, there must have been an explosion.

To prove this, No. 4 was put in 4 cubic inches of the explosive mixture; absorption with dense watery vapour first appeared in two or three seconds the ball became dull red, and passed rapidly, but during an appreciable space of time, to vivid redness, at which moment an explosion happened. The activity of these balls is greatly increased by heating them gently: I have known No. 4 in that case inflame an explosive mixture of less than half a cubic inch.

The balls No. 5, 6, 7, and 8, though not so powerful as the preceding, acted, nevertheless, with considerable energy. I put No. 8., for example, in four cubical inches of the explosive mixture, expecting only a silent action; to my surprise it became red, and caused an explosion. Nor would I venture to put Nos. 9, 10, and 11. in contact with much of the explosive mixture, for they act with an energy by no means to be expected from the small quantity of platinum present. No. 12. was put into a half cubical inch of the explosive mixture: a very sluggish action ensued, but in the space of five minutes the gases were completely condensed.

These experiments will convey a distinct idea of the extraordinary power of platinum, in causing the combination of oxygen and hydrogen gases, even when the quantity of metal present does not exceed one quarter of a grain, and that made up into a ball with fourteen times its weight of foreign matter. It is necessary to mention, that the gases I used in these experiments were of great purity, quite dry, and mixed in the exact proportion to form water. The oxygen was carefully prepared from dry chlorate of potash, and received over mercury. I prepared the hydrogen from zinc and dilute sulphuric acid, and collected it in a gas-holder over water. Successive portions of it were drawn off when required, and dried over mercury by fused potash. In the more delicate experiments I always ascertained its purity, and made due correction for the air it contained. If the gases are not mixed in the proper proportion to form water, a silent but rapid combination is produced. I have never had any explosion when one volume of hydrogen was mixed with four of oxygen, nor when one of oxygen was added to four of hydrogen. Atmospheric air and hydrogen may be mixed in any proportion without danger of explosion.

The ball itself suffers no chemical change; at least I have used the same one in fifty different experiments, without observing the least diminution of its energy. Its activity does indeed diminish after repeated use, if not ignited during the

VOL. II. NO. 3.

30

experiments, but is completely restored by heating the ball to redness. The method I pursued in all the succeeding experiments, was to throw the flame of an oil or spirit lamp upon the ball, by means of a blowpipe, so as to keep it intensely hot for the space of one minute, after which it was allowed to sink to the temperature of the room before being used. Sometimes, however, it was necessary to plunge the ball through mercury while strongly heated, and this occasionally destroyed its action most completely; though its energy was restored by keeping it for a few minutes in the strongest heat of the blowpipe. I have sometimes observed the energy of an active ball so completely destroyed by passing it through mercury, that it did not produce the slightest effect in a highly explosive mixture.

Eleven volumes of atmospheric air were mixed with one volume of hydrogen. An electric spark was passed through one portion of the mixture, and the ball No. 3. brought in contact with another. Detonation occurred in the first case, and in the second, diminution of the volume of gas, with deposition of water. A strong charge of electricity was passed through a mixture, composed of 13 air to 1 hydrogen, without any detonation; whereas the balls, Nos. 3 and 5., afterwards caused immediate diminution. Repeated strong charges from a Leyden jar were sent through a mixture of 15 air to 1 hydrogen, without any detonation, while balls Nos. 3 and 5. acted instantly. A mixture, composed of 19 air to 1 hydrogen, was not at all influenced by electricity; but balls Nos. 3 and 7. caused formation of water, as did No. 8. when warm. The ball No. 3 was put into a mixture, composed of 1 volume of hydrogen to 4 of air: an abundant deposition of water followed; and in one minute after the entrance of the ball, the residual gas was transferred into a eudiometer and electrified, but without any detonation. The very same occurred in a mixture of 5 air to 1 hydrogen. Ball No. 3. had a distinct action in mixtures of 30 and 40 air to 1 of hydrogen.

It is then apparent from these experiments, that the platinum balls caused the formation of water, when a strong elec trical charge did not do so. It appears, too, that the balls acted so powerfully in a mixture of air and hydrogen, that electricity, afterwards employed, had no effect. The platinum is likewise shown to act with great velocity; for, after the short space of one minute, so large a quantity of hydrogen had disappeared, that the residual gas was not explosive. I have, within a few days, seen a short article by Professor

Gmelin of Tubingen, published in a recent number of the Anno les de Chimie et de Physique, in which he states, that a mixture of air and hydrogen, after it had been acted on by platinum, still contained so much hydrogen as to detonate from the electric spark. He concludes, therefore, and justly, were his experiments correct, that platinum cannot be employed in eudiometry. He does not mention how he operated, and therefore I cannot perceive the source of his error.

I now proceeded to inquire, whether the exact quantity of hydrogen was indicated by the action of platinum.

In a mixture of 5 air to I hydrogen the diminution occasioned by an active ball indicated within a fraction the quantity of hydrogen present; and in a mixture of 9 air to 1 hydrogen the quantity of the latter was again correctly indicated.

These experiments were performed once and again with similar results.

In mixtures of 13 air to 1 hydrogen, 15 to 1, 17 to 1, and 19 to 1, the quantity of hydrogen was very accurately indicated, and the brilliancy of the mercury round the ball was in each case sullied by the formation of dew.

I mixed 5 volumes of hydrogen with 500 of air, or in the exact proportion of 1: 100. The ball No. 2., gently warmed, was put into 180 volumes of this mixture, which, of course, contained 1.8 of hydrogen. In half an hour there was diminution of 3 volumes, which indicates the presence of 2 hydrogen.

The ball No. 2., gently warmed, was put into 186 volumes of the same mixture, containing of course 1.86 hydrogen. In three quarters of an hour there was a diminution of 2.8, which indicates the presence of exactly 1.86 hydrogen.

No 2, gently warmed, was put into 132 volumes of the same mixture, containing 1.3 volumes of hydrogen. In three quarters of an hour there was a diminution of 2.5, which indicates 1.6 hydrogen.

These experiments were made in a tube of 0.4 inch inter nal diameter, divided into deci-millilitres, with such accuracy, that I could measure to within one quarter of a deci-millilitre, or the 650th of a cubic inch. In a tube of such small diameter, a much longer time is requisite for perfect action than in a more capacious vessel, since in a jar of 1.3 inch diameter, the effect is quite complete in five minutes. In all such experiments as the preceding, where great nicety of observation is necessary, the thermometer was always observed, and due correction made for the least difference. The duration of

the experiments was not cousiderable enough to render ba rometrical observations requisite, more particularly as I drew no inference from single experiments. I think it then fair to conclude, that hydrogen may be detected in atmospheric air by platinum, when it does not exceed th of the whole volume, it follows too, if my experiments are not in fault, and I took all possible care to avoid error, that even the quantity of hydrogen under such circumstances is exactly indi

cated.

A mixture of 6 volumes of oxygen to 1 hydrogen was fired by electricity, and a platinum ball was brought in contact with the residue, but occasioned no diminution.

A mixture of 8 volumes of oxygen to 1 hydrogen was fired by electricity. The residue was not at all diminished by platinum.

A mixture of 9 volumes of oxygen to 1 of hydrogen was fired by electricity. The diminution was quite to the sufficient extent, nor had platinum any effect upon the residue.

In a mixture of 11 parts of oxygen with 1 of hydrogen, electricity occasioned only a feeble detonation, attended with partial decrease of volume, but a platinum ball caused formation of water in the residuc.

A mixture of 13 oxygen to 1 hydrogen was acted on by electricity and platinum, just as in the preceding case.

A mixture of 15 volumes of oxygen and 1 hydrogen was electrified, without either detonation or decrease in volume. The ball No. 2. then caused a diminution, which indicated the exact quantity of hydrogen.

These results with respect to the influence of electricity correspond exactly with the experiments of MM. de Humboldt and Gay Lussac, detailed in the Journal de Physique for 1805.

I made various mixtures of oxygen and hydrogen, in which the latter wasth, 7th,th,th, th, of the whole mass, and in all these cases ball No. 2. caused a diminution, which indicated the quantity of hydrogen with much accuracy. In the two last experiments the ball was slightly warmed.

7 volumes of hydrogen were mixed with 700 volumes of oxygen, being the exact proportion of 1: 100. The ball No. 2. gently warm, was put into 138 volumes of this mixture. The process was conducted in a capacious jar; the gases were carefully transferred into the graduated tube at the end of 10 minutes, and the diminution was 3 volumes. Hence 2 volumes of hydrogen were indicated, while 1.38 were present.