ISOMORPHISM, ISOSTENISM AND CO VALENCE 253 



potassium chlorate and sodium nitrate closely resemble each other. He 

 does not believe that mixed crystals can be formed because the molecular 

 volumes are so different. However, monoclinic mixed crystals of potas- 

 sium chlorate and potassium nitrate up to 15% of the nitrate have been 

 obtained, as well as orthorhombic mixed crystals, with 25% of the chlorate. 

 This is considered to denote similarity of structure, notwithstanding the 

 complete difference of crystal form between potassium nitrate and chlorate. 

 Silver chlorate and bromate are tetragonal while silver nitrate is ortho- 

 rhombic. The chlorate also forms cubic crystals isomorphous with sodium 

 chlorate. Silver chlorate thus shows no resemblance to the nitrates. 



Anhydrous calcium, strontium, and barium nitrates crystallize in the 

 cubic system, but strontium chlorate Sr( 003)2 is orthorhombic with axial 

 ratios, 0.9174:1 :o.6oo3. It also exists in 3 other modifications, no one of 

 which is cubic. However, even here it has been possible to make mixed 

 crystals containing 7 to 12% of chlorate. 



The salts Mg(N03)2.6H20, Ni(N03)2.6H20, Co(N03)2.6H20, are 

 monoclinic, the nickel and cobalt salts being isomorphous with each other 

 but not with the magnesium salt. On the other hand, the corresponding 

 chlorates of nickel and cobalt, Ni(C103)2.6H20, etc., and the bromates of 

 magnesium, nickel, cobalt and zinc, Mg (Br03)2.6H20, etc., are cubic. 

 Only in one instance were mixed crystals obtained : Zn(Br03)2.6H20 and 

 Co(N03)2.6H20 crystallizing together as cubic crystals. 



Looking back over this comparison of chlorates and nitrates we see 

 that among the 20 compounds considered (belonging to 5 different crystal 

 systems), there is only one instance, namely that of sodium nitrate and 

 the unstable trigonal modification of sodium chlorate, in which correspond- 

 ing nitrates and chlorates even belong to the same system. 



This evidence seems sufficient to prove that nitrates and chlorates have 

 fundamentally different constitutions. A careful study of the trigonal form 

 of sodium chlorate should be made, preferably by the X-ray method, to 

 determine if its structure is actually like that of sodium nitrate. If it turns 

 out to be so, it is probably to be explained as a very unusual form in which 

 two of the electrons in the atoms constituting the chlorate ion become 

 unavailable, perhaps by being imprisoned within one of the octets. Under 

 such conditions the chlorate ion might become "pseudo-isosteric" with the 

 nitrate ion, so that the chlorine atom might then act with a covalency of 4. 



There are no available data by which to test isomorphism between 

 chlorates and metaphosphates. 



These considerations show that the ordinary valence theory not only 

 fails to predict cases of isomorphism which do exist ( for example, potas- 

 sium chlorate and strontium sulfate) but predicts isomorphism (sulfites 

 and carbonates) where none exists. The octet theory does not fail in either 



