I have laid particular stress upon these gravels, | as they appear to form a connecting link with the diamondiferous deposits upstream, and those of the recently-discovered deposits on the coast of what was recently German South-West Africa.

These deposits were first discovered in 1907, when the sands of Luderitzbucht, a little desolate seaport on the coast north of the mouth of the Orange were found to be full of small glittering diamonds. At first the story was scoffed at, for there had been a German settlement there since 1884, and no one had ever heard or dreamed of diamonds being there. But diamonds they proved to be, and the deposit at Kolmans Kop, quite close to the settlement of Luderitzbucht, soon proved to be but one of many, indeed they have been proved to exist in varying quantities practically all up the coast from a little north of the Orange River well nigh to Walfish Bay.

The whole coast is a dreary and desolate one, bare sand dunes stretching for many miles inland; it is waterless, and with the exception of Luderitzbucht, was, at the time of the discovery, quite uninhabited—indeed, uninhabitable!

The diamonds, which are very small, and very brilliantly polished by the constant blowing sand, are found in a surface deposit of sand and grit. They are so uniform in size as to give the impression of having been graded, as indeed they apparently have, by the action of both wind and water. And though the main deposits have been found at some small distance inland, rather than on the present beach, there is evidence that the coast line has been gradually rising, and that they came from the sea.

Farther transportation has been effected by the prevailing wind, which blows with great force during most of the year, and is the cause of the shifting sand dunes, which are a feature of the country, and which blows pebbles, grit, sand, and even diamonds impartially.

Of the many deposits located, by far the richest is at Pomona, south of Luderitzbucht, where the lucky first-comers literally filled their pockets with little diamonds. And everywhere along the coast, in close proximity to the diamonds, are found pebbles identical with those of the River Diggings I have described, smaller, it is true, but having undoubtedly a common origin.

And the theory, held by many geologists, is that the diamonds were, in common with gravel, at one time washed down the Orange and into the sea, whence the north-east setting currents have borne them along the coast, and the tides have deposited them where they are found to-day. Another theory is that they have been washed up from a pipe in the sea, but this would not account for the gravel, and it is also a fact worth noting, that the size of the stones is larger near the southern extremity of the fields, near the Orange River, and becomes smaller and smaller at each successive field as one journeys north.

In any case, and whatever their origin, their discovery proved a godsend to the German Administration of the country, which had been a drain upon the Imperial German Government. Rules and regulations were hurriedly framed to control the new source of wealth, prospectors and would-be diggers flocked to Luderitzbucht, where at that time a condensing plant had to supply the only driking water, except that obtained from the

ships, and where conditions were as unfavourable either for prospecting or digging as can well be imagined.

In spite of this, and at risk of their lives, men spread up and down the coast, and extraordinary finds were made-many of them to be confiscated by the German officials; and within a month or so of the first finds, the Deutscher Bank in Luderitzbucht had an enamelled bucket three parts full of the precious stones, which had been placed in their keeping pending a decision as to the rightful

owners.

To add to the general confusion many claimants for certain valuable portions of the coast now put in an appearance, their claims being generally based upon old concessions granted by natives prior to the taking over of the land by the German Government; and many such claims had never been registered, as the land had always been considered worthless.

In addition to these claimants, several large companies laid claim to the land, others dug up. long-forgotten clauses in their charters which gave them a right to "minerals," and in some cases even specified "precious stones should such be found," and in the midst of this confusion new companies were continually being formed. So for a time chaos reigned, during which the licensed prospector was at a great diadvantage, since he never quite knew in pegging a claim whether his title to it would in the end hold good.

And he, as had been the case in Kimberley nearly forty years earlier, gradually disappeared, giving place to various companies, amongst which the better-known, such as Kolmans Kop, and Pomona, had proved, when war broke out in 1914, highly productive. The stones are very small, ranging from one-quarter to an eighth of a carat, but are good quality, and up to August of that year about 5,400,000 carats had been extracted, of an approximate value of nine-and-a-quarter millions sterling. The output was strictly controlled by the German Government, and the sales limited to a little over a million carats annually. (To be continued)

THE PHYSICAL CHEMISTRY OF BASIC SLAGS.

By CECIL H. DESCH, D.Sc.,

Professor of Metallurgy in the University of Sheffield.

The

ANY account of the physical chemistry of basic slags at the present day must of necessity be largely a confession of ignorance. We are still unaware, in spite of much excellent research work on the subject, of the form in which the principal components are combined in the slag. chemical analyses indicate the proportions in which the metallic and non-metallic oxides occur, and the sulphur percentage is also known, but there is much difference of opinion as to the nature of the silicates and phosphates which compose the slag, and as to the nature of the sulphides which occur in it as minor constituents. Large and very perfectly formed crystals have been isolated in certain instances, but it is not clear that these are identical with the usual constituents of massive slags, or even that they are mineralogically homogeneous substances. As an illustration of this last point, the case of the large and apparently perfectly formed plates sometimes

found in tap cinder from the puddling process may be mentioned. Microscopical examinations of these plates in transverse sections proves them to be intergrowths of fayalite and ferrous oxide or magnetite, resembling the "colonies" of eutectic in white iron, which also have the appearance of definite crystals, but have been shown by Prof. Benedicks to consist of two constituents, closely intergrown.

The essential components of basic slag are the oxides of iron, manganese, calcium, magnesium, silicon, and phosphorus, whilst sulphides of calcium and manganese, and calcium fluoride, may be regarded as accessory components. Only one state of oxidation need usually be considered, but iron may present itself either in the ferric or ferrous form.

As a starting point, the excellent investigation of acid open hearth slag by Messrs. Whiteley and Hallimond may be taken (J. H. Whitely and A. F. Hallimond, J. Iron and Steel Inst., 1919, i., 199). It is shown by these authors that iron occurs chiefly in the form of the orthosilicate, Fe,SiO,, fayalite, with which the corresponding manganese compound, Mn,SiO,, tephroite, forms solid solutions. When the proportion of manganese is high, the metasilicate, MnSiO,, rhodonite, is formed, and this may contain some iron in solid solution, but a distinct ferrous metasilicate is not observed. This is confirmed by the present writer's observations of puddling cinders, in which a metasilicate has not been found, even when large proportions of silica are deliberately added. Any such excess appears, both in cinders and in slags, as free silica, usually in the form of tridymite. Excess of iron oxide appears as magnetite, or as solid solutions of similar character, ranging from nearly pure ferrous oxide to crystals richer in ferric iron than magnetite. The growth of such crystals is very characteristic, and they form delicate branched crystallites. In acid slags containing more than 8 per cent of lime, an anorthic metasilicate of calcium, magnesium, iron and manganese is formed, giving to the slag a marked acicular structure. A later paper by the same authors (J. H. Whiteley and A. F. Hallimond, J. Iron and Steel Inst., 1919, ii., 159) shows that eutectics are formed between fayalite and mag netite, and between fayalite and tridymite. The writer has also observed these eutectics, but the further eutectics mentioned by Whiteley and Hallimond, namely, of magnetite and tridymite, and of all three minerals, may perhaps be attributed to imperfect equilibrium. Should they prove to occur in the system in a state of equilibrium, then ferric oxide must be regarded as a separate component.

Leaving aside the question of phosphorus, the basic slags differ from acid slags chiefly in containing a much higher proportion of lime. A glance at the ternary equilibrium diagram for the system lime-magnesia-silica (J. B. Ferguson and H. E. Merwin, Amer. J. Sci., 1919 (iv.), xlviii., 81) will show that a slag containing anything like 40 per cent of lime must be largely made up of the orthosilicate, CaSiO1. The solid solutions of this silicate with the silicates of other metals are very liable to undercooling, with the formation of glass. Consequently, a glassy ground mass is often observed in basic slags, rendering the interpretation of their constition much more difficult. Acid slags are usually very free from

glass, and even puddling cinders mostly contain mere traces of it. Sections of basic slag may exhibit glass in which are embedded small crystals of a phosphoric mineral; and it is evident that the constitution of such specimens could only be determined by making experiments in the direction of annealing, and so permitting the crystallisation of minerals from the glass.

Turning now to the phosphorus, it is remarkable that no phosphate of iron, either ferric or ferrous, is known which is stable at high temperatures. The naturally occurring phosphates are hydrated minerals, and do not yield homogeneous anhydrous products on heating. On the other hand, the phosphates of calcium are well known. Several investigators, however, have shown that silicophosphates are the principal constituents of phosphoric basic slags. The first of these compounds was described by Carnot, who attributed to it the formula Ca,(PO1)2, Ca2ŠiO, (A. Carnot, Richard and Daubre, Compt. Rend., 1883, xcvii., 136). In 1887, Stead and Ridsdale described crystals of a compound also containing phosphate and silicate, 4Ca,(PO4)2, Ca,SiO, (J. E. Stead and C. H. Ridsdale, Trans. Chem. Soc., 1887, li., 601). Other similar compounds have been described, and their composition is reviewed by Kroll (V. A. Kroll, J. Iron and Steel Inst., 1911, ii., 126), who added yet another, Thomasite, to which he attributed the composition Ca,(PO1)2, 3CaO, Ca2SiO1. At an earlier period, Hilgenstock (G. Hilgenstock, Stahl u. Eisen, 1883) had described a tetrabasic phosphate, 4CaO, P2Os, and this was long regarded as the essential constituent of all slags which were of agricultural value. It is more likely that the silicophosphates are the valuable compounds. As a rule, these compounds have been described by their external characters only, and very little is known as to their mineralogical individuality. It must be remembered that calcium is replaceable to a greater or less extent in such compounds by iron and other metals, and it is quite possible to find slags, the analysis of which will permit of inclusion under the formula of Thomasite given above. It does not follow that these slags are pure individuals; they may be eutectic intergrowths simulating pure crystals, as mentioned above. When phosphorus is added in moderate quantities to puddling cinders, no new constituent makes its appearance, and phosphorus compounds can evidently pass into solid solution in siliceous iron minerals. The examination of specimens polished on one surface only and etched, as in the examination of metallographic specimens, usually gives much more information than the study of thin sections, although the application of polarised light to the latter affords a useful means of idenifying minerals by their optical properties when their limits have been once defined. The fine eutectics in slags and cinders are generally overlooked in thin sections.

Sulphides enter into solutions to an unknown extent, but a part may remain in insoluble globules. Fluorides combine with phosphates to form minerals of the apatite class, and this, no doubt, accounts for the insolubility, under agricultural conditions, of slags containing fluorspar. Beyond this, little can be stated definitely as to the constitution of the basic slags without further research, the performance of which is highly desirable.-Transactions of the Faraday Society, December, 1920.

IT has been shown that gallium may be separated from many elements by precipitation as the ferrocyanide from a solution containing as much as a third of its volume of concentrated hydrochloric acid (Lecocq de Boisbaudran, Compt. Rend., 18821883, xciv., 1154, 1228, 1439, 1628; XCV., 157, 410, 503, 1192; 1332; xcvi., 152, 1696, 1838; xcvii., 142, 295, 522, 623, 730, 1463; Browning and Porter, Am. J. Sci., 1917, xliv., 221). Two problems arise in connection with the application of this reaction in analytical work, namely, filtration and the recovery of the gallium from the precipitated salt free from iron and the ferrocyanide radical. This paper gives the results of a study of these points.

The precipitate of gallium ferrocyanide is very gelatinous and, while it is possible to filter it on an ordinary paper filter, the filtration is very slow and it is difficult to wash the material. The large bulk of the precipitate and the slowness with which it settles make it impossible to decant the liquid with any success. The effect of changing the acid concentration and of precipitating the material in the presence of electrolytes such as potassium chloride and ammonium chloride was studied, but it was found that the precipitate does not settle any better under these conditions. Adding the reagent to the hot solution likewise failed to make any appreciable difference in the form of the precipitate and in its speed of filtration.

Various modifications of paper and asbestos filters were tested in an attempt to hasten the filtration. The use of asbestos or of paper on a Gooch crucible was found impossible, the precipitate passing through even a thick layer of the material, as it does likewise through paper on a funnel when suction is applied. If, however, a mat of fine filter paper fibre, best made by scratching some paper with a knife, is washed into the funnel, containing a double filter paper, the allium ferrocyanide is held on it, provided the suction is not applied too strongly. If, as is often the case, the first portion of the filtrate is cloudy, it should be poured through the filter again, the resulting filtrate being clear. This process can be carried out in much less time than is required for the filtration of the same amount of material without the use of suction. Of two solutions, containing equal amounts of the gallium precipitate in 10 CC. of liquid, one required 15 minutes to be filtered once through a paper filter without suction, whereas the second, by the of the method just just described, was completely filtered in two minutes, including the second filtration of part of the solution. Table I. shows the results of the quantitative determination of gallium as the ferrocyanide, the filtration having been made with suction. The material was ignited and weighed as the mixed oxides of gallium and iron (Compt. Rend., 1882, xciv., 1228), in the gallium ferrocyanide salt, Ga.(FeC.N.). If some of the iron remains as the carbide, as is probably the case, the resulting weight is not affected, on account of the identity of the molecular weights of Fe,O, and 2FeC, (Treadwell-Hall. "Analytical Chemistry," 4th ed., i., 320).

use

1.

2.

0.0386 0.0386

0'0400

0'0395

Error

G.

0'0014

0.0009

This method of direct ignition is not specially recommended, as it gives high results which in small amounts of material show high percentage It was, however, thought best to include it in the general study of the problem.

errors.

The recovery of gallium from the ferrocyanide by the ignition with ammonium nitrate and the subsequent separation of the gallium from iron by means of sodium hydroxide has been described by us (Browning and Porter, loc. cit.). The possibility of recovering the gallium without the formation of ferric or ferrous salts and the necessity of separating it from these was studied. When the precipitated gallium ferrocyanide is treated with an alkali in solution it is decomposed into the soluble alkali salts of gallium and of ferrocyanide. It was found that when carbon dioxide is bubbled through such a solution to saturation the gallium is quantitatively precipitated as hydroxide, or basic carbonate, in a form easy to filter and readily washed free from ferrocyanide. This behaviour is analogous to the reaction utilised in the detection of the constituents of zinc ferrocyanide (Treadwell-Hall, "Analytical Chemistry," 4th ed., i, 319). Table II. shows some results on the quantitative estimation of gallium after its precipitation as the ferrocyanide and the recovery of the gallium as described above. The ferrocyanide precipitate was in each case filtered on paper after standing for two days. In Expt. 1 the paper with the precipitate was treated directly in a beaker with sodium hydroxide solution and carbon dioxide, the precipitate and paper being filtered off. In Expt, 2. the precipitate was filtered and dissolved off the paper by sodium hydroxide solution and then treated with the carbon dioxide. In the other cases the ferrocyanide was filtered on paper with suction in the presence of paper fibre, dissolved in sodium hydroxide solution and finally precipitated by carbon dioxide from a volume of 100 cc.

When an alkaline solution of gallium ferrocyanide is boiled with ammonium chloride a precipitate of gallium is formed, which consists chiefly of the ferrocyanide. If, however, the ferrocyanide is oxidised over to the ferricyanide by hydrogen peroxide in alkaline solution before the ammonium chloride treatment, the precipitate obtained is gallium hydroxide free from any cyanide radicals, as shown by qualitative tests. oxidation may be made with a nitrate, but the peroxide method is the more satisfactory. This method for the recovery of gallium from the ferrocyanide works well and is rather quicker than the carbon dioxide method. Quantitative results are given in Table III. of some determina

The

Of the elements commonly associated with gallium, whose salts are soluble in alkali, zinc is the most common one that forms an insoluble ferrocyanide. The insolubility of zinc ferricyanide in acid and in ammonium hydroxide makes the separation of gallium from the ferrocyanide by means of the peroxide method impossible if zinc is present. In such a case, however, either of the following two methods may be employed.

In the first case the mixed ferrocyanides are dissolved in sodium hydroxide solution and the bases precipitated together by carbon dioxide. The gallium may now be separated from the zinc by dissolving the precipitate in hydrochloric acid and boiling the solution with ammonium acid sulphite to precipitate the gallium (Porter and Browning, Jour. Amer Chem. Soc., 1919, xli., 1491). The following procedure may also be used. The alkali solution of the ferrocyanides is treated with hydrogen sulphide to precipitate the zinc (Browning and Porter, loc. cit.), which even in the presence of the ferrocyanide is precipitated as the sulphide. From the filtrate the gallium may be recovered free from ferrocyanide either by precipitation with carbon dioxide, or by boiling with ammonium chloride after the oxidation of the excess of sulphide and of the ferrocyanide by hydrogen per oxide.

Summary.

I. A method of filtering gallium ferrocyanide with suction has been described.

2. Methods for the recovery of gallium from its ferrocyanide salt by alkali and carbon dioxide by sodium hydroxide, and by hydrogen peroxide and ammonium chloride have been developed.

3. Methods for the recovery of gallium from its ferrocyanide in the presence of zinc have been described. Journal of the American Chemical Society, January, 1921.

THE FARADAY SOCIETY.

GENERAL DISCUSSION ON "THE FAILURE OF METALS UNDER INTERNAL OR PROLONGED STRESS."

THE General Discussion on this subject which is being organised jointly by the Faraday Society, Institution of Mechanical Engineers, Institute of Metals, and the Iron and Steel Institute will take place on Wednesday, April 6 next. There will be both afternoon and evening sessions. The NorthEast Coast Institution of Engineers and Shipbuilders, The Institution of Engineers and Shipbuilders in Scotland, and the West of Scotland Iron and Steel Institute are also participating in the Discussion.

A preliminary programme has been issued which states that the Discussion will be opened by Dr. W. Rosenhain, F.R.S., who will give a general

By R.

"The Spontaneous Cracking of Necks of Small Arm Cartridge Cases." By W. C. Hothersall. "The Internal Stresses in Brass Tubes." H. N. Vaudrey and W. E, Ballard. "The Failure of Condenser Tubes."By Commdr. W. Campbell, U.S.N.R.F.

"Note on Corrosion and Season-Cracking." By G. D. Bengough.

"Corrosion Cracking of Non-Ferrous Materials.” By W. R. Woodward.

"Stress and Season-Cracking in Cold-Worked Brass Articles." By Dr. F. Rogers.

"Effects of Prolonged Stress on Metals at High Temperatures." By Dr. F. Rogers.

"Internal Stresses in Relation to Microstructure." By J. C. Humfrey.

"Failure of the Lead Sheathing of Telegraph Cables." By L. Archbutt.

Other contributions will probably be made by Prof. C. Benedicks, Mr. W. O. Ellis, Dr. F. C. A. H. Lantsberry, Dr A. McCance, Dr. E. A. Smith, Prof. T. Turner, Mr. J. E. Howard, Dr. F. C. Thompson, Prof. C. H. Matthewson, Dr. R. S. Hutton, Mr. B. S. Haigh, Mr. R. W. Webster.

An exhibition of specimens will be held in connection with the meeting, and those desirous of lending specimens are asked to communicate with Mr. F. S. Spiers, Secretary to the Joint Committee, 10, Essex Street, Strand, W.C.2.

THE ACTION OF WATER ON WOOL.-When a small sealed tube, containing woollen threads in water is heated to a temperature of 130°-135° C., the wool undergoes a decrease in volume. The shrinkage, easily noted, corresponds with a profound change in the structure of the wool. The fibres placed between crossed Nicol prisms no longer display the characteristic birefrigency of the primitive wool, and they are also pulverisable dry. The wool can be attacked by water under pressure or by steam; in fact, in the autoclave the same results are observed, but more slowly than with liquid water. With HC instead of water, results are analogous. Working with a diluted solution of potash (1/20 N for example) the wool shrinks at a lower temperature and at 130° C, disaggregation commences and finally the fibres are completely dissolved.-Bull. Soc. Chim, de France, January 20, 1921.

PROCEEDINGS OF SOCIETIES.

PHYSICAL SOCIETY,

Annual General Meeting, February 11, 1921.

PROF. SIR W. H. BRAGG, F.R.S., President, in the Chair.

THE report of the Council was read by the Secretary, Mr. F. E. Smith. The report was unanimously adopted. The report of the Treasurer was read by Mr. W. R. Cooper. The report was unanimously adopted. Votes of thanks to the Honorary Auditors, the Officers and Council, and the Governing Body of the Imperial College were unanimously passed, the respective proposers and seconders being Dr. Chree and Dr. Lister, Mr. Clark and Dr. Borns, Dr. Thomas and Mr. Paul. The scrutators, after examination of the ballot papers, declared the Officers and Council for the ensuing year to be elected in accordance with the following list :

President-Prof. Sir W. H. Bragg, C.B.E., M.A., F.R.S. Vice-Presidents (who have filled the office of President)-C. Chree, Sc. D., LL.D., F.R.S.; Prof. H. L. Callendar, M.A., LL.D., F.R.S.; Prof. R. B. Clifton, M.A., F.R.S.;. Sir Richard Glazebrook, K.C.B., D.Sc., F.R.S.; Sir Oliver J. Lodge, D.Sc., F. R.S.; Prof. C. H. Lees, D.Sc., F.R.S.; Prof. A. W. Reinold, M.A., F.R.S.; Sir Arthur Schuster, Ph.D., Sc.D., F.R.S.; Sir J. J. Thomson, O.M., D.Sc., F.R.S.; Prof. C. Vernon Boys, F.R.S. VicePresidents-Prof. W. Eccles, D.Sc.; Prof. A. S. Eddington, M.A., M.Sc., F.R.S.; The Rt. Hon. Lord Rayleigh, F.R.S.; Prof. Sir Ernest Rutherford, D.Sc., F.R.S. Secretaries-F. E. Smith, O.B.E., F.R.S., National Physical Laboratory, Teddington; D. Owen, B.A., D.Sc., 62, Wellington Road, Bush Hill Park, N. Foreign Secretary-Sir Arthur Schuster, Ph.D., Sc. D., F.R.S. Treasurer-W. R. Cooper, M.A., B.Sc., 82, Victoria Street, S. W. 1. Librarian-Prof. Á. O. Rankine, D.Sc., Imperial College of Science and Technology. Other Members of Council-G. B. Bryan, D.Sc.; C. R. Darling, F.I.C.; Prof. C. L. Fortescue, O.B.E.; E. Griffiths, D.Sc.; F. L. Hopwood, D.Sc.; E. H. Rayner, M.A., D.Sc.; A. Russell, M.A., D.Sc.; T. Smith, B.A.; J. H. Vincent, D.Sc., M.A.; Prof. W. B. Morton, M.A.

After the conclusion of the general business, a discussion on "Absolute Measurements of Electrical Resistance, and Instruments based on the Temperature-Variation of Resistance" was held.

SIR RICHARD GLAZEBROOK opened the discussion with a historical review of the work of early workers on absolute resistance measurements and the gradually increasing accuracy which had been obtained in such measurements with the development of improved methods and apparatus.

Mr. F. E. SMITH wished to confine himself to two suggestions. The first was that the accuracy of absolute determinations being now in excess of the accuracy with which the International Mercury Standard could be reproduced, such standard was now useless, and workers should express their results in true ohms rather than in international ohms. His second suggestion was that teachers should familiarise their students

a

with absolute methods of resistance measurement, and he showed a simple apparatus with which such measurements could be made in the laboratory in a few minutes.

Prof. H. L. CALLENDAR explained the principles of his temperature-compensated resistance bridge, and exhibited various resistance thermometers for special purposes and several instruments for radiation measurements.

Mr. C. R. DARLING described the early work of Siemens on resistance thermometry.

Major TUCKER exhibited and described his hotwire microphone.

Prof. J. T. MCGREGOR MORRIS exhibited and described a robust and portable hot-wire anemo

meter.

Miss HARGOOD-ASH, in the absence of Prof. Leonard Hill, exhibited and explained the latter's "Calcometer."

Mr. E. A. GRIFFITHS exhibited and described a resistance gauge for measuring the depth of liquid in a tank.

Mr. A. H. DAVIS read a Paper on "An Instrument for Measuring Convected Heat."

Dr. J. S. G. THOMAS exhibited and explained his Directional Hot-wire Anemometer.

Dr. DAYNES, in the absence of Dr. G. A. Shakspear, exhibited and explained the latter's gas permeameter.

Other instruments were exhibited, but owing to the lateness of the hour were not described.