834 



NATURE 



8, 19: 



(.1 



x\m\ thfr presence of « suffiident proportion of niclcH 

 the iron in the facc-ccntrcd «ul.i 

 1 phane at a temperature ui u 



iron would have reverted 

 ! ihf alpha phase. Presum- 



n llir ^'aiiiiiia < undition beCftUSe 



1 II lain a larj^cr pru{K)rtion of 



, and because this arrange- 



. .w.r.r.- fl...i, ..i,« .Itj.rn .tive. 



ice- 



!■: thr 



ably tlic II' 



'\\\ \\v.\\ I I .1 



1 n 1 



all 



I H 



'II 



liiii i.iMir w It hunt undue Uisiui'ii 11. im irunslomied 

 iiiti) aiKulur latticf — Still essentialU a lattice of copper 

 — but capable of carrying a larui r number of zinc 

 atoms, and that at a still higher < oik i utration of zinc 

 a further tnndit'u ation of the lattiie might occur. 

 Each suet ( ■'si\ (• inodification would, in such a case, 

 be expected to show an imrtasing approxiinatidn to 

 the hexagonal lattice of /.ini: iusclf. Actual dcicrniina- 

 tions of the lattices of the beta and gamma phases 

 of the copper-zinc system, made by Owen and Preston 

 in consequence of this suggestion, have completely 

 verified it. The two phases show no compound 

 lattice, but a modified copper latti( < . 



^'Vnother point of some interest in tlie distinction 

 between solid solutions and compounds may be briefly 

 considered. A good deal of consideration has been 

 given by Tammann ^ to the chemical properties of 

 solid solutions, Tammann's work being based on the 

 idea of a substitution structure, but without reference 

 to lattice (li^toition. If this principle is applied, 

 however, an interesting conclusion may be drawn, to 

 which attention was first directed by Mr. Preston, in 

 regard to what may be termed " symmetrical " solid 

 solutions. In any alloy system forming a continuous 

 series of solid solutions between two metals, alloys 

 must occur in which the two kinds of atoms are present 

 in some simple ratio such as one to one, two to one, 

 three to one, etc. According to the particular nature 

 of the latti( r system in each case, some of these simple 

 ratios will allow the atoms to arrange themselves in 

 a perfectly symmetrical manner. Such perfectly sym- 

 metrical atomic arrangement, however, is not likely 

 to be attained or approached except in specially favour- 

 able circumstances. Ver>' gradual cooling from fusion 

 and a considerable rate of diffusion are essentials, 

 but there is the further condition that the symmetrical 

 arrangement in question should be a simple one. Thus 

 in any lattice, an arrangement in which alternate 



* Tammann, ZtUxhr.f. Anorg. u. AUgtm. Ckemit, July 1919. 



layers or planits of atoms con»i«t eafh exclusively 

 ould seen uch a simi 



face-coit; lattice a oni 



to-«in i.itiu iillowb of such an aiiaii-tnient, all 

 atonic ai the cube comers and t!. • at the cent 



of two opposite faces l)cing o< 

 atom and the remaining foi 

 second kind of atom. In .si 

 simple symmetrical ar*^ •" " " • 

 formed by the proc« 

 prr-rnt throrv. is tli.ii m •.■.^•ji .u 

 .111 .)i(iii)i((l 1)\ one kind of alotii 



one kind 



itrrn liv th 



jiiired by the 



• farc-rcntrcs 

 all the <u!.e 



corni In 1i\ the oiIki . ' !'ly'"g «in atomic ratio 



of tlin T til niu-. 



f 'ti ial interest which attaches to such • 



; H al arrangements is that, if fully 

 there will be in .such a lattice a perfectly unin ni 

 atomic spacing. The consequence must be, if the 

 present theory is correct, a single melting-point and 

 relatively low hardness and electrical resistivity. Jn 

 some alloy systems, this state of affairs is so cIom ly 

 approached that it becomes plainly \asible on t!ie 

 equilibrium diagram as experimentally determined, and 

 the presence of a compound at the simple at«iT:,i< 

 ratio in question has sometimes been inferred — wrongly, 

 according to the prcs* nt \ ii w. In other system-, 

 where diffusion is slow and uniform L'eometri' .il 

 arraiiLidTunt, therefore, is practically unattainable, the 

 ideal i (nidition is never reached expcrimcntallv. ' ;t 

 the " solidus " curve shows an infl' 



"liquidus" in the neighbourhood of 1 1.- .. 



composition — in several examples near the one-to-cne 

 ratio of atomic concentration. In some alloy - < • 

 two such inflexions, corresponding approxii: 

 two such ratios, have been observed. Althoi 

 inflexions have become increasingly definit' 

 best - determined diagrams, so that they couid nol 

 be ascribed to experimental error, no explanation 

 as yet Ix < ;i offered. The fact that these details 

 only to 1h seen clearly in recent diagrams, prej 

 by methods of extremely slow cooling of the alio] 

 tallies well with the requirements of our theor>'. 



Inferences from the substitution and lattice-distor<^ 

 tion theory could be pursued at much greater lengthj 

 so far, no failure of such an inference, when t( 

 by means either of older well-established fact or 

 special experiments, has yet been found. There 

 thus some hope that a small but real step has 

 taken towards the better understanding of the nat 

 of alloys, and particularlv of solid solutions and int€ 

 metallic compounds. 



Weather Influences in the British Isles. 

 By C. E. P. Brooks. 



'X'HE sun is the only source of terrestrial weather 

 -*- in the sense that the difference between the 

 amount of solar radiation received in different latitudes 

 is the driving force of the atmospheric circulation. The 

 complexity of the earth's surface combined with its 

 rotation about an axis introduces corresponding rom- 

 plexities into this circulation, but H. H. Chivton ron- 



siders that " if there were no variation in solar radiation j tion 

 NO. 2823, VOL. 112] 



the atmospheric motions would estabi al 



system with exchanges of air between equator ana ; 

 and between ocean and land, in which the only va 

 tions would he daily and annual rhamres set in opera- 

 tion by tlie relative motions of earth and sun ; the 

 existing ohangis \vr call weather have their origin 

 hitrtv. it not entirt.lv. in the variation of solar radia- 



It has been found that some parts of the 



