CRYSTALLOGRAPHY 557 



Numerous investigations of the structure of individual 

 substances have recently been made, and the time seems not 

 far distant when most of the elements and simpler chemical 

 compounds will have been examined. According to A, W. 

 Hull {Phys. Rev., 18, 88, 1921), the hexagonal elements, titanium, 

 zirconium, cerium, and osmium, each consist of two triangular 

 prismatic lattices with dimensions 2-97A and 4-72A, 3-23A 

 and 5-I4A, 3-65A and 5-96A, and 2-71 A and 4-32 A respectively. 

 It is possible, however, that cerium may be more complex. 

 Thorium is composed of a face-centred cube with edge 5 -04 A. 

 Lithium is described by J. M. Bijvoet and A. Karssen [Proc. 

 Acad. Amsterdam, 23, 1365, 1922) as belonging to the centred 

 cube type with edge 3 -50 A, and beryllium by L. W. McKeehan 

 {Proc. Nat. Acad. Set., 8, 270, 1922) as consisting of two 

 interpenetrating hexagonal lattices. The former authors have 

 also examined lithium hydride (Proc. Acad. Amsterdam, 25, 26, 

 1922). According to the summary by P. Niggli {loc. cit.), and 

 a paper by V. M. Goldschmidt {Zeit. Metallknnde, 13, 449, 

 192 1 ), eighteen elements are known to crystallise in the face- 

 centred cube lattice, six in the centred cube lattice, four in 

 lattices of the diamond type, ten in the hexagonal type, four 

 in rhombohedral lattices, and one in a tetragonal lattice. 



An interesting investigation of some of the compounds 

 present in iron and steel has been made by A. Westgren and 

 A. S. Lindh {Zeit. phys. Chem., 98, 121, 1921). They confirm 

 the results obtained by A. W. Hull {Phys. Rev., 9, 94, 191 7 ; 

 10, 661, 191 7) that a~ and /3-iron have essentially the same 

 structure, both belonging to the centred cube type, while they 

 also show that both V-iron and austenite belong to the face- 

 centred type. Martensite, on the other hand, gives the struc- 

 ture of a-iron, and must be largely composed of it, or of /S-iron. 

 In a further paper, A. Westgren and G. Phragmen {Zeit. phys. 

 Chem,., 102, i, 1922) describe the structure of cementite as 

 orthorhombic, there being four molecules of FcgC in the 

 elementary parallelepiped, which has edges of length 4- 5 3 A, 

 5-1 1 A, 6- 77 A. These results confirm the generally accepted 

 metallurgical opinion that martensite is very largely yS-iron and 

 that austenite is a solid solution of 7-iron and carbon. The 

 question of the form of the carbon in the latter is interesting. 

 If, as seems probable, the carbon is present as the iron carbide, 

 cementite, then it is feasible to assume that the temperatures 

 at which the austenite solid solution is formed, the lattice of 

 cementite is pseudo-cubic. The structures of some metallic 

 solid solutions have been investigated by E. C. Bain {Trans. 

 Inst. Min. Met. Eng. Bull., 11 39, 1922). In the case of the 

 system nickel-chromium, which shows a continuous series of 

 solid solutions, the alloys containing up to 65 per cent, have 



