them the general formula MO, HO+12(MO, SbO3+nHO), in which M expresses the positive metal, and n the number of atoms of water. This relation of 13 equivalents of base to 12 equivalents of acid was so unusual, that the formulae were generally received with some mistrust. But on calculating Heffter's results Iwith the new equivalent of antimony, numbers are obtained >which agree very well with the general formula MO, SbO+nHO, thus establishing their relation with a large class of salts.

Mallet has recently made a determination of the atomic weight of lithium. He estimated the chlorine in a very pure specimen of chloride of lithium prepared from spodumene, and took much larger quantities than had hitherto been done. As the mean of three very closely agreeing experiments, he obtained the number 86:89, or 6.95 on the hydrogen scale. This number is higher than that found by Berzelius, 654, and which has hitherto been usually adopted; but Mallet shows that the method used by Berzelius was not entirely free from objection. Mallet points out, that if we take the mean of the equivalents of lithium and potassium, we get

which is almost exactly the equivalent of sodium, 23.00, as determined by Pelouze.

Troost has published some experiments on lithium and its compounds. The method by which he obtained the lithia salts is suitable for their preparation on a large scale. It consists in heating in a crucible in a good wind-furnace, lepidolite mixed with carbonate and sulphate of baryta in certain proportions. At the bottom of the crucible there is obtained a glass perfectly melted but viscous, and above a very fluid mass, which may readily be poured off while hot. This consists of a mixture of sulphate of baryta, potash and lithia. By treating this with water, the alkaline sulphates are dissolved out. Lithia may also be obtained from petalite by the same process, if a sufficient quantity of alkalies be added to make the whole proportion the same as in lepidolite.

Troost obtained lithium by Bunsen's method. He also made a great many attempts to procure it by the method used by Deville in the preparation of sodium. To this end he heated a mixture of carbonates of lithia, chalk, and charcoal to a white heat for several hours. It was found, however, that lithium was not volatile. By acting on chloride of lithium with sodium, an alloy of lithium Silliman's Journal, November 1, 1856.

+ Comptes Rendus, November 10, 1856.

and sodium is produced. Lithium has the greatest analogy in its chemical relations to magnesium, and it appears to play the same part in the series of alkaline metals as magnesium in the series of alkaline earths.

M. Riche has investigated tungsten and some of its compounds. He prepared the metal by passing hydrogen over tungsten heated to a very high temperature. At lower temperatures bodies containing more or less oxygen are formed. The tungsten thus produced is not melted, or even aggregated; it is in the form of small crystalline grains, which by friction assume a metallic lustre, and scratch glass with great ease; exposed in a wind-furnace to a temperature sufficient to soften the crucible in which it was contained, it remained solid, and it could only be melted by employing a Bunsen's battery of 200 elements.

Tungsten is acted on by air or even by oxygen only at a high temperature, and is only attacked by chlorine at 300 degrees. It is slowly oxidized by nitric acid to tungstic acid, and by strong sulphuric or hydrochloric acids it is converted first into the oxide of tungsten, and finally into tungstic acid. Heated with iodide of methyle in a closed tube, an organo-metallic compound is produced which has the formula 3(C2 H3) W, I.

Riche determined the equivalent of tungsten by passing hydrogen over perfectly pure tungstic acid. He obtained results which fixed the number at 87. This is a little lower than the number usually admitted, and probably arises from the greater purity of the substance used in this case.

The oxychloride is readily obtained by passing chlorine over a mixture of tungstic acid and powdered charcoal heated to dull redness, and redistilling the product formed in a current of hydrogen. It has the formula W, C12 O, and forms with water tungstic and hydrochloric acids,

W C120+2HO=W 03+ 2HCI.

Terchloride of tungsten, W CP3, is obtained by passing chlorine over metallic tungsten, moisture being carefully excluded. It crystallizes on sublimation in long steel-gray needles which melt at 218 degrees, and the fracture of which presents the appearance of iodine. The bichloride W C12 is obtained in small quantities by passing hydrogen over the terchloride. Bisulphide of tungsten is obtained by melting together equal weights of tungstate of potash and sulphur until gentle fusion. On treating the cooled mass with water the bisulphide is left in small acicular crystals of a black colour, which are changed to red by contact with the air.

* Comptes Rendus, February 4, 1856.

Peligot describes a method for obtaining pure metallic uranium. He prepares green protochloride of uranium by the action of chlorine on a mixture of an oxide of uranium and carbon at a

high temperature. A quantity of sodium sufficient to decompose the protochloride of uranium is placed in a porcelain crucible and on this a layer of chloride of potassium; and this is covered with a mixture of the same salt and of protochloride of uranium. The crucible with its cover is placed in another crucible, which is filled up with charcoal powder. The addition of chloride of potassium has the object of rendering the action less active and less instantaneous.

The crucible is heated until the reaction is set up, which is known by a sound heard at the moment; the crucible is then removed to a wind-furnace, and heated to a white heat for 15 or 20 minutes. When the crucible is cold, a melted slag is found which contains globules of uranium.

The metal thus prepared has a certain malleability: it is not so hard as steel; its colour is like that of nickel or iron.

It assumes in the air a yellowish tint in consequence of a slight superficial oxidation. Heated to redness, it exhibits a sudden incandescence, and becomes covered with a voluminous black oxide, in the interior of which some unoxidized metal is found if the heat has not been carried too far.

Its specific gravity is 18.4, and it is therefore next to gold and platinum, the heaviest body. The metal may also be prepared by the action of aluminium on the green chloride of uranium. This action is evidently due to the greater volatility of the chloride of aluminium.

M. de St.-Claire Deville† gives a description of two forms of apparatus which he uses for the production of very high temperatures. His description is accompanied by drawings, without which it is difficult to give an adequate idea of their construction. One of them is a lamp-forge which he uses in the production of high temperatures in mineral analysis; by its means a temperature very near that of an iron assay may be produced; felspar and albite melt and become liquid. The fuel used is oil of turpentine, but other volatile oils may also be employed. The principle of the furnace is to increase as much as possible the surface of the carbon, and to restrain the combustion within a very small height. For fuel he uses cinders freed from slag and ash, and about the size of a hazel-nut, by which a higher temperature is obtained than with gas-coke.

At the elevated temperature obtained by this furnace, the best

Comptes Rendus, January 21, 1856.

† Annales de Chimie et de Physique, February 1856.

earthen crucible becomes as liquid as glass. This arises from the impurities accompanying the clay, for the silicates of alumina, especially those containing an excess of alumina, do not melt easily. In one experiment, Deville found that platinum was melted at a temperature at which topaz was not attacked. Deville used crucibles of quicklime, which are simply cut with a knife from pieces of lime well baked and slightly hydraulic. For certain experiments he used crucibles and tubes of carbon, cut in a lathe from the graphite of gas retorts. They may be purified by passing a stream of chlorine over them while heated to redness. This deprives the graphite of the sulphur, iron, silica, and alumina, which are converted into the chlorides and volatilized. Crucibles of alumina are also used for certain operations.

With this furnace Deville has succeeded in melting many substances of the most infusible nature. Platinum melts in a lime crucible, and has then properties very different to those of the ordinary metal. Ordinary platinum may be shown to be full of minute holes; the melted platinum is quite free from these: melted platinum does not determine very readily the union of hydrogen and oxygen; it has great malleability and ductility. When the heat is raised a little above the melting-point of plati num, it volatilizes with great facility. When the experiment is made in two crucibles, both hermetically sealed, there is found on the cover of the outer crucible a multitude of little globules of platinum, some as large as a pin-head, others only to be seen with the lens, and presenting very much the appearance of the globules of mercury obtained in blowpipe analysis.

17

Metallic manganese is obtained by mixing pure peroxide with a quantity of sugar-charcoal not quite sufficient for its complete reduction, placing the mixture in a lime crucible which is enclosed in another and heated under certain, precautions.

T

The metal is found united to a regulus, and surrounded by a reddish-violet crystalline mass, which may be a spinelle of manganese and lime, Mn 03, CaO. The metal is pure; it contains no carbon, having been melted in an excess of oxide. It has a rose-coloured reflexion like bismuth, and is brittle like that metal, although it is very hard. Its powder decomposes water at a point a little above the ordinary temperature.

TO

Chromium was obtained by melting pure oxide of chromium with a quantity of carbon insufficient for complete reduction. It was melted, but not to a regulus, although the heat employed would have fused and volatilized platinum. Chromium cuts glass like diamond; it is about as hard as corundum. It is attacked by hydrochloric acid, but not by nitric acid, whether dilute or concentrated.

Nickel is obtained like manganese; it melts to a very homo

geneous regulus, which is forged with great ease. Its ductility is almost unlimited, and its tenacity is as great as that of iron. A nickel wire which requires 90 kilogrammes to break it, has the same diameter as an iron wire which is broken by 60 kilo. grammes.50

Cobalt is prepared in the same way. It is as ductile as nickel, and more tenacious. Its tenacity is to that of nickel as 115 to 90. Cobalt is hence nearly twice as tenacious as iron. Silica is the most refractory body on which Deville has experimented, but even this he has succeeded in melting.

༣.

་་་་

XX. Polarized Light vibrates in the Plane of Polarization. By C. H. A. HOLZMANN*.

IF

F we take a diffraction grating with parallel slits, which, for the sake of shortness of expression, I will assume shall stand vertical, and if upon this grating a horizontal pencil of light be allowed to fall, the light will be spread out by diffraction in a horizontal direction. If the incident light be plane polarized, we shall be able to decompose its vibrations into vertical, which are thus parallel with the slit, and into horizontal, The vertical vibrations will not be changed by diffraction; the horizontal ones, however, will; they will be decomposable into such as lie in the direction of any assumed diffracted ray, and into such as are normal to this direction. The first will give condensed waves in this direction, and will not be perceived as light; the second, on the contrary, will proceed as transverse waves in the direction of the diffracted ray, and, together with the vertical vibrations proceeding in the same direction, will give the diffracted light.

"

The diffracted light has hence the same vertical, but another and smallment than incident light; and the direction of its vibration cannot therefore coincide with the direction of vibration of the incident light. It must be steeper than the latter. From this it is easy to determine the direction of vibration in the diffracted light. If the direction of the incident ray of light be normal to the surface of the diffraction grating, and if s be the distance of an æther particle at the time t from its position

6 distan, then, under the assumption that t

waves only meet the grating, the vertical projection of s, if a be the angle of the direction of vibration, or of s with the vertical, is equal to in: bit 109 - bud 1-11,1# bros antired jou

s cosa;

* Translated by Dr. E. Atkinson from Poggendorff's Annalen, No. 11, for 1856-1978 of silom in 19490mnasia el bomido at Da