6i8 



NA TURE 



[April 27, 1893 



a mechanical sifting of the elements, iln fmaking these experi- 

 ments, a cup-like receptacle was used, which had a vertical 

 partition of sheet glass placed in it, so that the cup was divided 

 into two parts by the glass. Sodium amalgam was placed on 

 one side and pure mercury on the other ; the whole was then 

 heated to a temperature of 200° C, at which the glass became 

 slightly conducting. By the aid of a battery, the sodium atoms 

 of the sodium silicate were set in motion, and after 30 hours 

 it was found that a considerable quantity of sodium, amounting 

 to 0"05 gramme, had passed into the mercury which was 

 originally pure. A corresponding amount of sodium had been 

 lost by the amalgam, but the glass had exactly preserved its 

 original weight and clearness. The glass was partly composed 

 of neutral molecules of sodium silicate, together with free 

 molecules both of sodium (base) and of the acid, and of the free 

 sodium capable of being transported under the influence of the 

 electric current. When Tegelmeier replaced the sodium 

 amalgam by lithium amalgam and repeated the experiment, the 

 sodium of the glass passed as before into the originally pure 

 mercury, and the glass became opaque on the side touching the 

 lithium amalgam ; but after a time the opacity extended right 

 through the thickness of the glass, and the metallic lithium 

 began to accumulate in the previously pure mercury. It is 

 not possible thus to chase out all the sodium present in 

 the glass ; but the free sodium atoms are replaced by 

 those of lithium. Analysis showed that the glass originally 

 contained 2*4 per cent, of potassium and 13 "i per cent, of 

 sodium ; but after the experiment, while retaining the same per- 

 centage of potassium, it had only 4^3 per cent, of lithium, and 

 only 5*3 per cent, of sodium. The glass in which lithium had 

 thus replaced part of the sodinm was very tender, opaque, and 

 friable. The conclusion to be drawn is that the atoms ol 

 lithium, having an atomic weight of 7, and an atomic volume 

 of 1 5 '98, can pass along the tracks, or molecular galleries left 

 in the glass by the sodium atoms, the atomic weight and volume 

 of which are 23 and 16 04 respectively. When a metal of 

 superior atomic weight and volume to sodium was substituted for 

 the lithium — such as potassium, with atomic weight 39 and 

 atomic volume 24 — it was found not possible to chase out the 

 sodium, the new atoms being too big to pass along through the 

 spaces where the sodium had been. We are thus confronted 

 with a molecular porosity which can in a sense be gauged ; 

 and the mechanical influence of the volume of the atom 

 is made evident. Proceeding to the details of the experi- 

 ments made by the committee, the influence of im- 

 purities on copper was next referred to. The question was 

 raised whether normal copper can be made to assume an 

 allotropic state, analogous to that in which there is reason to 

 believe iron can exist, and if so are the properties of normal and 

 of allotropic copper as widely difi'erent from each other as those 

 of the distinct varieties of certain well-known non-metallic 

 elements. The point is one of considerable interest, and Prof. 

 Roberts-Austen seems to have little doubt that copper can be 

 prepared by electrolytic deposition, in an allotropic state, in 

 which the density of the metal is from 80 to 82 as compared 

 with 8 "92, which is that of normal copper. The effect of 

 mechanical and thermal treatment upon copper was then referred 

 to, and some interesting figures were given, showing how 

 different may be the properties of a metal chemically pure ; for 

 instance, rods of very pure electrolytic copper, all the same sample, 

 but variously treated, broke under stresses varying between 

 8"2I9 tons and 18,750 tons to the square inch ; the former 

 being the tensile strength of cast rods, and the latter of cast 

 rods worked and not annealed ; whilst cast rods carefully 

 worked, and annealed gave a tensile strength of 18-259 per 

 square inch. The experiments show a difficulty in determining 

 a standard tenacity for copper. 



The effects of arsenic, bismuth, and nickel upon copper 

 afforded one of the most interesting parts of the investigation, 

 and from the engineer's point of view an extremely im- 

 portant section of the series of experiments. It has been too 

 often accepted as a matter of fact that pure copper is the best 

 that can be used for engineering purposes, and specifications 

 are generally framed to this effect. The Research Committee, 

 however, show that the metal may be, and frequently is, as a matter 

 of practical fact, too pure for the purpose ; thus, it has been 

 found that a very fair percentage of arsenic improves the copper 

 used in fire boxes of locomotives. It is well known that of old 

 these parts of the boiler lasted for a much longer period of time 

 than they do in the present day. In fact, as Mr, Tomlinson, an 



old railway engineer, said in the discussion, they used to expect 

 to get half a million miles of running out of a copper fire box, 

 whereas about half that distance is all that is obtained in the 

 present day. This he attributes to the effect of electrical matters 

 upon engineering practice. The electricians insist on their 

 copper being absolutely pure, and that has raised the standard, 

 so that now the copper smelters get all the impurities out of the 

 metal, whereas in old times a considerable percentage of alloy, 

 especially arsenic, was present. Antimony appears to behave 

 like arsenic, and when present in proper proportion greatly 

 strengthens the copper. Bismuth, on the other hand, renders 

 copper singularly weak. Witho'l percent, of bismuth a sample 

 of copper was too brittle to work, and had at the ordinary tem- 

 perature a tenacity of 18,000 tons to the square inch ; but at a 

 higher temperature the fall in tenacity was very rapid, and 

 there was practically no elongation. The prejudicial effects 

 of bismuth did not seem to disappear, even though but a trace 

 were present. In one test of a singularly pure copper, con- 

 taining only o'oo2 per cent, of bismuth, although the metal was 

 strong and worked well, the elongation was very small. The 

 variation in the effect of arsenic and antimony on the one hand, 

 and of bismuth on the other, is of considerable interest, for ac- 

 cording to the classification of Mendeleeff, arsenic, antimony, 

 and bismuth all belong to the same family, of which nitrogen is 

 a type. The atomic volume of bismuth — 20 9 — is, however, 

 higher than that of arsenic — 13*2, or of antimony— 17 '9, and 

 therefore, according to the principle laid down by Prof. Roberts- 

 Austen, bismuth ought to diminish the tenacity of copper, of 

 which the atomic volume is only 7"i. But in accordance with this 

 reasoning the influence of arsenic and antimony should be exerted 

 in the same direction, even though in a less degree. The author 

 has turned his attention to this matter, and has already been 

 conducting a series of experiments which have extended over 

 nearly twelve months. The investigations are, we believe, not yet 

 complete, but the results will be given subsequently. A diagram 

 was, however, exhibited at the meeting, in which curves were 

 shown, illustrating the behaviour of various alloys of copper and 

 bismuth during cooling, and the wholly unexpected fact was re- 

 vealed that the copper passed below the freezing point before it 

 actually became solid. On each curve there was a second or lower 

 point of solidification, which occurred at a constant temperature 

 in all the alloys, and wasvery close to the melting- point of bismuth 

 itself The existence of this second point was very evident, 

 even when the copper contained only one per cent, of bismuth, 

 and this fact goes far to explain the peculiar action of bis- 

 muth on copper. It would appear that whether very poor 

 or very rich in bismuth, the alloy of copper may be a portion 

 of bismuth, containing perhaps a little copper, always re- 

 mains fluid until the temperature of the mass has fallen to 

 260° C, which is the point at which bismuth tself solidifies. 

 The presence. Prof. Roberts-Austen stated, of a fluid con- 

 stituent in an alloy long after the mass itself had become 

 solid, is doubtless the determining cause which enables the 

 metal to assume a highly crystalline, and consequently 

 an intensely brittle structure. So far as he was aware the 

 cause of the peculiar behaviour of bismuth could not have been 

 revealed by any other method of investigation than the one 

 adopted. In connection with this point, a fact brought forward 

 during the discussion by Mr. Gowland, is of interest. In the 

 course of his metallurgical work at the Japan mint, he had 

 brought before him a large number of bars of silver for the 

 purpose of coining, but they were so brittle that it was impos- 

 sitile to work them at all. On investigation he found that there 

 was an appreciable quantity of bismuth in the silver. The 

 structure was coarsely crystalline, and though the whole mass 

 was so hard and brittle, the crystals themselves were very 

 ductile. The conclusion he came to at the time was that the 

 crystals of silver had become separated, as it were, by a film 

 of bismuth. The fact bears out the correctness of Prof. Roberts- 

 Austen's mode of reasoning. Judging from their polished sur- 

 faces, the alloys of copper rich in bismuth are to all appearances 

 as coherent as the alloys of copper and tin, which have great 

 strength. The report gives some interesting particulars of the 

 effect of pressure. The passage of iron from one allotropic 

 modification to another is accompanied not only by a change of 

 heat capacity, but also by a change of volume. This matter was 

 referred to in the previous report, but the author gave some further 

 interesting particulars of experiments carried out by compres- 

 sing a piece of steel in a hydraulic press, in order to obtain 

 recalescence at a lower temperature than would be the case if the 



NO. 1226, VOL. 47] 



