salt is a powerful oxidizing agent, and produces greenish precipitates with metallic solutions.-Bull. Soc. Chimique, 1875, October 5.

IMPROVEMENT IN THE AMMONIA-SODA PROCESS.

The following improvement in the ammonia process for the manufacture of soda has been patented in England. The solution of chloride of ammonium formed is mixed with carbonate of magnesia and distilled, and the carbonate of ammonia, which passes over, is again used in the manufact ure of soda. The solution of chloride of magnesium remaining, which also contains some chloride of sodium, originally present with the chloride of ammonium, is concentrated, and the chloride of sodium removed from time to time as it crystallizes out. What remains is then calcined with the forma tion of hydrochloric acid and magnesia. The latter is reconverted into carbonate by the gaseous products of combustion from the furnace. The separation of the chloride of sodium before the calcining is important, as otherwise the chloride of magnesium will not be decomposed by heat alone. A partial decomposition of the chloride of magnesium, however, is said to be more favorable to the conversion into carbonate than complete decomposition.-5 C, XXXVIII., 1875, 303.

MANUFACTURE OF SODA FROM SEA-WEED.

Herland, in Comptes Rendus, calls attention to some defects in the treatment of sea-weed for the manufacture of soda, and makes some suggestions for an improvement of the process. He remarks that in the ordinary method the weeds destined to be calcined are dried on the shore in the open air, involving decomposition and a loss of salts by moisture, and that the process of incineration is itself de fective. The method by which he proposes to remedy the difficulty is as follows: Place the fresh weed in baskets of iron wire, moved by a turning crane, and steeped in a series of vats containing about fifty kilogrammes of good quicklime per cubic meter, and so placed that it can pass in succession from vat to vat until it is exhausted of its useful salts. The time of steeping in each vat is from forty to sixty minutes. The successive saturated lyes are to be reduced to dryness and treated with carbonate of potash. This will yield a

soda very rich in soluble salts and alkaline iodides. While the incineration gives but fifteen per cent. of potash and one per cent. of iodide, the new process yields forty to fifty per cent. of potash, and sometimes six per cent. of iodide. The residual weeds are applicable to agricultural purposes.-1 A, July 21, 1876, 32.

ATOMIC WEIGHT OF CERIUM.

Bührig, by means of a series of careful combustions of cerium oxalate, has very accurately redetermined the atomic weight of the metal. This, based upon the values given by Stas for carbon and oxygen, he finds to be 94.1782. This value is the mean of results obtained in fifteen experiments. -Journal für prakt. Chemie, XII., 209.

THE CERIUM METALS.

The rare metals cerium, lanthanum, and didymium have at last had their atomic weights definitely fixed by Hillebrand. He determined the specific heat of each metal by means of Bunsen's ice calorimeter, getting the following values: Cerium, 0.04479; lanthanum, 0.04485; didymium, 0.04563. These figures, considered with reference to Dulong and Petit's law, correspond respectively to the atomic weights 138, 139, and 144.78. These observations prove that the lower oxides of these metals are really sesquioxides, and the higher oxides, dioxides. Until lately the atomic weights of this group have been rated at only two thirds of the above values.-Poggendorff's Annalen, 1876, No. 5.

CERIUM, LANTHANUM, AND DIDYMIUM.

These rare metals have recently been isolated in quantity by Drs. Hillebrand and Norton, students under Bunsen at Heidelberg. Cerium outwardly resembles iron, is very ductile and malleable, and tarnishes readily. In hardness it about equals silver, and its melting-point is at a full red heat. Heated in the air it kindles very easily, at a lower temperature even than magnesium, burning with great brilliancy. Upon shaving a cerium wire with a knife the thin fragments of metal ignite as they are cast off, while with an ordinary fire-steel sparks can be struck from it as from a flint. Didymium and lanthanum resemble cerium closely, but are less

inflammable, and do not tarnish so easily. The specific gravities are as follows: Cerium, 6.628 to 6.728; lanthanum, 6.049 to 6.163; didymium, 6.544. In their electrical relations the three metals lie between magnesium and zinc; they are electro-positive to the latter, electro-negative to the former.Poggendorff's Annalen, March 25, 1875.

EXPERIMENTS ON STEEL.

MM. Trève and Durassier have communicated to the French Academy the following results of their investigations into the relation existing between the chemical and physical properties of steel. They prepared fifteen bars of steel, divided into five series of three bars, each series being of a different degree of carburation, and each tempered in a different manner. These bars were then magnetized to saturation, and their magnetic force determined by the method of deflection. The bar containing 0.95 per cent. of carbon, and tempered with cold water, caused the maximum of deflection, represented by 47. The bar containing the same amount of carbon, but tempered with oil at 10° Centigrade (50° Fahr.), gave 43. That the magnetic capacity of steel is influenced by the nature of the liquid used for tempering seems therefore to be manifest. The bar containing 0.25 per cent. of carbon gave a deflection represented by 13, showing how decided is the influence which it exerts. In preparing diagrams of variations for the different series of bars, the influence of the percentage of carbon and of the tempering liquid became manifest. It was remarked, however, that at the top of the scale in carbon, the nature of the tempering liquid no longer exerted a very marked influence. The study of these diagrams brought to light, likewise, the noteworthy fact that a remarkable similarity existed be tween the magnetic curves and the curves of elasticity of the bars. The investigators assert from their experimental trials that carbon gives elasticity to the steel, and at the same time enhances its capacity for magnetism.-3 A, V., 556.

FLUENCE OF FOREIGN INGREDIENTS UPON THE PHYSICAL PROPERTIES OF STEEL.

erience that has of late been gained in the prohosphuretted steel may be characterized as a

triumph which promises the most substantial results for this important branch of metallurgical industry. It has been demonstrated that the peculiar properties of steel are not governed absolutely by the presence of a certain percentage of one or two substances, like phosphorus or carbon, but that these properties may be secured and modified at pleasure by judiciously controlling the relative proportions of a number of foreign ingredients introduced into the metal. The progress that has been made in this direction has been intelligently summarized by The Engineering and Mining Journal: "By securing the proper relative proportions of carbon, phosphorus, silicon, and manganese, a steel of great softness and strength can be obtained, while the same percentage of phosphorus in an ordinary steel would have indicated very different properties (which means, in other words, that it would have been quite worthless). There can no longer be much doubt that manganese exerts upon steel a body-giving, toughening influence, as well as a neutralizing effect upon the hardening or cold-shortening due to phosphorus. Though these properties of manganese have been suspected for some time, the mutual dependence, and to a certain extent interchangeability, of carbon and phosphorus were not fully appreciated until M. Tessié du Motay succeeded in producing with ferro-manganese a good steel rail, containing about 0.12 per cent. of carbon, 0.25 of phosphorus, and 0.75 of manganese." In the light of the recent investigations, therefore, phosphorus is no longer entitled to the evil distinction of being, as a well-known metallurgist has expressed it, "the very scourge and pestilence of the steel-maker," and the time is possibly not far away when many rich deposits of ore now esteemed to be worthless will find ready utilization.

GALLIUM.

M. Boisbaudran states that in his first discovery of gallium he did not possess more than one hundredth of a milligramme of the metal dissolved in a minute drop of liquid. Lately, with a larger supply of material, he has determined some of its characteristics. The spectrum is exceedingly brilliant, giving with a gas flame the line 417; in a spark spectrum line 403.1 also appears. The metal can be deposited

electrolytically upon copper, forming a hard layer, difficult to burnish and rather whiter than platinum. It was originally found in a blende from the Pyrenees, but has since been detected in blendes from various other sources, so that the discoverer thinks it may be quite widely diffused. Its most important chemical feature is that it forms apparently a true alum, thus belonging with indium and aluminium in classification. This fact is particularly interesting when considered in connection with a brief article by Mendelejeff. This author some few years ago, in his paper upon the “periodic law," ventured upon theoretical grounds to predict the existence of several then unknown metals. In all probability gallium is one of these, and will fill a gap which Mendelejeff pointed out in the aluminium group. The verification of such a prophecy might well afford a parallel to the discovery of the planet Neptune.-1 A.

GALLIUM.

M. Boisbaudran has succeeded in isolating about half a gramme of this new metal, and has redetermined some of its leading characteristics. In the fused state it is of a fine silver white color; but in crystallizing it assumes a pronounced bluish tint, and loses considerably in brilliancy. The metal, by cooling from fusion, may be obtained in isolated crystals, which are octahedral. The melting-point, previously put at 29.5° Centigrade, is now, as a mean of six closely agreeing measurements, fixed at 30.15°.

Great interest attaches to the specific gravity of gallium. A few months ago it was determined to be 4.7 at 15° Centigrade, the determination being made with a specimen of metal weighing only six centigrammes. But, calculated upon the basis of Mendelejeff's law-the law by which the existence of gallium was predicted-the value should have been 5.9. Two new determinations, now published by Boisbaudran, give 5.935 and 5.956. This agreement with theory is very close and very striking.—6 B, LXXXIII., 611, September 18.

DESILVERIZATION OF ARGENTIFEROUS ZINC.

The desilverization process introduced by Parkes in 1850, but subsequently neglected in consequence of the difficulty