Messrs. Longmans & Co's List. A TREATISE ON LIGHT. By R. A. HOUSTOUN, M.A., Ph.D., D.Sc. With 328 Diagrams. 8vo. 7s. 6d. net. "Dr R. A. Houstoun's Treatise on Light,' now before us, occupies a place of its own. It will be welcomed as a manual for classes of a more advanced character than those in which optics is taken merely as a part of a general physics course. It is, indeed, alive with modern information and research; and, as numerous passages reveal, it is written by one to whom optical laboratorv work is familiar, and who directs it to bring out useful and important results "-Nature. R. H. A. PLIMMER, D.Sc. With a Coloured Plate of Absorption Spectra and 86 Illustrations in the Text. Royal 8vo. 12s. 6d. net. An AMATEUR'S INTRODUCTION to CRYSTALLOGRAPHY. By Sir WILLIAM PHIPSON BEALE, Bart., K.C., M.P. 8vo. 4s. 6d. net. An INTRODUCTION to the MECHANICS of FLUIDS. By EDWIN H. BARTON, D.Sc., F.R.S.E. 8vo. 6s. net. By WILLIAM WATSON, A.R.C.Sc., D.Sc., F R.S. Assistant Professor of Physics at the Royal College of Science, London. TEXT-BOOK of PRACTICAL PHYSICS. THIRD IM- 1914. TEXT-BOOKS OF PHYSICAL CHEMISTRY. A SYSTEM of PHYSICAL CHEMISTRY. By WILLIAM Vol. 2. Thermodynamics and Statistical Mechanics. , NEWS , Assistant Chemist required in large Factory near London. Write, stating experience and salary required. No man residing more than 10 miles away or already engaged on Government work need apply.-Address, Box 620, care of Mitchell and Co., Ltd., 1, Snow Hill, E.C. The THEORY of VALENCY. By J. NEWTON Chemist, young, ineligible, desires post in The FRIEND, D.Sc. SECOND EDITION, REVISED. 5s. net. MONOGRAPHS ON PHYSIOLOGY. Edited by ERNEST H. STARLING, M.D., D.Sc., F.R.S. INVOLUNTARY NERVOUS SYSTEM. By WALTER HOLBROOK GASKELL, M.A., M.D., F.R.S. With 8 Coloured Figures. 6s. net. [Just Published. MONOGRAPHS on BIOCHEMISTRY. Edited by R. H. A. PLIMMER, D.Sc., and F. G. HOPKINS, F.R.S., D.Sc. The RESPIRATORY EXCHANGE of ANIMALS and MAN. By AUGUST KROGH, Professor of Zoophysiology, University of Copenhagen. Royal 8vo. 6s. net. [Just Published. MONOGRAPHS ON PHYSICS. Chemical Laboratory. Good experience in Inorganic Analysis and in Organic Chemistry.-Address, C. Y., CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, E.C. FE EMALE LABORATORY ASSISTANTS. A number of Female Laboratory Assistants possessing a knowledge of quantitative chemical analysis is required by a large Munitions Factory.-Applications, giving full details of training and experience, and marked on the outside "L/A," to be forwarded to William Porteous and Co., Glasgow. No applications from those already employed on Government work will be considered. Lady, with Chemical Training, required by Firm In South London, to read for Research Work and summarise notes. Starting salary £140. Apply by letter only, stating age and experience. - Address," L 140," CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, E.C. Edited by Sir J. J. THOMSON, O.M., F.R.S., and FRANK HORTON, D.Sc. Lady, with Secretarial Training, required in The PHYSICAL PROPERTIES OF COLLOIDAL SOLUTIONS. By E. F. BURTON, B.A. (Cantab.), Ph.D. (Toronto). 8vo. 6s. net. [Just Published. Chemical Department of a Firm in South London. Starting salary £120. Apply by letter only, stating age and experience.Address, L 120, CHEMICAL NEWS Office, 16, Newcastle Street, Farringdon Street, London, E.C. MONOGRAPHS ON INORGANIC AND PHYSICAL MILLER'S ELEMENTS of CHEMISTRY, CHEMISTRY. Edited by ALEXANDER FINDLAY, M.A., Ph.D., D.Sc. The MOLECULAR VOLUMES of LIQUID CHEMICAL COMPOUNDS, from the POINT of VIEW of KOPP. By GERVAISE LE BAS, B.Sc. (Lond.). With Diagrams. 8vo. 7s. 6d. net. last Edition, in three volumes, in good condition, wanted to purchase.-Reply to "Chemist," care of W. J. Wilcox and Co., 26, Martins Lane, Cannon Street, London, E.C. MICA Telephone No. 2248 Avenue. F. WIGGINS & SONS, 102/8/4, Minories, London, E. MICA MERCHANTS, LONGMANS, GREEN, & CO., Manufacturers of Mica Goods for Electrical and ALL purposes 39, Paternoster Row, London, E.C. Contractors to His Majesty's Government. In all forms of the apparatus hitherto described the length of the manometer is determined by the extent of the heating apparatus, and in the most advantageous cases cannot be very great. It is, however, quite possible to have the manometer outside the heating apparatus (Fig. 6); then it may be made of any length (the longer the better), and an ordinary large beaker may be used as a heating vessel. Several cc. of mercury must be left over in the bulb to ensure that at t2° the capillary bore from the bulb to well outside the heating vessel is filled with mercury (otherwise if the bore of the capillary from the neck of the bulb to VaLc (273+13) Lc (273+13) al (273 +11) 1 (273+t1) (3) The pressure of the air only in the bulb at t20 (273+2)/(273+ti). = (4) The volume of the vapour at o° C. and 760 mm. pressure is 11160 w/d (assuming that I grm. of hydrogen at o° C. and 760 mm. pressure occupies 11160 cc.), and at to the volume is still V, and therefore the pressure is11160 X 760 w (273 +12) anywhere without the heating vessel be empty the vaporised substance will naturally tend to condense into the outside cooler region, and if the mercury thread extends from the neck of the bulb not well outside the heating vessel the air in the manometer will not be throughout at one temperature, viz., the room temperature), and also on account of the necessity of having this excess of mercury within the bulb substance whose vapours react upon mercury cannot be experimented upon with this form of apparatus, but if the apparatus have a very long spirally-twisted neck (Fig. 7) and the mercury thread be so arranged that it begins about a quarter to half of the length of the neck from the bulb then the apparatus can be so used. The necessary calculating formula for the vapour density with this apparatus is derived as follows : Let t3° C. be the external temperature when the reading / has been made, d = the vapour density, a = the area of the cross section of the capillary bore. (1) The initial internal pressure ti° = Results. = CH3.CO.CH3.-p=757 mm. ; L = 847 mm.; Lc 858 mm.; V=257'3 cc.; w=0'5116 grm. ; 1= 383 mm. ; t1 = 17° C.; t2 = 100° C.; t3 = 18.5° C.; vapour density (determined) = 29.2 (theoretical, 29'0). C6H6 (benzene).-p=763 mm. ; L=819 mm. ; Lc 832 mm.; V 246 2 cc.; w=0.7620 grm. ; 1= 329 mm.; t = 19.5° C.; t2 = 100° C.; t3 = 160 C.; vapour density (determined) 38.6 (theoretical, 39'0). If this form of apparatus be used in the quantitative analysis of a binary mixture it can easily be shown that the formula for calculating the percentage weights (w, and 100-w1) isd2wi+dı(100-w1) = = 100 didapLV [Lc (273 +13) -1 273 +t2)] 31068 w/Lc (273+1) (273+ (2) number of experiments, several devices were resorted to in order to use over and over again the same tubes; thus, the bulb at one end had a tube several inches in length, and just wide enough to admit the weighing tube (w), but which on cutting off the end to determine the volume V did not greatly reduce the bulb's capacity, and when this was entirely cut off after a number of determinations another similar tube was fused on; again, the end of the capillary tube was generally cut off only after every third or fourth experiment for cleaning purposes, the cleaning and drying being carried out as often as possible by aid of a long thin drawn-out "re-fill" (Fig. 8), and when too short was cut off altogether and another affixed. Also, the substance to be experimented on was enclosed (sealed) in previously weighed very thin drawn-out glass tube, and thus weighed introduced into the bulb with a heavy piece of glass-rod, and after the bulb was sealed and Lc was measured this extremely fragile tube was easily brokento liberate the enclosed body-by gently shaking the glassrod against it. From the illustrations one might think that only long bulbs were used; occasionally round bulbs were employed, made from ordinary small distilling flasks). Hackney Technical Institute, London, N.E. Russian Dyemaking Scheme.-A Petrograd report states that the scheme for establishing coal-tar colour works in Russia has not been abandoned. A project is under discussion which is said to have the support of the leading dyestuff consumers in the country, and it is thought to have very good prospects of success. The propose capital of the concern is 5,000,000 dols., the whole o which is to be provided by home consumers. It is state that the seven largest calico-printing companies in Russia (including the Baranoff Company, the well-known printer and Turkey-red dyers, and the E. Zündel Company Moscow), representing an aggregate capital of ove 25,000,000 dols., have together promised about 2,200,000 dols. toward the new undertaking. It is proposed to erec tar-distillling plants and manufacture right from the raw product.-Chemical Engineer, xxi., No. 3. THE PROPERTIES OF SOLID SOLUTIONS OF METALS AND OF INTER - METALLIC COMPOUNDS.* By F. C. THOMPSON, M.Met., B.Sc., Demonstrator in Metallography in the University of Sheffield. It is a profoundly interesting fact that almost without exception the alloys which are of industrial utility consist of one or more solid solutions. The brasses, almost the whole of the bronzes, all the copper-zincnickel alloys ("nickel-silvers") which find practical application, most of the coinage, and that most important branch of light aluminium alloy for aeroplane and motorcar construction, all consist entirely, or almost so, of such a constituent. Even in the case of the steels the all important property of hardening is due to the formation, and the more or less complete preservation on quenching, of a solid solution. The nickel and other special steels also which find so great employment in engineering construction owe their special properties, in part at least, to the improvement in the ferrite arising from the presence of the alloying metal in solid solution. The practical explanation of the overwhelming superiority of those alloys which consist of solid solutions when compared with the pure metals is found in the remarkable "toughness." or combination of strength and ductility, which is their dominating characteristic. Concerning the hardness bestowed on a metal by solution in it of another much has been written. Even now, however, the real explanation of the facts is by no means fully understood, so that in view of the supreme practical importance of the question it seems advisable to pursue it further. In addition to the hardeess of the alloys under consideration, another specially characteristic property is their high electrical resistance. Now, hardness and high electrical resistance are also characteristic of metals in the hard-worked state, and in general the physical properties of hard-worked metals and of metallic solid solutions are so much alike as to lead almost inevitably to the conclusion that some common cause is operating in each. In comparative tensile tests of a metal in the annealed and in the totally work-hardened states, the maximum stress has in the latter case been raised, and the elongation before fracture reduced until it has become inappreciable. These properties find their parallel in the solid solutions. The brass with 46 per cent of zinc has a maximum stress of 30 tons per square inch, that of copper being about 14 tons per square inch, while the elongation has fallen to under 5 per cent. The parallelism does not end here. It has been pointed out more than once that by cold-rolling the tenacity of a metal may be exactly doubled (Charpy, "Contributions à l'Etude des Alliages," Paris, 1901, p. 1). The tensile strengths of copper and of the copper alloy with 46 per cent of zinc, already given, show the same effect. In the silver-gold alloys -metals which form an uninterrupted series of solid solutions-the maximum sclerometric hardness is almost exactly double that of either pure metal, and similarly almost exactly equal to that of, for instance, silver which has been hammered dead hard. The results of electrical conductivity measurements point in the same direction. The addition of nickel to the copper-zinc alloys raises the specific resistance just as cold-rolling does, and to such an extent also that the further influence of deformation in the cold becomes negligible. A "nickel-silver" with 28 per cent of nickel, 12 per cent zinc, and 60 per cent copper has, when in the hard-worked state, a specific resistance of 41.5 microhms per cc., and after full annealing at Soo° C. a resistance of 41.3. The specific volume of a metal is influenced to only a very small extent by cold work. It is therefore of interest that, in the simplest cases, solid solutions in metals are * A Paper read before the Faraday Society, May 9, 1916. formed from the components without change of volume, the specific volume being a strictly lineal function of the concentration. Considering therefore the really remarkable parallelism between the properties of a metal in the "écroui" condition and when in solid solution with another metal, the assumption is perfectly justifiable that distortion of the crystalline matter is equally present in the latter case and in the former. From whence, then, does this internal strain to be assumed in the solid solutions arise? It is now a thoroughly established fact that the atoms of any crystal are arranged in space with perfect regularity. Taking in each atom a similar point, the space lattice on which the crystal is built is obtained. The unit of this lattice is therefore, in the case of the majority of metals, a cube whose side is, where v is the atomic volume. That metallic solid solutions are truly crystalline needs no labouring; cleavage slipping under stress, beautifully definite etching pits, single and repeated twinning, can' point to no other conclusion. In the space lattice of such a solution the lattices of the two or more constituents are regarded as interpenetrating. The conditions may be more easily followed, however, by considering one space lattice only. In the case of an element A to which a second B, which passes into solid solution, is added, the atoms of A in the space lattice are replaced progressively by those of the isomorphous B, and the lattice passes imperceptibly from that of the one pure element to that of the other. Now, this atomic replacement in the lattice must react on the physical properties. It implies, as Gossner has suggested (see Tutton, “Crystalline Structure and Chemical Constitution," 1910, p. 128), an attempt at the equalisation of the atomic volumes of the components in the solu. tion, and if the atom of A occupies a greater volume than that of B, the replacement must result in an expansion of B and a contraction of A in the act of crystallising together. The resulting distortion must give rise to elastic strains, probably of a high order, throughout the whole mass, to which the increased hardness of this class of alloy must be ascribed. The extent of the strain on this theory will depend, among other factors, on the extent of the atomic replacement, and should be greatest when half the atoms have been replaced by others. It is valuable evidence, therefore, in favour of the view propounded that in those series of binary alloys, such as gold-silver or gold-copper, which form an uninterrupted series of solutions the maximum hardness is invariably found when 50 per cent of each component is present. We may now consider in what way the hardness of an alloy may be expected to vary with composition, if the condition of internal strain which must arise if the conception be true of the equalisation of the atomic volumes of the constituents in such crystalline aggregates be the true cause of the hardening. If the solution of B and A be very dilute, and if the reaction on the atoms of A be distributed uniformly over all, the change in the volume of each atom will be small. On the other hand, since the atoms of B are few the change in volume in each will be great. The atoms of B therefore are compelled to assume practically the atomic volume of A. But as more of B is added the change of atomic volume, and hence the rate of increase of hardness with concentration, will become less and less. The additions of B already made will have produced a tendency for the mean atomic volume to approach more nearly that of the added metal, so that the increased strain produced by each further addition will become smaller. The theory of the increased hardness here propounded would therefore, in the simplest case where the component metals are completely isomorphous, lead to a curve for the relationship of the hardness of an alloy to composition which rises sharply at each end, as small quantities of each metal are dissolved in each other, and then rises less and less steeply, reaching a flattened maximum with the alloy containing 50 per cent of each metal. The parabolic curve obtained experimentally for such alloys, e.g., gold-silver, fulfils these requirements of the theory absolutely. It is, however, possible to predict still more closely the type of the hardness-concentration curve of a binary orseries of alloys of isomorphous metals. Consider an alloy containing x atomic per cent of an element A whose atomic volume is v1 dissolved in a metal B the volume of whose atom is v2. The tendency to equalise the volumes of the atoms of both A and B will increase that of B (assumed the smaller) and decrease that of A. Now, since there is no evidence of any marked change in the closeness of the packing of the atoms with alterations of concentration in such solutions, and since the greatest divergence of the specific volume of any of the alloys from that calculated according to a straight line relationship does not exceed 0:2 per cent, it may be assumed that the mean atomic volume of the crystalline solution will be the arithmetical mean, i.e., NEWS X 10 dynes per sq. cm. = 6× 10" dynes per sq. cm. 6 × 105 megabars. It is probable, however, that the hardness of a metal is determined by its intrinsic pressure. and the value of this constant for silver is 1.6 x 105 megabars. The hardness of the 50 per cent alloy of silver and gold is twice that of pure silver, whence the value for f of 6× 105 megabars may be considered as sufficiently satisfactory. The high value, however, of the applied stress which would produce the same increase of strain-energy as is produced by the changes of atomic volume may be explained on the assumption that the process of mutual crystallisation of the components of the solid solution does not result in exact equalisation of the two atomic volumes, but merely in the tendency towards equivalence. The actual strains would then be smaller than are here calculated, the more so the less exactly did the quality of the atomic volumes ensue. The result then would indicate a considerable divergence of the real conditions from those postulated. The theory here advanced to account for the characteristic properties of solid solutions has regarded them as being due to the distortion of the crystalline material resulting from the difference of atomic volume of the solvent and solute. Which of these has the larger atomic volume is, from the point of view of the theory, comparatively unimIn either case the result is the same-the alloy portant. is hardened. This is directly opposed to the conclusion arrived at by Roberts-Austen in 1888, that the effect of alloying one element with another depends on its atomic Gold volume (Phil. Trans., A, 1888, clxxix., [5], 339). In 1896, Arnold and Jefferson carried out further experi- alloy is increased. work on the same point in iron alloys, showed beyond Arnold (Journ. Iron and Steel Inst., 1894, 107), in other doubt that the atomic volume has no influence whatever on the physical properties of the resulting alloy, and that, so long as the alloying element passed entirely into solution, |