2 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES voL. 13, No. 1 
mixtures of diopside and nephelite, diopside may separate early and 
then, by reaction with the liquid, be replaced by forsterite (olivine) 
and akermanite (melilite). This was in accord with the mineral 
paragenesis in the Cadieux rocks, and it was suggested that an alka- 
line (nephelite-rich) liquid reacted with augite to produce monticellite 
and melilite, acting, as it were, as a desilicating agent. It was sug- 
gested also that an analcite-rich liquid was produced by the reaction 
and that this gave rise to the analcite dikes. 
Scheumann in a more recent paper discusses the application of these 
conclusions to the rocks of the Polzen region and decides that, while 
these reactions were in part responsible for the production of lime- 
rich minerals, especially melilite, yet addition of lime to the magma 
must be considered an important factor particularly in the production 
of monticellite. He suggests, too, that since melilite separates ‘‘in 
excess” it may later react with the liquid to reproduce pyroxene and 
of this reaction he finds evidence in the Polzen rocks.‘ In the artificial 
mixtures, no doubt, this reversal of the reaction with reformation of 
pyroxene and nephelite would occur at lower temperatures, but this 
fact cannot be demonstrated on account of the sluggishness of reaction. 
The reversal did not occur in the Cadieux rocks for the reason, as 
believed, that the reacting liquid was separated (squeezed out) and 
formed the analcite dikes. Even if it had not been separated, anal- 
cite might have persisted as a ground-mass mineral in the presence 
of melilite just as quartz may occur in the ground-mass of an olivine- 
bearing rock. The requisite condition is very rapid cooling. Though 
Scheumann accepts the formation of some melilite by the reaction 
method demonstrated experimentally, he nevertheless appears to be 
offering an objection to it when he says “‘ Melilite rocks in which the 
more silicic alkaline residue has crystallized as analcite are unknown.” 
With this very question in mind I have myself been examining a few 
melilite rocks, sections of which were readily obtainable. In two of 
these I have found analcite as typical interstitial, residual material. 
The one is the melilite-nephelite basalt of Moiliili, Oahu, Hawaiian 
Islands which has been described by Cross.’ In places in this rock 
the interstitial material. consists entirely of anhedra of nephelite but 
in other places it is a mixture of nephelite and analcite in which the 
nephelite occurs as euhedra in an analcite base and in which there can, 
therefore, be no question of the analcite being formed by alteration 
of the nephelite. The attack of the alkaline liquid on the augite is 
- shown by the outlines of the augite “interlocking most irregularly 
4 Neues Jahrb., Centralbl. 1922: 495-545. 
®’W.Cross. U.S. Geol. Surv. Prof. Paper 88: 20-22. 1915. 
