Summary: From thermodynamic considerations of the system anorthite-åkermanite it appears that the effective molecular weights of the ‘molecules’ in the melts of this system are 0·5 the molecular weights of anorthite as well as that of åkermanite.

The effective molecular weight value of 0·5 anorthite is in significant contrast to that of 1 anorthite obtained by Adams and Cohen (1966) from a similar analysis of the system anorthite-diopside (Bowen, 1915, and Osborn, 1942). The analysis by Adam and Cohen would lead to a non-ionic silicate melt structure in the system anorthite-diopside in contrast to the accepted electrolytic view of silicate melts. It is suspected that the reason for this apparent difference lies in the published liquidus morphology of the anorthite primary fields of the system anorthite-diopside. This same liquidus morphology is also responsible for the distortions in the isofracts and isotherms in the high-temperature area of the primary field of anorthite of the system anorthite-åkermanite-diopside (de Wys and Foster, 1958).