1902.] on the Constitution of Crystals. 189 



In the case of the principal angle of the monoclinic series, which 

 determines the inclination of the inclined axis characteristic of that 

 system of symmetry, the change of angle is directly proportional to 

 the change in atomic weight. 



Before concluding what may be said about the exterior mor- 

 phology, reference may be made to one other fact, which follows from 

 that just referred to and from the results of careful determinations 

 of the specific gravity of the salts. The latter enables one to arrive 

 at the molecular volume, which in the case of salts belonging to the 

 same series does actually represent the relative volume of the chemical 

 molecules. The size of the molecules of the rubidium salt was in 

 every case found to be intermediate between the sizes of the mole- 

 cules of the potassium and csBsium salts. By combining these 

 molecular volumes with the lengths of the crystallographical axes as 

 afforded by the angular measurements, it has been possible to de- 

 termine the relative dimensions, in the three directions of space, 

 of each chemical molecule. 



This was preceded by a proof, which has been confirmed by Dr. 

 Fock, of Berlin, from an entirely different point of view, that the 

 unit of the crystal structure in these salts was identical with the 

 chemical molecule, and was not an aggregate of such. 



The result has been to show that the sizes, or at any rate the 

 distances apart from centre to centre, of the chemical molecules in 

 the three rectangular directions of space are intermediate in the 

 case of every rubidium salt. Thus the directional dimensions of the 

 molecules, like the angles of the crystal structure, follow the order 

 of the atomic weights of the alkali metals. 



This fact is clearly exhibited in exaggerated fashion by the model, 

 and it will be observed that the intermediate position of the rubidium 

 salt is somewhat nearer to that of the potassium salt than to that of 

 the csBsium salt. 



In turning now to the 'consideration of the optical characters of 

 the crystals of the series of salts in question, it will be necessary to 

 remember two main facts. The first is, that the symmetry of the 

 rhombic system, in which the sulphates and selenates of the alkalies 

 crystallise, determines that the velocity of light transmission shall 

 be different along the three morphological axes of the crystals, and 

 that two of these directions shall be those of maximum and minimum 

 velocity respectively. In other words, the ellipsoid which, as is well 

 known, in general represents the velocity of light transmitted through 

 a crystalline medium, is one whose three rectangular axes are of un- 

 equal lengths, but which is fixed in direction, for these axes are identical 

 in direction with the three morphological axes. The second is, that in 

 the case of the monoclinic double sulphates and selenates the sym- 

 metry of the system demands that only one of the three rectangular 

 axes of the optical ellipsoid shall be identical with a morphological 

 axis, the latter being the one symmetry axis of the system. The 



