Evidence from Longonot volcano, Central Kenya, lending further support to the argument for a coexisting CO2 rich vapour in peralkaline magma

SIR-Bailey (1978) argued that if isochemical melting has taken place, the world wide chemical coherence of peralkaline magmas generated below rifted continental regions requires that the lithosphere below each complex be chemically preconditioned prior to melting. He suggested that metasomatic preconditioning through the action of a separate vapour is the only way of producing the required convergence of lithospheric compositions. The chemistry of oversaturated peralkaline obsidians, which are potentially generated from such a source, demonstrates that the vapour must be deficient in H2O and, in order to leave little record of its previous activity in the final volcanic glass composition, must have a low solubility in silicate melts. Bailey argued, using field and experimental evidence, for a vapour phase rich in carbon gases; CO2 at low temperatures and pressures, CO and CH4 at high temperatures and pressures. New textural and mineralogical data from the crystalline peralkaline trachyte lavas of the Quaternary volcano Longonot, Central Kenya, provide further evidence in support of this argument. The Longonot peralkaline trachytes are aphyric to mildly porphyritic. containing a maximum volume of 8% phenocrysts. Anorthorclase (0.7-9.0 mm), fayalitic olivine (0.4-1.2 mm), iron rich clinopyroxene (0.15-1.0 mm) and titanomagnetite (0.05-0.6 mm) comprise euhedral phenocryst phases which are set in a groundmass of alkali feldspar, alkali clinopyroxene, aenigmatite, alkali amphibole and, in the earlier flows, titanomagnetite. Low temperature secondary alteration of the silicates and oxides is absent, the lavas being completely fresh in both hand specimen and thin section. Within the groundmass of many massive and vesiculated lavas, widely scattered patches of fine grained carbonate also exist. The carbonate in each patch occupies an interstitial position with respect to groundmass feldspar, pyroxene, aenigmatite and amphibole. The unaltered nature of the enclosing silicate phases demonstrates that the carbonate is a primary, residual crystallization product in equilibrium (or near equilibrium) with the enclosing silicates. A similar occurrence of primary carbonate associated with analcime lenses was also recorded by Webb (1973) in the flow banded groundmass of a phonolite from South Turkana, Kenya. In the Longonot trachytes, the occurrence of carbonate patches in the groundmass immediately surrounding some vesicles suggests the presence of small residual quantities of a CO2 rich vapour phase. Turning to the phenocryst assemblage, many of the euhedral olivines have suffered partial replacement. In the most advanced example observed (Plate 1 a) the replacing mineral assemblage is alkali pyroxene, an opaque phase and a carbonate arranged in zones parallel to the original crystal boundary. The optical properties that could be determined on the fine grained carbonate, and its association with fayalitic olivine, suggest it may be siderite or ankerite. Many less advanced replacement structures show only alkali pyroxene and a minor opaque phase replacing the olivine phenocryst (Plate Ib). Microprobe analyses from such a structure (Table 1) reveal a pyroxene of sodic hedenbergite composition replacing fayalite. The replacement is interpreted as having taken place during the late stages of crystallization by reaction with a high temperature, coexisting CO2 rich vapour for the following reasons. (i) Several olivines are entirely fresh and unreplaced indicating that replacement is selective and not caused by the wholesale reaction of olivine with the surrounding melt. Even the altered phenocrysts show no rim corrosion, simply marginal replacement from the rim inwards. (ii) The replacing phases constitute a high temperature assemblage, not a low temperature, secondary alteration assemblage which would be expected to include iddingsite, serpentine or chlorite. . (iii) Only olivine has suffered such partial replacement, all other phenocryst and groundmass minerals being entirely fresh and unreplaced. This suggests that the olivine became unstable owing to a crticial change in oxygen fugacity, a change which did not affect the other minerals. A residual CO2 rich vapour phase could have promoted the reaction of olivine. In order to produce the selective