JuLY 19, 1923 FERGUSON: OXIDES OF IRON 279 
ferro-ferric oxide phase and vapor in the system, (Fe-O). Analyses of 
these samples are given in table II. 
The free iron was determined by the copper sulfate method.® The 
ferrous iron reported was the value obtained by titration corrected for 
the metallic iron present. The ferric oxide was obtained by difference. 
Each of these samples was divided into two parts. One part was 
kept and the other part was heated at a low temperature in vacuum 
until it contained an appreciable amount of free iron. We shall refer 
to this latter part as the converted part. Portions of each kind of 
material were then heated in vacuum in sealed pyrex glass tubes for 
five hours at carefully regulated temperatures. The free iron in each 
portion was then determined. The results are shown graphically in 
figure I. The graphs represent in order samples I, II, and III. The 
right hand curves were obtained with the original materials. 
If all the ferrous oxide in samples I and II were converted to free 
iron and magnetic oxide, the resultant free iron content would be, 
respectively, 13.3 and 16.7 per cent. This calculation assumes the 
equation 4 FeO=Fe+Fe;0:. We were able to obtain conversions of 
13.5 and 16.9 per cent iron and this is as good a check as one could 
expect from such materials. 
With sample I, but one reaction appears to take place. Iron does 
not begin to form in the iron-free material until a temperature of 526°C. 
isreached. The iron in the converted part begins to combine at 577°C. 
and the reaction proceeds until all the free iron has combined. 
With samples II and III, there appears to be an additional reaction 
by means of which a small quantity of iron is liberated. ‘This may be 
noted even above 600°C. and the quantity of iron so formed does not 
appear to vary much in the temperature interval between this tem- 
~ perature and the temperature at which the major reaction takes place. 
The iron liberated in this reaction does not recombine when the con- 
verted materials are heated at the higher temperatures and the major 
reaction takes place. 
Hilpert could not tell whether the compositions richer in ferrous 
oxide were unstable at the lower temperatures, or-whether they were 
stable but could not be prepared owing to the slow reaction velocity. 
Our experiments indicate that they are not stable. It seems probable 
that ferrous oxide in the pure state is unstable at a fairly high tempera- 
ture but that by the solution of ferro-ferric oxide the temperature of 
transition is lowered. The rates at which iron would form on cooling 
_.° Williams and Anderson, Journ. Ind. Eng. Chem. 14: 1057. 1922. 
