one in a place so remote as Kasan? From the position of these two places, it is obvious that these irregularities cannot in any way be connected with the diurnal motion of the sun, as the hourly variations obviously are. What is most remarkable in the irregular variations is, that the motion of the needle is much more frequently towards the east than in the contrary direction.

From a series of observations on the irregular variations at different places, M. Kupffer has discovered that there is a regular period during the night in which the needle has a uniform movement as it has during the day. He finds that at St. Petersburgh the needle marches towards the west from nine o'clock in the evening till almost one o'clock in the morning; that the arc travelled over by the needle is three minutes, whereas the diurnal arc travelled over from nine in the morning till two in the afternoon is nearly fifteen minutes at St. Petersburgh, and twelve at Kasan. The observations have been chiefly made at the same moment at St. Petersburgh, Nicolaïeff, and Kasan; and such is the coincidence, that choose any hour in a whole table of observations, at which the needle has suffered irregular oscillations at St. Petersburgh, and it will be found to have suffered similar perturbations at Nicolaïeff. In a long series of observations, only two exceptions have been found to this general rule. The author observes, the sign of the diurnal variation changes in crossing the irregular line called the magnetic equator, and gives a simple rule for remembering the direction in which the daily variation takes place. If we stand at the southern extremity of the needle, in the northern magnetic hemisphere, with the face to the needle, its north pole turns towards the left from eight o'clock in the morning till two in the afternoon; the same thing takes place from eight o'clock in the evening till two in the morning; but from two o'clock till eight, whether in the morning or evening, it marches towards the right. Over the whole of the southern magnetic hemisphere, the periods are the same, but the deviations are in the opposite directions.' The author remarks, that the extent of the hourly variations depends on the length of the projection of the needle on the plane of the magnetic equator; that in all probability it depends on the diurnal revolution of the magnetic equator, and is probably a phenomenon closely allied to those of revolving discs of metal discovered by M. Arago*.

II.-CHEMICAL SCIENCE.

1. MATTEUCI ON THE ORIGIN OF THE ACTION OF THE VOLTAIC PILE. A HIGHLY important discussion is at present in progress, relative to the original source of electricity in the voltaic pile, not originating with, but to a considerable extent renewed by, the endeavours of M. A. de la Rive, to prove that chemical action is the sole cause;

Voyage dans les Environs du Mont Elbrouz dans le Caucase. Par M. Kupffer, Membre de l'Académie des Sciences de St. Pétersbourg.

contact of dissimilar metals having no effect. This has been vigorously controverted by MM. Pfaff, Marianini, &c.: but we wish here merely to quote the experiment of M. Matteuci, who has joined in the investigation.

He was convinced by Pfaff's experiments, that electricity could be developed by contact only; but to be more convinced, he made certain experiments on frogs. First, he assured himself that there was no chemical action between distilled water perfectly free from air, and zinc, either alone or in contact with copper. After contact of many hours, the most sensible tests could not shew the presence of the oxide of zinc or copper. It would, therefore, be very unjust (in the present question) to assume that there is chemical action because there is development of electricity. Being sure upon this point, a prepared frog was then suspended from a rod of zinc, which was fixed at the bottom of a gas jar, and connected with a long copper wire, so that nothing more was required to produce the well known contraction than to touch the muscles of the legs with the copper wire.

To remove every suspicion of chemical action, the frog was washed in distilled water, freed from air, so as to remove all animal fluid. It was then suspended by the nerves from the zinc, the jar filled with distilled water, and then with pure hydrogen: but on touching the limbs with the copper wire, the same contractions took place as in common air. The experiment was repeated in vacuo, carbonic oxide, carbonic acid, and oxygen, sometimes dry and sometimes damp, but always with the same result.

M. Matteuci, therefore, remains convinced that the mere contact of different metals is able to develope electricity; although he admits with most philosophers, that chemical action exerts an influence over this force just as heat does in thermo-electric experiments.*

2. CONDUCTING POWERS OF LIQUIFIED GASES.-
(K. T. Kemp.)

By making liquified sulphurous acid gas a part of the circuit in a galvanic battery of 250 pairs of plates, shocks were received, water was decomposed, and the galvanometer was acted on as if a continuous metallic communication had existed. Liquid sulphurous acid is therefore an excellent conductor of electricity. Cyanogen, on the contrary, was found to be a perfect non-conductor, even to a voltaic current from 300 pairs of plates. Liquified chlorine was also found to be a perfect non-conductor of electricity from a battery of 250 pairs of plates. The author then tried liquified ammoniacal gas, but could not ascertain whether it was a conductor or non-conductor of electricity. It is, in all probability, a non-conductor †.

3. GENERATION OF STEAM BY HEATED METAL.

In boilers of high pressure engines, the heat applied to a part not † Edin. Journal of Nat, and Geog. Science. May, 1831. 2 S

* Ann. de Chimie, xlv. 106. VOL. I.

containing water, sometimes raises that part of the boiler to a dull red heat, which suddenly coming in contact with a portion of the contained water, generates steam with such rapidity as to burst the boiler. In order to ascertain the effects of different metals, raised to different temperatures, in generating steam from boiling water, a series of experiments have been undertaken by Mr. Johnson, of Philadelphia, which contain the following results. He found by immersing iron raised to different temperatures in boiling water, that more steam was generated in a given time by iron of a red heat, just visible in day light, than by the same piece of iron raised to a white heat. This may arise from the greater quantity of steam forming an atmosphere around the white hot iron, and thus preventing the water coming in contact with the iron. The steam generated bears a direct relation to the weight of the metal, being about one pound of steam for every nine pounds of iron. In comparing cast iron with malleable, he found that cast iron raised to the same temperature generates more than wrought iron, being about one pound of steam for every eight pounds and a quarter of iron*.

4. ON THE PREPARATION OF IODIC ACID.-(Serullas.)

In the course of experimental investigations, M. Serullas had made out that iodic acid was insoluble in alcohol, and that the perchloride of iodine acted upon, and was decomposed by, water. Upon these two points he has founded a process for preparing pure iodic acid in an economical manner. For this purpose the perchloride of iodine, saturated to the utmost with chlorine, and in the solid state, must be prepared, and a small quantity of water, or, what is better, a solution of the perchloride put into the flask with it; some fragments of glass, for the purpose of detaching the solid chloride from the vessel, are also to be added, and then this mixture of glass, fluid and solid chloride of iodine is to be transferred by a funnel into a stoppered 8 or 10-ounce phial, in which it may afterwards be strongly agitated-the funnel retains the fragments of glass, which may be washed with a little saturated solution of perchloride. The bottle is then to be violently shaken to reduce the solid chloride to powder, that all parts may come in contact with the liquid, and that it may be freed from subchloride as much as possible. The whole is then to be poured into a capsule, as much liquid as possible decanted, and then small quantities of ether, or strong alcohol, added, agitating the whole with a glass rod. Almost immediately the solid part becomes white, and the liquid yellow: it is then to be decanted, and the washing repeated, until the liquid comes off colourless. In this way the acid is ob tained as a white crystalline powder, perfectly pure, which, being dried and pressed under the finger, feels like very fine sand; or it may be dissolved, filtered, and crystallized by mixture with sulphuric acid, as formerly described t.

* Silliman's Journal, vol. xix. p. 292, † Quarterly Journal of Science, xxix. 411.

It is important to wash away all the subchloride, which may be done by using colourless saturated solutions of the perchloride. It appears probable, from the change of colour, that the first contact of water converts nearly the whole of the perchloride into iodic acid *.

5. ON THE PRECIPITATION OF THE VEGETO-ALKALIES BY IODIC ACID. (Serullas.)

On a former occasion M. Serullas shewed that solution of iodic acid, added to solutions of the vegeto-alkaline salts, produced an abundant precipitate of acid iodate. In this way the smallest quantity of a vegeto-alkali may be shown by this acid, or by a solution of perchloride of iodine, because of the iodic acid there present. The sensibility is so great that the acid may be considered in this respect as one of the most delicate tests which chemists possess: the hundredth part of a grain of quinia or cinchonia, dissolved in several thousand times its weight of alcohol, may in this way be precipitated, and the precipitate quickly collected. The iodic acid should be so dilute as not to be precipitated itself by the alcohol; that is always the case when solution of the perchloride of iodine is used. All the vegeto-alkalies are not equally evident to the test, but the least sensible is discovered when one-fifth of a grain is present.

Great excess of the acid is required in these experiments; it should, therefore, be added by drops. From this circumstance, the alkalies are no test of iodic acid. M. Serullas hopes to found on these effects a process for trying the strength of Peruvian bark.

The iodine with which the acid is formed ought to be quite pure, and the alcohol also, with which the iodic acid is washed, ought to be purified; the latter by mixture with a few drops of sulphuric acid, and distillation to remove lime; the former by using iodine precipitated from its alcoholic solution by water.

The precipitates formed by iodic acid in alcoholic solutions of the vegeto-alkalies, when dry, are decomposed, with explosion, by temperatures of 240° to 248° F.; even when heated upon paper, upon using a tube, the detonation is of considerable force. The results are easily understood; the effect forms a character of the compounds †.

6. ON THE ACTION OF BROMIC AND CHLORIC ACIDS ON ALCOHOL.

(Serullas.)

In consequence of some suspicions relative to a process for the preparation of bromic acid, in which alcohol was used, M. Serullas mixed about equal quantities of solution of that acid and strong alcohol; the liquids immediately became coloured, much heat was evolved, the fluids boiled, and very powerful vapours of acetic ether

escaped. The acid acted nearly as powerfully as nitric acid under similar circumstances.

Concentrated chloric acid, poured into strong alcohol, acted powerfully; producing ebullition, and vapours of acetic acid. When the quantity of alcohol was small, the whole was converted into strong acetic acid; and when there was still less alcohol, inflammation occurred.

Dried folded filtering paper dipped into chloric acid takes fire, and exhales the odour of nitric acid.

In the acetification of alcohol by these hydrogenative agents, no carbonic acid is produced.

The bromic and chloric acids had been prepared by the process of Mr. Wheeler, namely, adding silicated hydrofluoric acid in excess to the heated bromate or chlorate of potassa, continuing the heat till the excess was driven off, cooling and filtering the liquid, evaporating with the usual precaution, and refiltering the concentrated liquid through glass *.

7. ON PERCHLORIC ACID AND ITS FOSSIL FORMATION.

(Serullas.)

The chloric acid just described, differed a little from that de. scribed by other chemists; but, upon heating it, it was found to become the same. A singular result of the action of heat, applied to a greater extent, was observed in the change of one part only into chlorine and oxygen; the other part, nearly one-third, became perchloric acid, probably often considered by chemists as chloric acid. After distilling chloric acid for some time, therefore, the aqueous products may be rejected, but there remains in the retort, adhering to its sides, a dense colourless liquid, which, by raising the heat sufficiently at last upon all the surface of the retort, may be driven over into a clean receiver.

This is perchloric acid, and, though very concentrated, it will not inflame paper like chloric acid, but enables it to burn with small sharp detonations. It may be distilled at high temperature without further change. At first it has a reddish colour, from a little manganese, perhaps; but that passes off on a second distillation. It is not altered when heated with muriatic acid or alcohol; the process is one which will give perchloric acid far more readily than that of the discoverer Stadion.

A perchlorate of potash was made with this acid and the alkali. This being decomposed carefully by heat, gave 46 of oxygen per cent. This result corresponds with the composition of the acid already given, i. e. 1 proportion of chlorine and 7 of oxygen †.