244 PHENOMENA, ATOMS, AND MOLECULES 



salts of both acids are readily soluble in water, while the lead and silver 

 salts are very difficultly soluble. 



Groth ^ gives the following crystallographic data for anhydrous potas- 

 sium cyanate and trinitride. Both belong to the tetragonal system. The 

 ratio of the axes a:c is i : 0.5766 for the cyanate and i : 0.5798 for the 

 trinitride. Both have "strong negative double refraction." The most com- 

 mon faces of both crystals are given as C (001) and O (m). Cleavage 

 was not observed in either case. No other data are given by Groth from 

 which a comparison of cyanates and trinitrides can be made. 



From the octet theory and the above data it is therefore safe to predict 

 that the physical properties of the salts of cyanic and hydronitric acids 

 will be found to be practically identical. This resemblance should be at least 

 as close as that between nitrous oxide and carbon dioxide and should cover 

 solubility in water, in alcohol, etc., density in crystalline form and in 

 solutions, viscosity of solutions, optical and magnetic properties, electric 

 conductivity of solutions, etc. The densities should be alike because the 

 molecular weight of N.>, is the same as that of NCO. 



The available experimental data, although meager, are sufficient, I 

 think, to show the complete isomorphism of cyanates and trinitrides. The 

 similarity of structure thus follows directly from Mitscherlich's rule. 

 These experimental data furnish direct evidence against such structural 

 formulas as 



K — N<ll, K — N = N = N, K — O — C = N, 



but strongly support the octet theory structures : 



K+ (N = C = O)- and K+ (N = N = N)" 



These formulas show that the covalence of potassium is zero ; the nitrogen 

 and oxygen in the cyanate are dicovalent. while in the trinitride one of the 

 nitrogens is quadricovalent and the other two are dicovalent. These data, 

 as far as they go, thus constitute experimental proof of the octet theory 

 of valence. Further experimental work on the physical properties of 

 cyanates and trinitrides is highly desirable. 



The octet theor\- also indicates that diazomethane should have the struc- 

 ture H2C = N = N and should thus be isosteric with HoC = C = O, a 

 compound apparently not given in Beilstein. This compound should closely 

 resemble diazomethane in all its physical properties such as freezing point, 

 vapor pressure, viscosity, etc. 



No direct comparison can be made of the physical properties of isosteres 

 ^ P. Groth, "Chemisclie Krystallographie," Leipzig, Part I, 1906, Part II, 1908. 



