ALLOYS
327
than the lead-tin series given in Fig. 2. It represents, more- of pure gold itself. If the compound formed in the alloy over, a series in which the freezing-points are well within is dissociated, the portion of the curve which marks its range of the ordinary mer- presence will be rounded. curial thermometer, and of With regard to the goldwhich the complete freez- aluminium series, Heycock ing-point curve is exactly and Neville have shown that of the same order as that it presents a very compliwhich represents the freez- cated case, as not merely ing of a solution of salt, as one compound (AuA12), but Roberts-Austen has shown. several compounds of gold C o m Pos i-r i o r*, c O ro f»o © i T For alloys of which the and aluminium are formed. Fig. 4. Fig:. 2. melting-points are beyond In the third case, the two the range of the ordinary mercurial thermometer, some metals which are dissolved in each other form isomorphous form of pyrometer must be employed. Heycock and mixtures. In this group the freezing-points of the alloys Neville, in a very elaborate and prolonged series of of the series lie on a straight line, Fig. 4, as do the silverinvestigations on the freezing - point curves of alloys, gold alloys examined by Schertel. This question is very have used the platinum resistance pyrometer, and they important, and Gautier has given several instances of such determine the points at which alloys solidify or metallic isomorphous alloys, among w'hich that of antimony and constituents fall out of solution by noting the points bismuth may be cited. Metals do not appear to have been studied from the at which the electrical resistance (as measured by a Wheatstone Bridge) remains constant. Roberts-Austen, point of view of surfusion until 1880, wdien A. D. van on the other hand, has devised a recording pyrometer for Riemsdijk showed that gold and silver would Surfusion_ use with a thermo-junction. This instrument traces auto- both pass below their actual freezing-points graphically on a sensitized plate the cooling curve of any without becoming solid. Roberts-Austen pointed out that definite alloy, the halting places corresponding to the surfusion might be easily measured in metals and in alloj s arrests in the fall of the mercurial column in an ordinary by the sensitive method of recording pyrometry to thermometer being indicated by more or less horizontal which reference has already been made. He also showed portions. H. le Chatelier has thrown much light on the that the crossing of curves of solubility which had already question of the fusibility of alloys. He points out that been observed by H. le Chatelier and by Dahms in the in considering them as solutions it is necessary to modify case of salts could also be measured in the lead-tin alloys the view in which ordinary solutions are regarded. In (Fig. 2). The investigation of the mutual relations of ordinary chemical language, it is usual to distinguish the partially miscible liquids due to Alexejew, Konovalow, solvent from the body which is dissolved, but really the and to Duclaux was extended to alloys by Alder Wright. two bodies play the same part. Many aqueous solutions, The addition of a third metal will sometimes render the as he shows, can exist below the freezing-point of water : mixture of two other metals homogeneous. Heycock and that of chloride of calcium, for instance, which is liquid Neville proved that when one metal is alloyed with a small clown to - 55° C. He demonstrates that the typical curves quantity of some other metal the solidification obeys the which represent the fusibility of alloys have in each case law of Raoult. They also showed that the molecule of a corresponding curves in saline solutions, and in confirma- metal when in dilute solution often consists of one atom. tion quotes experiments by Loewel andBakhuis-Roozeboom, There are, however, numerous exceptions to this rule. In together with his own results, especially referring to the the case of aluminium dissolved in tin and mercury or freezing-point curves of isomorphous mixtures of certain bismuth in lead, it is at least probable that the molecules organic substances first recorded by Kuster. The result of in solution are AL, Hg.2, and Bi2 respectively, while tin Le Chatelier’s very interesting work is the division of the in lead appears to form a molecule of the type Sn4. Matthiessen in the paper already referred to showed that freezing-point curves of alloys into the following classes : In the first the metals dissolving each other give neither the electrical conductivity of all alloys may be represented definite compounds nor isomorphous mixtures. The curves graphically by three types of curves. Gold and Electrical of this class consist of two branches as is. shown by the silver alloyed afford an instance of the first of conductiyit ylead-tin curve in Fig. 2, in which composition and tem- ive types, there being a rapid decrease in the conductivity (or increase in the resistance) when either perature are co-ordinates, and it would appear that when a curve is composed of two branches only it is safe to silver is added to gold or gold to silver. The tinconclude that after solidification the two isolated metals copper alloys are, as regards electrical conductivity, types of the L-shaped curve. There is a rapid diminution of are simply in juxtaposition. conductivity when tin is added to copper, and at a certain In the second case, the two point there is an abrupt bend, and the curve proceeds metals give rise to one or more or less horizontally. More recent work has, however, more definite compounds. shown that there are certain breaks where the definite Here, as in the case of a compounds SnCu3 and SnCu4 occur. The conductivities solution of sulphate of soda of tin and lead are both low, and the lead-tin series is in water, one branch of the represented by a straight line connecting the two metals. curve is peculiar and charcompos.x.o* The conductivity curve of these alloyed metals represents acteristic. If the compound ^ 3. the third type. For the details reference must be made which is formed between the Matthiessen’s memoir. At low temperatures, Dewar two metals fuses without dissociation, there will be a maxi- to Fleming have shown that the conductivity of pure mum point in the curve which corresponds to the definite and metals increases greatly as the temperature is diminished. compound, and the curve will be of the form shown by Roberts-Austen’s gold-aluminium curve in Fig. 3. H. le Gold, in the purification of which elaborate precautions Chatelier has found similar curves for the copper-tin were taken, appears to present a very good instance. On the other hand, the conductivity of alloys does not increase alloys. In the case of the aluminium-gold senes the with diminution of temperature in the same manner, and a maximum point occurs, as Fig. 3 shows, at the °y a marked contrast between the behaviour of metals and AuAR the freezing-point of which is higher than that
