556 



NA TURE 



[February 13, 1902 



in the I to 8 percent, alloys, but dift'enng from them in nevei 

 occurrin;; hollow. On uxidalion it becomes very dark and is 

 easily distinguished from the other two constituents of the alloy. 

 In form it is plate like, and around it crystallises out the bright 

 constituent characteristic of the i to S per cent, alloys, either as 

 s rough envelope of fairly uniform thickness or as projecting 

 crystals. Stead was the first to draw attention to the fact that 

 the crystals of this division of the series were composite. As 

 the copper approaches 40 per cent., the plate-like crystals are 

 grouped together in parallel bunches. Casting masks the com- 

 posite character of the crystals, if, in the lower percentages, it 

 does not destroy it ; for under 20 per cent. Cu, the crystals can- 

 not be resolved into (wo components under high powers. 



/^\ to b\1 per cent. Copper.— M 41 percent. Cu the crystals 

 are small, lath-shaped, and arranged more or less in groups. 

 The alloy is brittle, and this brittleness increases with the per- 

 centage of copper. With each addition of copper, the groups of 

 composite crystals become more and more compact and the 

 amount of eutectic diminishes until at 56 per cent. Cu it dis- 

 appears (Stead, Journal of the Society of Chemical Industry, 

 June, 1897), and the bright constituent of the crystals forms the 

 groundmass. When 617 per cent. Cu is reached, the bright 

 constituent disappears and we have a homogeneous mass of 

 SnCuj, probably a definite compound. Casting tends to harden 

 and toughen these alloys. Seeing that these alloys up to about 

 56 per cent. Cu show four breaks in their cooling curves, one 

 would naturally expect to find four different constituents in 

 each. Only three, however, can be distinguished, fjuenching 

 below the first and second breaks gives a difference in structure 

 only. As in the alloys containing 617 per cent. Cu and 

 onwards, branch e of the curve (Fig. i) corresponds to a re- 

 arrangement in the solid, and as the difference between the 40 

 and 41 per cent. Cu alloys is one of structure only, we may 

 assume that the second retardation in the cooling curve (.) is 

 one of rearrangement also. 



617 to (&■■& fcruiit. Copper, SnCu.^to SnCUi. — The changes 

 which take place between these two points can only be observed 

 when the alloys are very slowly cooled. The alloys set as a 

 whole at the first break and tend to rearrange themselves sub- 

 .secjuently in the solid, on branch e (Fig. l). Each addition of 

 copper to SnCu3 brings in more and more of a bright con- 

 .stituent, probably SnCuj. (Quenching and casting produces 

 structures entirely new. Figs. 2-5 show the 66 per cent. 

 Cu alloy differently cooled. Fig. 2 was quenched on the first 

 break. There is a cell-like structure with light-coloured walls 

 or boundaries. In places the change has gone further, and we 

 get the fine cross-hutching characteristic of Fig. 3, which has 

 been quenched below the first break. The cell-like struclure 

 has entirely disappeared. Fig. 4 has been quenched below the 

 second break and resembles a slowly-cooled alloy, except that 

 in the latter there are distinct traces of a eutectic structure. 

 I'ig. 5 has been cast on an iron plate, and the "schiller" structure 

 is well developed. At 68-2 per cent. Cu the alloy is homo- 

 geneous, has a conchoidal fracture and is extremely brittle. 



6S-2ii to 7$ per eciit. Co/t/^r.— Immediately the copper is in- 

 creased above 68 3 per cent., the second eutectic makes its 

 appearance. As the copper increases, the grains of SnCuj split 

 up into bright veins and dendrites, surrounded by the eutectic. 

 The veins and dendrites decrease and disappear at 75 per cent. 

 Cu, where the mass is made up entirely of the eutectic. The 

 alloys are best studied when furnace-cooled ; their surfaces 

 above 71 per cent. Cu are seen to consist of a network 

 of dendrites or skeleton crystals resembling those seen on the 

 surface of a pure metal. This surface structure continues right 

 up to the copper end of the series. It was .soon noticed that 

 the internal structure of the alloys from 70 to 75 per cent. Cu 

 showed no trace of these dendrites, and so the surfaces of several 

 were rubbed down and polished. In each case their structure 

 was the same as that of the centre of the alloy, which shows 

 that these dendrites have split up and rearranged themselves 

 after .solidification, and all that remains of them is this surface 

 .structure. Casting makes the structure very minute, and about 

 73 per cent. Cu traces of the skeleton crystals can be seen in 

 the centre of the ingot. They appear dark and structureless, 

 as if they had been unable to resolve themselves into their 

 two constituents. 



75 to 100 per cent. Copper.— V^'hen 76 per cent, copper is 

 present, two new constituents make their appearance and the 

 alloy assumes a yellow lint. In section we find yellow grains, 

 surrounded by a bright white border, set in the second eutectic, 



NO. 1685, VOL. 65] 



in which small bright white grains also occur. As the total 

 copper is increased, the yellow grains increase, forming den- 

 drites and skeleton crystals, the white borders and grains 

 merge together and the eutectic decreases till at about 90 per 

 cent, it disappears. The yellow grains become darker and 

 darker (contain less and less tin in solid solution). The 

 light borders diminish and disappear, about 95 per cent, 

 leaving copper dendrites behind. These dendrites vary in 

 composition from centre to outside, and so the centre etches a 

 darker colour. They darken with increase of copper until 100 

 per cent, is reached. Casting tends to make the copper 

 grains solidify, containing a considerable quantity of tin. In 

 this way the eutectic can be made to disappear considerably 

 below 90 per cent Cu. Quenching shows that (he upper break 

 corresponds to the solidification of the copper ; break 2 to the 

 solidification of the groundmass which splits up into a eutectic 

 when branch e is reached. Fig. 6 contains 80 per cent. Cu 

 furnace-cooled, whilst Fig. 7 shows the surface of the .same and 

 also that with this percentage of copper the dendrites of copper 

 have directed the formation of the surface skeletons. Fig. 8 

 is the same alloy quenched below first break. The dendrites of 

 copper are seen set in a structureless matrix. Fig. 9 is the same 

 alloy quenched below the second break. The dendrites of copper 

 (light, because of a different etching process) are seen, set in a 

 fibrous matrix — the eutectic of which the formation has been 

 faced. Fig. 10 shows the same alloy cast. As its appearance 

 would indicate, the alloy is very tough and cuts well. 



It seems clear then that branch e of the cooling curve is one 

 of change in the solid, and this conclusion has been proved 

 beyond doubt by the beautiful work of Heycock and Neville 

 published by the Koyal Society. When one considers the 

 many and distinct different stiuctures in the series produced 

 by quenching at different temperatures and by rehe-iiting and 

 then quenching, it is quite evident that the changes which take 

 place during the cooling of an alloy of copper and tin, especially 

 in the neighbourhood of the second eutectic, are even more 

 numerous than those of the carbon-irons. 



UNIVERSITY AND EDUCATIONAL 

 INTELLIGENCE. 



Camisridof.. — The Allen studentship, value 250/. for one 

 year, for research in connection with medicine, mathematics, 

 physics and chemistry, biology and geology, or moral science, 

 will be filled up at the end of the present term. It is open to 

 graduates under the age of twenty-eight on January 8. 



Principal K. II. Griffiths, F.R.S., of Cardiff University 

 College, has been approved for the degree of Doctor of Science. 



The Rede lecture will be delivered next term by Prof. 

 Osborne Reynolds, F.K.S., of Owens College, Manchester. 



Mr. W. N. Shaw, F. R.S., will give three lectures, on 

 February 13, 20 and 27, on the physics of the ventilation of 

 buildings. 



Prof. Tilden, F.R.S. , has been appointed an elector to the 

 chair of chemislry ; Lord Rayleigh, F.R.S. , an elector to the 

 chairs of chemistry and of mechanism ; Dr. Hill, to the anatomy 

 chair ; Mr. F. Darwin, F.R.S., to the botanychair ; Dr. Ilinde, 

 F.R.S., to the geology chair (Woodwardian) ; Sir G. G. Stokes, 

 F'.R.S., to the Jacksonian and Cavendish chairs; Dr.D. M.ac- 

 Alister, to the Downing chair of medicine ; Dr. Hugo Miiller, 

 F.R.S., to the chair of mineralogy; Prof. E. Ray Lankesier, 

 F.R.S., to the chair of zoology and comparative anatomy; 

 Prof. McKendrick, F.R.S., to the chair of physiology; Lord 

 Lister, F.R.S., to the chair of pathology; and Prof. Marshall 

 Ward, F.R.S., to the chair of agriculture. 



Dr. J. Reynolds Green, F.R.S., has been elected to a fellow- 

 ship at Downing College. 



TnK University College, Hristol, does not receive the generous 

 support given to similar colleges elsewhere, but the report of the 

 council lor the session 1900-01 shows that much valuable 

 woik has been done in spite of limited means and opportunities. 

 Important papers have been published by various members of 

 ihe scientific staff and others are in progress. The clinical and 

 bacteriological research laboratory, which has been at work 

 under Prof. Stanley Kent for little more than a year, has, among 

 other matters, betn able to afford valuable aid to the Medical 

 Officer of Health in reporting upon the presence of plague in- 



