426 Dr. L. Vegard on 



are arranged in a lattice of the zircon type ; but are we, 

 then, to draw the conclusion that also xenotime is to be 

 given a constitution formula, YO 2 P0 2 ? I tliink not, in 

 the usual meaning of the word, although we found * that 

 the geometry of the zircon lattice would indicate that in 

 a " molecular element " M0 2 oxygen atoms probably were 

 attached to the central atom M in another way to that in 

 which they were attached to the other neighbouring atoms. 



Indeed, it will be difficult to prove anything so long as we 

 do not know exactly what realities are to be associated with 

 a constitution formula and with the conception of valency. 



Probably our ideas in this respect will have to be 

 fundamentally changed. 



A constitution formula serves to express certain chemical 

 properties of the substance, but it may not at all be able 

 to express the true relation in space of the atoms which 

 constitute the molecule. 



Let us return to our case of xenotime, and assume it is 

 to be considered as a phosphate, the constitution formula of 

 which should be 



/Ox 



This formula would put one of the oxygen atoms into a 

 singular position ; and if the valency units are to indicate 

 forces, the Y atoms would be attached to the P atoms, as far 

 as chemical attraction is concerned, only by means of the 

 oxygen atoms. 



Comparing the constitution formula with the arrangement 

 found for the atoms in the solid substance, we see that the 

 constitution formula does not give any indication of the true 

 arrangement. Thus there is no oxygen atom which takes up 

 any singular position, and the oxygen atoms seem to be 

 associated in a similar way with both Y and P. 



A constitution formula which should express the way in 

 which the atoms are arranged in the crystal — or the con- 

 stitution of the solid form — should be written Y0 2 P0 2 . 



This result may be best expressed by saying that the 

 constitution formula of the solid structure and the con- 

 stitution formula giving the chemical properties may be 

 two different things ; and we cannot, from the atomic 

 arrangement, find out the "chemical constitution" of the 

 substance. 



* Phil. Mag. xxxii. p. 92. 



