4 o SCIENCE PROGRESS 



seems impossible to assign any one chlorine atom to a particular 

 sodium atom to form a molecule in the ordinary chemical sense. 

 The structure ascribed to zircon, however, indicates that in the 

 solid state the formula should be Zr0 2 Si0 2 and not ZrSi0 4 , 

 while rutile and cassiterite should be (Ti0 2 )2 and (Sn0 2 )2 

 respectively. Xenotime, in the solid state, should be repre- 

 sented by Y0 2 P0 2 and not YP0 4 . Vegard explains this depar- 

 ture from the usual formula by assuming that the constitu- 

 tional of a substance varies with the state, and that the structure 

 of the solid form cannot be expressed by the ordinary constitu- 

 tional formula. Hence xenotime, despite its structure, may be 

 regarded as an orthophosphate and rutile as an oxide or as a 

 titanate. 



Fock (Cent. f. Min. 191 6, 392), on the other hand, argues 

 that chemical molecules do exist in the solid state, and regards 

 the atomic conception of crystal structure as erroneous. Rinne 

 (Zeit. Anorg. Chem. 96, 317, 1916) also subscribes to this view 

 and holds that Bragg 's work shows the existence of groups of 

 atoms linked together to form molecular complexes. The 

 affinities of an atom may be divided so that, for example, in 

 rock-salt the single valency of each sodium atom may be par- 

 titioned among six chlorine atoms and in the cubic form of 

 zinc sulphide the two valencies of zinc among the four sulphur 

 atoms which are equidistant from the zinc atom. 



In a long theoretical paper on the structure of crystals 

 (Neues Jahrb. f. Min. 1916, ii. 47) the same author discusses the 

 structural differences between the various forms of matter and 

 points out the utility of the Laue radiograms in the study of 

 morphotropy and isomorphism. 



Pfeiffer (Zeit. Anorg. Chem. 97, 161, 1916), again, argues 

 that the crystal structures so far determined can be explained 

 in terms of Werner's co-ordination theory. Thus the co-ordina- 

 tion number of sodium is six, which is in harmony with the fact 

 that in rock-salt each sodium atom is equidistant from six 

 chlorine atoms, while the numbers " eight " and " four " for 

 calcium and fluorine respectively correspond with the relations 

 in fluorspar. Radicles such as S0 4 and C0 3 are considered to 

 occupy the place of one atom, and hence to have a single co- 

 ordination number. 



Smits and Scheffer (Proc. Akad. Wetensch. Amsterdam, 19, 

 432, 1916) maintain that the distance between adjacent mole- 



