226 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1921. 



CaO About 4.9 per cent by weight. 



MgO About 3.7 per cent by weight. 



Na 2 About 3.3 per cent by weight. 



K 2 About 3.0 per cent by weight. 



H 2 About 2.0 per cent by weight. 



C0 2 About 0.7 per cent by weight. 

 An understanding of the chemistry of these oxides and their com- 

 binations is essential to the progress of petrology. Their study 

 should proceed from the simple to the complex, i. e., should begin 

 with the individual oxides, then proceed to their two-component 

 systems, then the ternary systems, and so on. Upon this fundamen- 

 tal basis is then erected the structure of physical properties for each 

 system — densities at all accessible temperatures, mechanical properties, 

 fluidity, surface tension, specific and latent heats, etc. 



The study of these systems may be divided on practical grounds 

 into (1) investigations of the anhydrous oxides and silicates (taking 

 in the first eight oxides in the list above) ; (2) investigations involv- 

 ing, hydrous silicates, as well as combinations containing both carbon 

 dioxide and water. 



(1) Anhydrous silicates.— Work on the first group involves high- 

 temperature researches under ordinary atmospheric pressure con- 

 ditions, except in the case of systems containing the oxides of iron, 

 where the oxygen pressure must be controlled, and systems contain- 

 ing the alkali silicates, where attention to moisture, carbon dioxide, 

 and volatility of the oxides is necessary in certain cases. Consider- 

 able progress has been made in the study of the anhydrous silicates. 

 The phase rule diagrams of the four ternary systems of SiO, Al 2 O n , 

 MgO, and CaO are now complete, and a large amount of data is at 

 hand on the alkali feldspars, the forms of silica, portions of several 

 quaternary silicate systems, etc. 



(2) Silicates with volatile components. — Work on systems involv- 

 ing the volatile components C0 2 and H 2 must be done, for the 

 most part, under pressure, and with apparatus designed especially 

 for this purpose. The methods are well in hand and progress is be- 

 ing made in assembling experimental data. The theoretical side, in- 

 volving the complications due to pressure as a variable in addition 

 to temperature, is also being carried forward by several investi- 

 gators. 



So much for the 98 per cent. But the remaining 2 per cent con- 

 tains many natural substances of such great economic as well as 

 geologic interest that they must also receive attention. These may 

 be roughly classified as in the following examples : 



The sulfide ores (e. g., sulfides of iron, nickel, zinc, copper, lead, 

 cobalt, cadmium, mercury, silver). — These must be studied both in 



