Summary: The course of assimilation and differentiation illustrated in the rock assemblages may be schematically represented in the following diagram (p. 465). The important part that complex pyroxenes play in the genesis of these assemblages will have become apparent from the preceding discussion.

Assimilation of lime by the olivine-dolerite magma leads first to the formation of pyroxene-rich dolerites and ultimately pyroxenites, by conversion of the hypersthene molecule of the augite and conversion of olivine to monoclinic pyroxenc. This process is accompanied by enrichment of the pyroxenite in Mg relative to Fe. The liquid is correspondingly enriched in Fe and this is one of the most striking chemical features of the rock series, as given below : Fe/Mg Pyroxenite (Pyroxene) 0·71 Pyroxene-rich dolerite 1·25 Olivine-dolerite 1·6 Titanaugite-melilite-rock 8·3 Nepheline-dolerite 13·0 Melilite-rock 19·0

The formation of the pyroxene-rich rocks leaves a magma enriched in plagioclase, soda, and iron (strongly), which corresponds to the olivine-free augite-dolerites with their dark-eoloured augites. Further assimilation of lime leads to the formation of an augite-plagioclasenepheline-rock, in which the evidence for the resorption of plagioclase is well seen, giving place to nepheline and a titaniferous lime-augite rich in alumina.

These assemblages are limited to the hybrid zone, and occur in segregatory patches and even in small veins, pointing to a later consolidation and a mobility superior to that of the pyroxenite type. In them areas composed almost wholly of titanaugite and nepheline (hydronephelite) give the nepheline-dolerite assemblage.

More intense absorption of lime at the immediate contact introduces a new mineral among the crystallizing products, viz. melilite. This may arise primarily in the liquid from pyroxene and plagioclase molecules and also by a reaction of already crystallized pyroxene (which becomes resorbed) with the liquid. Ultimately all the plagioclase is used up in the production of melilite, or titanaugite and nepheline, and a melilite-titanaugite-rock, usually with nepheline commonly partly converted to thomsonite, is produced.

Lastly, the melilite-rocks, with their constituent wollastonite, perovskite, and aegirine, arise by interaction of lime and an alkali magma of nepheline-dolerite composition. It is only in these titan-augite-free rocks that perovskite figures as a rock constituent.

The conversion of limestone into lime-silicates in solution in igneous magma is undoubtedly a strongly endothermic process. It is in harmony with this fact that we find the zone of hybrid rocks is limited and local in its distribution. It is clear, however, that sufficient heat was available to incorporate carbonate into silicate solution with the precipitation of abundant diopsidic pyroxene and a resultant modification of the residual liquid. The nature of the assemblages of the exogenous contact-zone and the occurrence of wollastonite with basic plagioelase in place of grossular in the hybrid zone are additional features which point to a high temperature at the contact. It will be remembered that grossular is formed only at a late stage of the reaction process, corresponding to lower temperatures, and then at the expense of melilite. In fact, the series titan-augite-melilite-grossular is a reaction series in the same sense that augite-hornblende-biotite forms a reaction series in normal differentiation. The coarse grain-size of the hybrid zone, together with its vesicular character, is a very striking feature. The carbon dioxide derived from the assimilation of the chalk has, doubtless, assisted in decreasing the viscosity of the melt and permitted the free growth of comparatively large crystals, besides being responsible for the abnormal development of vesicles formed at a late stage of the consolidation.

In conclusion, we may add that the production of a basic alkali residuum in the manner detailed in the foregoing is unmatched among the described occurrences of alkali-rocks. The occurrence of a leucite-bearing zone surrounding inclusions of limestone from pyroxene-andesite at the volcanic centre of Merapi, Java, described by Brouwer,¹ has not yet been clearly worked out to yield any concise idea of the manner of production of this alkali-rock. These two occurrences, among others, have been used by Shand,¹ in a recent paper, to support the assimilation hypothesis of the origin of alkalirocks. To this suggestion the writer would strongly demur, Rather is the Scawt Hill contact-zone, with its very limited hybrid zone and dominance within it of pyroxene-rich dolerite and pyroxenite assemblages, to be taken as an example of the restricted potentiality of igneous magma, to generate alkali types by assimilation.