sosman: problems of the oxides of iron 55 



The viscosity of a 39.99 per cent sucrose solution was meas- 

 ured at temperatures varying from 0°C. to 95°C. The observed 

 values correspond to the formula : 



t = 0.597 + 20) - ]^^ + 38.24 



+ 20 



On repeating these measurements practically the same values 

 were obtained. 



Measurements were made with 20.007 per cent and 59.96 per 

 cent sucrose solutions also. The fluidities at the latter con- 

 centration correspond to the formula: 



t = 1.472 (^ + 5) - ®®2 + 58.62 



+ 5 



The values found in the present investigation indicate a 

 higher viscosity than those of previous investigators. 



CHEMISTRY. — Some problems of the oxides of iron. 1 Robert 

 B. Sosman, Geophysical Laboratory. 



Starting with the general principle that the soundest prog- 

 ress in science is made by proceeding from that which is simple 

 and understandable to that which is more complex and less 

 understandable, various investigators have sought to unravel 

 the complexities of natural rocks and minerals by studying 

 first their simplest constituents. Since it appears that under 

 most rock-forming conditions silica (Si0 2 ), alumina (A1 2 3 ), 

 lime (CaO), and magnesia (MgO) always act as units, these 

 oxides may be taken as the pure components in the fundamental 

 experiments. 



With the oxides of iron the case is different. The equilibrium 

 between ferrous and ferric iron in a mineral or rock can vary 

 with every change of temperature and with every change in 

 the amount or pressure of available oxygen. In order to under- 



1 Presidential address, Chemical Society of Washington, January 11, 1917. 

 Presented before a joint meeting of the Chemical Society and the Washington 

 Academy of Sciences. 



