19, 6 Perkins: Chlorine Dioxide and Compounds 735 
Returning now to fluorine, it is easy to see that if its elec- 
tron binding strength in the unicovalent condition is too great 
for combination with any but a few atoms, still fewer compounds 
of dicovalent fluorine are to be expected. There is no reason, 
however, why certain atoms of very high electron binding 
strength cannot combine with dicovalent fluorine. The hydrogen 
atom is to be considered in this class, although occupying a 
unique position in it. In hydrofluoric acid we have some mole- 
cules containing dicovalent fluorine, and also, as has been pointed 
out by Latimer and Rodebush " some atoms of hydrogen in an 
unstable dicovalent condition. 
Aside from hydrogen, the “inert” atoms seem to be the only 
ones capable of matching the electron binding strength of 
dicovalent fluorine. This suggests the possibility of an ion, 
fot oe | 
The oxygen atom, in the unicovalent condition has a binding 
strength which is very weak compared to that of fluorine. In 
the bicovalent condition its strength of course depends some- 
what on the nature of the combined atoms, but it is comparable 
with that of unicovalent fluorine. Tricovalent nitrogen and 
quadricovalent carbon also have binding strengths of this same 
order of magnitude, so unicovalent fluorine combines with qua- 
dricovalent carbon in CF,, and with tricovalent nitrogen in ONF. 
THE CHLORINE OXIDES 
We are now in a position to return to the chlorine oxides. 
Unicovalent chlorine, due to the size of its kernel, has a binding 
strength which is much less than that of unicovalent fluorine or 
bicovalent oxygen, although somewhat greater than that of 
unicovalent oxygen. 
According to this view Cl—F and ClI—O—Cl would be 
very unstable molecules. CIF is not known, but Cl,0 is. Al- 
though the formula CL—O—Cl is often ascribed to this substance, 
the theory here presented indicates that Cl—Cl<O is the 
correct formula, because the evidence of chlorine compounds in 
general indicates that unicovalent chlorine is too weak to share 
a pair with bicovalent oxygen. 
We might expect that bi- or tricovalent chlorine could share 
electrons stably with bicovalent oxygen, but apparently this is 
* Latimer, W. M., and Rodebush, W. H., Journ. Am. Chem, Soc. 42 (1920) 
1425. 
