tinctly indicated in fig. 1, b, and that the action of this differs in no respect from that of P except in its longer period of revo lution, and that this period of revolution is the same with that of the signal wheel, it will be obvious that if several supplementary openings be made in the signal wheel, as for instance two, t and u, in two teeth diametrically opposite to each other, and two others, v and , in teeth adjacent to one of these on each side of it, the flashes of light through these openings will be seen by the eye at E to occur at points t', u', v', w', fig. 3, distributed around the circumference of a circle concentric with SDS'G, and at angular intervals from each other identical with those between the corresponding openings in the signal wheel. It will be further obvious that the observed angle of position of the diameter u't' will depend on the angle of position of the wheel K relatively to that of the signal wheel. The diameter u't' becomes, then, by aid of the more exact indication of S' DS', an index by which we know the required angle of position at which the wheel K arrives simultaneously with the arrival of the signal wheel S at a given point of reference. As the most convenient mode of procedure in practice, the observer at B may operate the satellite wheel until the index diameter u' t' is brouglit to a position, for instance the vertical one in the figure, made to denote zero, and the position micrometer for taking the angle of position of S'D S' may be graduated to thousandths and ten thousandths of a second.

Instead of the above described arrangement there is a modification of it which I am disposed to prefer, the type of which is a pair of telescopes at station B, placed side by side so that one may contain the rapidly revolving prism and the other the more slowly revolving one, each prism being in this case unintermitting in its action, and the supplementary openings of the signal wheel being replaced by the filling up of a single one of the primary openings. The omission of the flash of light from this one would be observable through the slow prism and give the required indication, while it would not probably injure in any material degree the distinctness of the are of light seen through the fast prism. Instead of a pair of complete telescopes, the equivalent of a pair of eye-pieces with a sliding object-glass to alternate between them at pleasure, would answer the same purpose. In this arrangement no rectification of the prisms by the observer would be necessary, it being always possible to observe the total deviation. This would be a great advantage on a line of very numerous stations, in which case it would, on the first described plan, be a somewhat critical matter to bring all the instruments on the line into the required correspondence for simultaneous observation.

As before intimated, it would be possible to employ but a single clock on the whole line of stations, but as this would require signal observations for every time observation at any other than the clock station, it would be more convenient to employ a clock at every astronomical station.

The question of the feasibility of the process described in this paper will depend primarily on the practicability of securing, with telescopes of moderate aperture, a sufficiency of light for such distances as from fifty to eighty miles, and next on the attainment of sufficient precision of rate in the uniform motion employed. I do not anticipate serious difficulty in either of these things. For the uniform motion, considering especially the light work it will have to do, the Fraunhofer regulator would I presume be everything that is required, or an electromagnetic regulator, similar to that described by me in a paper presented to the American Association at their meeting at Montreal, may be used if found reliable. From what a scientific friend has told me of his experience with distant lights, I think we are justified in anticipating the easy attainment of sufficiency of light.

A similar optical means can also be used for comparing a mean time clock at one station with a sidereal clock at another, by the method of coincidences, without other mechanism than the clocks themselves, though with diminished power of precision on a long line of stations. The pendulum of the one clock is made to carry in the focus of the telescope at its station, an opaque disc with a narrow slit, through which, at each oscillation, a flash of light is allowed to escape to the other station, and through the focus of the telescope at that station oscillates a wire carried by the pendulum of the other clock, which eclipses the flash of light at each coincidence of the two pendulums. Or the pendulum at the observing station may carry a mirror, in which either a flash or an interruption of light from the other station may be observed by reflection, and the coincidence noted when the flash or the break is seen at the same point of the field of view where it is observed with the pendulum at rest.

I have already observed that the visual method proposed in this paper might prove useful as a check, at least, upon the indications of submarine or subterranean lines of electric telegraph. But it seems less liable to uncertainty in its indications than even the air lines, the signals of which occupy a very appreciable and more or less ambiguous time in passing, and therefore on very extensive surveys it would be very instructive at least, and might be found to give increased accuracy, to add to the comparisons made by the telegraph wires, further comparisons by means of a sufficient number of the visual instruments to reach across the whole extent of the survey. In case it should

ever be undertaken, as has been proposed, to measure an extensive arc of the equator, the idea of such a visual method for the accurate determination of the differences of longitude, would be well worth considering.

I will close by suggesting one more obvious application of the method, and that is, the determination of the velocity of light, which, with a sufficiently high velocity of revolution of the prism and signal wheel, might be done with considerable accuracy by transmitting, in the same manner as before described from a second station to a third, a return signal from the second station to the first.

ART. VI.-On Osmious Acid, and the position of Osmium in the list of Elements; by J. W. MALLET, Prof. of Chemistry, &c., Univ. of Alabama.

31

2

31

IN most chemical text-books it is stated, on the authority of Berzelius, that there are five oxyds of osmium-OsO, Os,O, OsO2, OsO,, and Os0,-of which however the second and fourth have not been isolated, although compounds containing them are known. To these may be added a blue substance, first obtained by Vauquelin and supposed by Berzelius to consist of OsO united to either Os, 0, or OsO,, and the highest oxyd, probably OsO,, the existence of which was announced by Frémy in 1854.

While preparing osmium from some black platinum residues I have accidentally obtained, a substance which there is some reason to believe may be osmious acid-the hitherto unisolated teroxyd-mixed indeed with osmic acid, but still permitting certain of its properties to be observed.

Three or four ounces of the platinum residue were treated by a modification of the original process of Wollaston, now seldom adopted. The powder was mixed with three times its weight of nitre, the mixture was fused for some time in an iron crucible, and then poured out upon an iron plate. While still warm the fused cake was broken into fragments and put into a flask fitted with a cork, through which passed a tube two feet long, bent at right angles, and a funnel-tube, the latter drawn out to a very small bore at the lower end, and reaching to the bottom of the flask. The bent tube was well cooled, and undiluted oil of vitriol was very cautiously poured, by a few drops at a time, into the funnel.

The acid produced intense heat on coming in contact with the cake of potash salt, and oily drops of a bright yellow color began to make their appearance in the cooled tube. These drops very slowly congealed to a solid resembling unbleached bees-wax

SECOND SERIES, VOL. XXIX, No. 85.-JAN., 1860.

By the time the sulphuric acid had been added in slight excess a considerable quantity of this yellow substance had collected in the tube and in a receiver attached. By gentle heating the whole was obtained in the receiver, and united under a little water to a single mass. Towards the end of the distillation colorless needles and fused drops of the well known osmic acid came over, and doubtless a considerable portion of the yellow mass in the receiver consisted of the same.

At first it seemed probable that the yellow color of the latter was due merely to some impurity, and it was therefore cautiously resublimed, but it again collected of the same tint as before. It appeared to be even more fusible and volatile than osmic acid; it took a long time to congeal under a stream of cold water flowing over the outside of a tube in which it had been melted.

The water in which it was fused acquired a bright yellow color, and gave off fumes, the odor of which seemed to me somewhat different from that of osmic acid, and which irritated the eyes so insufferably that it was scarcely possible to finish work with the acid and put it up for preservation. It was removed as a single cake from the water, and sealed up hermetically in a glass tube which had been previously cleansed with care from all traces of dust or other organic matter. The water in which it had been fused was mixed with caustic potash, and gave a solution of very dark brown-red color, such a tint as would probably result from a mixture of the red* osmile of potash discovered by Frémy with the orange-brown osmiate of potash.

The sealed tube containing the fused cake or stick of yellow acid was allowed to remain upon a table exposed to the direct rays of the sun. The acid immediately began to sublime upon the sides of the tube, not in long needles and prismatic crystals like osmic acid (which seems to be monoclinic), but in feathery crusts like sal-animoniac, which under a lens had somewhat the appearance of minute octahedrons grouped together. The color was still bright yellow, but in a short time the sublimed acid began to turn black, and in twenty-four hours the whole inner surface of the tube was perfectly black and opaque. A tube containing pure colorless osmic acid has been exposed in a similar way to the sun for three weeks without any such blackening taking place. A tube closed by a cork, or one from which dust has not been carefully removed will often cause osmic acid to turn dark, but never exhibits anything like the absolute blackness and opacity of the whole tube noticed in the present in

stance.

* A rose-red color is also characteristic of the salt supposed by Berzelius to be the ammonio-terchlorid of osmium, corresponding in the chlorine series to osmite of

ammonia,

It is easy however to imagine the cause of this change undergone by the yellow acid if it be in fact the teroxyd of osmium. (mixed with osmic acid). The teroxyd probably broke up into osmic acid and one of the lower oxyds of osmium or perhaps the metal itself. We might have

In order to ascertain, if possible, which of the above changes had taken place, the tube was opened two or three months after it had been sealed, and the contents were examined. The fused stick of acid was found to be black and partially friable; on heating in another glass vessel most of it sublimed, leaving a little black powder behind, and condensed in needles, still slightly yellowish, but differing little in appearance from common osmic acid. The inner surface of the original tube was found coated with a thin filmy, adherent crust, of a black color and considerable lustre. This was scraped off, and a portion of it gently heated in a stream of dry carbonic acid gas until all traces of adherent osmic acid were driven off. After cooling, the carbonic acid was replaced by dry hydrogen, and heat was again applied. Water condensed on the tube beyond the heated. part, thus proving that an oxyd of osmium, not the metal, was under examination. Replacing again the hydrogen by oxygen, osmic acid was produced and carried off with the stream of gas. The black powder scraped off from the original tube was heated with hydrochloric acid, and seemed to be but slowly acted on; the acid however assumed a green color, and hence it is proba ble that the osmium existed as protoxyd.

It is not easy to see, without further investigation, how osmious acid could have replaced in part osmic acid in the attempt to prepare the latter as above described. Is there a particular stage of the decomposition of nitro by heat at which osmium may replace nitrogen in nitrite of potash (KO, NO,)? From the relations of the two elements, to be noticed presently, this would seem probable, and in fact Frémy has noticed the crystallization of osmite of potash from a solution in hot water of the fused cake of nitre and iridosmium. A reason for osmic acid (OsO,) being usually obtained from the latter, instead of osmious (Os0,), might perhaps be found in the fact that the chemists who of late years have worked upon osmium recommend the use of nitric or nitro-muriatic acid to neutralize the potash-sulphuric acid, to which Wollaston had recourse in his early experiments, is now seldom employed. Thomson, in his "Chem. istry of Inorganic Bodies," published many years ago, observes that osmic acid has sometimes a tint of yellow.