166 CARNEGIE INSTITUTION OF WASHINGTON. 



(2) The pseudowollastonite solid solutions whose compositions form an 

 area bounded by the following lines: (1) the CaO.SiOa - CaO.Mg0.2Si02 

 line; (2) a line running from the composition CaO 44.4, MgO 3.1, SiOa 52.5 on 

 the above-mentioned hne across to the composition CaO 46.7, MgO 3.5, 

 Si02 49.8, on the CaO.SiOa - 2CaO.Mg0.2Si02 line; (3) then either the last- 

 mentioned line back to CaO.Si02, or, more probably, an approximate continu- 

 ation of hne (2), to about the composition CaO 50, MgO 50, on the side-Une. 



(3) The woUastonite solid solutions. These extend to about 17 per cent 

 diopside or 3.2 per cent MgO at the higher temperatures. The most con- 

 centrated of these solid solutions along the diopside line (the 17 per cent) 

 decomposes at 1340° =t 5°C., and this sohd solution is the only one represented 

 on the liquidus. 



(4) The 5Ca0.2Mg0.6Si02 solid solutions. Only a few of these solid 

 solutions which are decomposed at the higher temperatures near the de- 

 composition-temperature of the pure compound are stable in contact with a 

 suitable liquid. 



(5) Certain members of the monticellite soHd solutions. Monticellite 

 takes up forsterite in solid solution to the extent of about 10 per cent, and 

 the decomposition-temperature of the solutions is thereby raised. Monti- 

 ceUite itself probably decomposes at too low a temperature to ever occur as a 

 primary phase. 



The temperature-cencentration relations of the liquids which may be in 

 equilibrium with each of these phases have been thoroughly investigated, 

 where necessary, by means of the quencliing method, and the results obtained 

 have been correlated with the existing data on the remainder of the ternary 

 system. The compounds 5Ca0.2Mg0.6Si02 and 2CaO.Mg0.2Si02 have not 

 been prepared previously. Attempts to prepare a compound of the formula 

 8Ca0.4Mg0.9Si02 (Schaller's akermanite) gave negative results. The 

 monticelhte solid solutions and the compound akermanite are discussed at 

 length, but the woUastonite and the 5Ca0.2Mg0.6Si02 soUd solutions are 

 only briefly mentioned, as they are made the subject of a subsequent paper. 

 (See abstract (24), below.) Experiments were made on the tridymite-cris- 

 tobalite inversion temperature, which was found for this system to be below 

 1500° C, in approximate agreement with Fenner's original value of 1470°. 

 The great sluggishness of the inversion precluded a more exact determination 

 on our part. 



(24) WoUastonite (CaO.Si02) and related solid solutions in the ternary system lime- 

 magnesia-silica. J. B. Ferguson and H. E. Merwin. Am. J. Sci., 48, 165-189 

 (1919). 



The study of the ternary system CaO-MgO-Si02 (described under No. 23 

 above) brought to Hght many perplexing Hquidus relations for which there was 

 no adequate explanation. An investigation of the solidus relations was there- 

 fore started in order to clear up the doubtful points, and the results of this 

 investigation are given in this paper. The salient features of these results are : 



(1) A confirmation of the earlier work in regard to wollastonite-diopside 

 sohd solutions, woUastonite taking up a maximum of 17 per cent of diopside. 



(2) The existence of solid solutions of pseudowollastonite and diopside 

 containing as a maximum about 16 per cent of diopside. 



(3) The finding of the new compound, 5Ca0.2Mg0.6Si02. 



(4) The existence of sohd solutions of akermanite (or perhaps of an un- 

 stable compound, 3CaO.Mg0.3Si02) in woUastonite and pseudowollastonite. 

 The woUastonite solutions extend to a composition containing between 60 



