October 3, 1907J 



NA TURE 



573 



without any exception over a wide range of substances, 

 it appears justifiable to conclude that they are univer- 

 sally aoplicable. The subject is therefore a very 

 extensive one, and the immediate researches which 

 arc dealt with here refer to only a small corner of a 

 very wide field. 



In the light of present knowledge it would now 

 seem as if the phenomena of the hard and soft states 

 are so striking that they might have been expected to 

 stimulate inquiry into their true meaning at a much 

 earlier date. One of the most obvious of these pheno- 

 mena has been perfectly familiar ever since metals 

 were first drawn into wire — that is, that the tenacity 

 of the metal is enormously increased by the operation. 

 By the simple operation of wire-drawing, the power 

 of pure soft iron to resist stretching is raised from 

 twenty tons per square inch to more than eighty tons. 

 Recent researches with metals in the highest state of 

 purity have shown that the resistance of gold to 

 stretching may be raised from 42 tons per square inch 

 to more than 14 tons, while silver and copper are 

 affected to an even greater extent. 



Until very recently the adjective " crystalline," when 

 applied to a metal, at once suggested hardness and 

 brittleness, and even yet among practical metallurgists 

 this association of ideas is not easily got rid of. It 

 is no paradox, however, to say that in the pure duc- 

 tile metals the crystalline state is actually the soft 

 state. In what follows it will be shown that a very 

 large part of this softness is directly due to the in- 

 stability of the crystalline structure. Conversely, the 

 non-crystalline or amorphous state is the more stable 

 niechanicaUy, and is therefore the harder. Not only 

 the softness, but also the malleability and ductility of 

 a metal, largely depend on its crystalline condition. 

 When the metal is mechanically worked, as by ham- 

 mering or rolling it into sheets or bars, or by drawing 

 it through dies into rods or wires, some of the crystal- 

 line is broken down and passes into the non-crystalline 

 form, and as the metal thereby becomes harder it is 

 also reduced to a lower condition of malleability and 

 ductility. 



It has been concluded from a long series of experi- 

 mental observations that in the passage from the 

 crystalline to the non-crystalline state there is an inter- 

 mediate stage during which the molecules have the 

 freedom and mobility of the liquid state, and that the 

 amorphous state results from the sudden congealing 

 of this mobile phase. It is well known that when a 

 substance passes from the liquid to the solid state, 

 time is required for the molecules to marshal them- 

 selves in the orderly formation which is the essential 

 feature of the crystalline state. If a liquid can be 

 congealed with sufficient suddenness, the solid which 

 results is non-crystalline or amorphous, e.g. glass, 

 sugar-candy, &c. If we were able to see the actual 

 molecules we may imagine that the amorphous solid 

 would present the appearance of an instantaneous 

 photograph of a liquid in which the molecules would 

 appear as if transfixed in the midst of their rapid 

 movements. It follows from the above that if it were 

 possible in a mass of metal simultaneously to break 

 down all the crystalline units of structure with suffi- 

 cient suddenness, the whole mass would for an instant 

 be in the liquid condition, and on re-solidification 

 would appear in the non-crystalline state. A little 

 consideration, however, will make it plain that these 

 conditions cannot be fulfilled in the ordinary 

 mechanical operations on metals in the solid state. 

 In a mass of metal, any stresses which are applied 

 mechanicallv must be applied from the outside, and 

 can onlv reach an internal point or surface after pass- 

 ing- through all the intervening layers. It follows 

 that the breaking down and " flow " of the crystalline 

 elements must take place step by step, so that the 



NO. 1979, VOL. 76] 



mobile condition occurs at successive surfaces within 

 the mass. A wave of mobility can in certain cases be 

 seen as it passes along a stressed rod, but it is in- 

 stantly followed by a wave of congelation which leaves 

 the metal behind it in a harder and more resistant 

 condition. In some cases a second wave of mobility 

 may be started by the application of a greater stress, 

 but as a rule each successive application of a uni- 

 formly increasing stress produces less and less effect. 

 The portions of metal which have yielded and flowed 

 and again congealed protect those portions which 

 still retain their crystalline structure. There appear 

 to be good grounds for believing that even in a gi i 

 leaf, in which the metal has been beaten to a thick- 

 ness of only 1/280,000 of an inch, there are still 

 minute units in the crystalline state which have 

 escaned destruction owing to the protective action of 

 the harder, non-crystalline metal in wh'ich they arc 

 embedded. Gold wires which have been drawn 

 through a wire plate until they are fifteen times their 

 original length show a microstructure in which de- 

 formed and broken down crystals are embedded in 

 non-crystalline substance. The hardened metal is a 

 complex structure built up of crystalline and non- 

 crystalline substance ; in studying its properties, 

 therefore, it is necessary to remember that no speci- 

 men, however drastic may have been its mechanical 

 treatment, can be entirely in the non-crystalline condi- 

 tion. 



Though an increase of hardness and tenacity is a 

 very conspicuous feature of the change from the one 

 state to the other, it is only one among a number of 

 equally definite indications of change. .\ comparison 

 of the heat of solution of a metal in the two states 

 shows that the molecular energy stored in the non- 

 crystalline form is greater than in the crystalline. In 

 this case the difference in solubility which results 

 fiom this greater energy is further accentuated as the 

 two phases of the metal act towards the solvent as a 

 galvanic couple. A thermoelectric couple made by 

 twisting together the ends of wires in the hard and 

 soft condition is affected by changes of temperature 

 in the same way as a couple made of two different 

 inetals would be. In the case of silver a thermo- 

 couple of this description can develop an e.m.f. of 

 27 microvolts for a temperature difference of 83°. In 

 all these cases the single chemically-pure metal 

 behaves like two distinct metals. 



When hardened metal is heated to a certain tem- 

 perature, its softness is completely restored. The 

 microscope shows that when this occurs complete crys- 

 talline rearrangement has also taken place. The 

 micrographs, Figs, i and 2, from the paper by Mr. 

 Beilby read before the Royal Society on June 27, 

 show the two types of structure, the hard and the soft. 

 In Fig. I the original crystalline grains have been 

 completely broken down and destroyed by wire- 

 drawing, giving place to masses of deformed 

 and shattered crystal units cemented or concreted 

 together bv that part of the metal which has 

 flowed and congealed. In Fig. 2 a new crys- 

 talline structure has been developed by heat, 

 and all traces of the other structure have disappeared. 

 This re-crystallisation in hardened metal occurs at a 

 temperature far below the melting point of the metal. 

 In gold the re-crystallisation temperature is about 280°, 

 while its melting point is ioSo° ; this profound change 

 of structure, therefore, occurs Soo° below the melting 

 point. 



In the crystalline state the molecules are disposed in 

 sheets or lamellse of uniform orientation, like soldiers 

 in a battalion. In the liquid state the molecules are 

 in free movement; they do not maintain fixed positions 

 with respect to each other. The effect of sudden con- 



