r.)i3] 



on Great Advance in Crystallography 



679 



positions (the rotation beinir also indicated in Fig. 2 by the dotted 

 ellipses). In a similar manner the dimensions of the striicturiil 

 units — the molecular volume (that of the "• brick " regarded as a 

 molecule) and its expression in the 

 three dimensions of space (which we 

 now have a means of determining, by 

 combining the density and the crys- 

 tallographic axial ratios, and which 

 are known as topic axial ratios) — 

 vary in regular progression as func- 

 tions of the atomic weight. Hence, 

 the structural dimensions of the two 

 extreme members of any of these 

 groups of salts, the potassium and the 

 caesium salts, are most divergent, and 

 Barker has shown that while the 

 rubidium salt will in general form 

 parallel growths with either the 

 potassium or the caesium salt, the 

 potassium and caesium salts them- 

 selves will never form satisfactory 

 parallel growths on eacli other, clearly 

 owing to the disparity in the dimen- 

 sions of their structural units. 

 Further, the ammonium salt of any 

 group has the interesting property of sidm by Rubidium or Cesium. 

 forming crystals of which the mole- 

 cular volumes and topic axial dimensions are almost identical with 

 those of the rubidium salt of the same group. Now it is most im- 

 portant and conclusive that the ammonium and rubidium salts form 

 the best of all parallel growths ; they are, indeed, no longer merely 

 parallel growths, but zonal growths and complete overgrowths. They 

 also form excellent mixed crystals, and in every way which has yet 

 been experimentally tested they show the nearest approach to true 

 iso-structure. Some photographs of such parallel growths of rubidium 

 and ammonium salts will render the matter clear. There is also one 

 case of iso-structure investigated by Barker, that of calcite and sodium 

 nitrate, which is of particular interest ; for these are not chemically 

 iso-morphous substances, but they happen to have similar rhombo- 

 hedral symmetry and almost identical molecular volumes and topic 

 axial dimensions. The slide of parallel growths of sodium nitrate on 

 calcite (reproduced in Fig. 3, Plate I, by the kindness of Mr. Barker) 

 is very striking and conclusive. Barker has recently suggested the 

 nature of the structure of these two substances, two molecules forming 

 the unit " brick," as represented in the slide exhibited. In the case 

 of potassium sulphate and its analogues, one molecule forms the 

 " brick," as represented by the next slide. 



Vol. XX. (No. 107) 2 z 



Fig. 2. — Rotation of the Op- 

 tical Ellipsoid of the Double 

 Sulphates on replacing Potas- 



