Buddington — Natural and Synthetic Melilites. 87 



compounds yield crystallized products essentially similar 

 in composition and properties to the minerals of the 

 humboldtilite series, varietal members of the melilite 

 group relatively poor in ferric iron. 



The humboldtilites are interpreted as essentially isomor- 

 phous mixtures of positive uniaxial akermanite (2CaO. 

 Mg0.2Si0 2 ) and a negative, tetragonal, uniaxial, moder- 

 ately birefringent form of 3CaO.Al 2 3 .3Si0 2 with minor 

 amounts of gehlenite, a ferrous iron compound, and 

 3KO.R 2 3 .3Si0 2 compounds. The compositions of the 

 humboldtilites lie in a zone which exhibits the lowest 

 temperatures of complete melting for the components 

 involved. 



Mixtures of 3CaO.Feo0 8 .3Si0 9 , 2CaO.Mg0.2Si0 2 , 

 3Na 2 O.Al 2 3 .3Si0 2 , 2CaO.Al 2 3 .Si0 2 , and 3CaO.Al 2 3 . 

 3Si0 2 , similar in composition to some of the ferric iron 

 rich melilites, were studied, and the properties of the 

 synthetic crystalline material were found to be quite 

 different from those of natural minerals of similar 

 composition. 



The natural ferric iron rich melilites probably form at 

 temperatures lower than those of the present experi- 

 ments since some of such melilites investigated decom- 

 pose or invert at temperatures at least as low as 850° C. 



Three new analyses of humboldtilites and one of ferric 

 iron rich melilite by Dr. H. S. Washington are given. 

 The ferric iron rich melilite differs from any hitherto 

 analyzed. 



This study serves to emphasize the great complexity 

 of this group of minerals and the necessity for further 

 data on their composition and properties, and for further 

 experiment in synthesis. The extent to which such known 

 and hypothetical compounds as 3MgO.Al 9 3 .3Si0 3 ; 

 3FeO.Al 2 3 .3Si0 2 ; 3K 2 O.AL0 3 .3Si0 2 ; 2CaO.FeO.2SiO* ; 

 2CaO.Fe 2 3 .Si0 2 , etc., may enter into the melilite min- 

 erals, and their corresponding effect, is yet to be studied. 



In conclusion, the writer desires to express his thanks 

 to Doctors N. L. Bowen, H. E. Merwin, and H. S. Wash- 

 ington for their friendly criticism and assistance in this 

 work. 



Geophysical Laboratory, 



Carnegie Institution of Washington, 

 Washington, D. C, June 20, 1921. 



