RECENT PROGRESS IN CRYSTALLOGRAPHY 109 



homogeneous structure as ensures that they are always 

 present in the same proportion. Moreover, the explanation of 

 poly-morphism is at once clear, for the equilibrium of the inter- 

 penetrating atomic point systems will be affected by temperature, 

 and a combined system, possible and stable at one temperature, 

 may become unstable at another considerably higher or lower, 

 when, if other conditions are suitable, another form, possibly 

 belonging to an altogether different system of symmetry, may 

 be produced. 



Having now shown that, from both the experimental and the 

 geometrical standpoint, a crystal may be structurally regarded 

 as an organised assemblage of points, which may be formed 

 either by the centres of gravity of the chemical molecules or by 

 the atoms of which the molecule is composed, and that the 

 topic parameters express, in the case of an isomorphous series, 

 the relative separations of these points along the axial direc- 

 tions, we are in a position to discuss the results of the writer's 

 experimental work as regards structure. 



In both series of salts, when potassium is replaced by 

 rubidium or caesium, a deformation of the crystal structure 

 occurs in the form of an extension of all the topic parameters — 

 that is, an increase occurs in the separation of the centres of 

 gravity of the molecules along each of the directions of the 

 crystallographical or topic (not always identical) axes. The 

 increase augments as the atomic weight rises, so that the 

 intermediate position for any rubidium salt is always closer to 

 the potassium salt position than to that for the caesium salt — 

 that is to say, the function of the atomic weight is of an order 

 higher than the first. Moreover, a similar increase of the 

 distances separating the points (molecular centres) is observed 

 when sulphur is replaced by selenium. 



It has been shown by Fedorow to be very important that 

 the correct type of homogeneous structure should be diagnosed 

 before the proper topic parameters can be arrived at. Now the 

 rhombic sulphates and selenates exhibit a marked pseudo- 

 hexagonal character, the primary prism angle being within 

 half a degree of 6o°. Hence, while <u represents as usual the 

 vertical topic parameter, and \jr that along one of the horizontal 

 axes of the rhombic system, %, instead of representing the other 

 horizontal axis at right angles, represents the parameters along 

 a pair of horizontal axes inclined at nearly 6o° to each other and 



