948 



SCIENCE 



[N. S. Vol. XLII. No. 1096 



solubility of the salt and of its hydrates both in 

 water and in alcohol and the ability of the alcohol 

 to displace the water of combination. 



Equilibria in Systems of Ketones, Water and 

 Salts with a Method for the Determination of 

 Methyl Alcohol in the Presence of Acetone: G. 

 B. Pkankfoeter and Lillian Cohen. 

 The Influence of Dissolved Substances upon the 

 Velocity of Crystallization of Under-cooled 

 Water: James H. Walton, Je., and A. C. 

 Beann. 



The velocity of crystallization of xmder-cooled 

 water has been mea,sured at — 9°, and the inhibit- 

 ing effect of dissolved substances — such as acids, 

 alkalies, salts, alcohols and some colloidal sub- 

 stances — determined. When possible tenth molec- 

 ular solutions were used. It has been found that 

 a rough proportionality exists between the num- 

 ber of atoms in the molecular and the inhibiting 

 effect. The greater the number of atoms in a 

 molecule of dissolved substance the slower the rate 

 of crystallization. 



The Partition Coefficients of Hydrogen Peroxide 

 between Water and Certain Organic Solvents: 

 James H. Walton, Je., and H. A. Lewis. 

 Hydrogen peroxide is soluble in many organic 

 solvents, such as phenol, aniline and certain esters. 

 In general, any organic liquid that dissolves water 

 will dissolve hydrogen peroxide. A determination 

 of the partition coefficient in most of these sol- 

 vents shows a normal molecular weight for the 

 peroxide. In Quinoline the partition coefficient 

 water : quinoline, is about 1 : 3. This ratio changes 

 with change of concentration of hydrogen perox- 

 ide, pointing to association of the molecule in the 

 quinoline solution. The partition coefficient 

 water : quinoline, has been studied at various tem- 

 peratures. 



The Preparation of Pure Iron and Iron Carbon 

 Alloys: J. E. Cain, E. Scheamm, and H. E. 

 Cleaves. 



It is shown that previous work on the iron-car- 

 bon diagram is unsatisfactory because of the great 

 variation in the materials used. It was therefore 

 thought necessary to produce a series of alloys of 

 great purity to form the basis of a redetermina- 

 tion of the diagram at the Bureau of Standards. 

 The general method pursued consisted in melting 

 electrolytic iron with sugar carbon in magnesia 

 crucibles. The electrolytic iron was prepared 

 from ingot iron anodes in a chloride bath with or 

 without the use of porous cups. The operation of 

 melting the iron with carbon gave great trouble 



at first, because the ingots obtained were full of 

 blow-holes and contained considerable quantities 

 of impurities. The difficulties were overcome by 

 melting in a vacuum furnace, and making cru- 

 cibles of especially pure magnesia, made and cal- 

 cined with great care at the Bureau of Standards. 

 A satisfactory procedure was finally worked out 

 and a series of alloys prepared of the composition 

 Ee + C = 99.96 per cent. 



The Oxides of Iron. I. Solid Solution in the 

 System FejO^ — Fe^Ot: Robert B. Sosman and 



J. C. HOSTETTEK. 



This investigation of the chemical relationships 

 of the iron oxides has been undertaken as a 

 basis for the study of the iron-bearing silicates at 

 high temperatures. Measurements of the dissocia- 

 tion pressure of the iron oxides were made in a 

 vacuum furnace with a heating tube of platinum- 

 rhodium. A study of the conditions of equilibrium 

 shows that reproducible oxygen pressures can be 

 obtained at a given temperature. Equilibrium is 

 attained in a few minutes at high temperatures, 

 although certain disturbing reactions go on slowly. 

 One of these by-reactions is the reduction of the 

 oxide by platinum, yielding oxygen and an iron- 

 platinum alloy. Another is a slow disappearance 

 of oxygen, which has not been satisfactorily ex- 

 plained. Ferric oxides from various sources yield 

 practically identical pressures. The same pres- 

 sures are also attained on both rising and falling 

 temperatures. The oxidation of magnetite gives 

 pressures which are a little higher than those pro- 

 duced by dissociation of FejOa. The pressure- 

 composition isotherm for the system Fe.O, — Fefii 

 at 1,200° indicates a continuous solid solution series 

 from Pe^Os over to a point very near FejO,, jf 

 not over the entire range to FosO,. The opacity of 

 the products prevents an optical demonstration of 

 the existence of solid solution in products with 

 more than 18 per cent. PeO, but its existence can 

 be shown optically in products which are more 

 ferric than this. The pressure-composition iso- 

 therm at 1,100° confirms that at 1,200°. The 

 major portion of the oxygen pressure curve of the 

 system at 1,200° lies between the limits 4 mm. 

 and 1.5 mm. The pressure drops rapidly near 

 PcsOj, and rises rapidly near Pe^Os. Since the 

 dissociation of Pe.Oa results in the formation of a 

 solid solution, the pressure of oxygen and the com- 

 position of the solid phase depend upon the re- 

 lation of the weight of the charge to the vol- 

 ume of the space into which the oxygen dissoci- 

 ates. This fact accounts for the variety and un- 



