The compositions of all the samples from Hadjer el Hamis fall within the rhyolite field in TAS diagram (Figure 2; [7] ). The peralkaline index (P.I. = molar [Na₂O + K₂O]/Al₂O₃) is higher than unity for the peralkaline rhyolites and low for the metaluminous (Figure 3) rhyolites (P.I.: 0.95).

Rhyolites from Hadjer el Hamis are both, dark green (peralkaline) and grey (metaluminous), with abundant feldspar phenocrysts in a groundmass of fine-grained dark and light minerals. Indeed, samples are porphyritic with partly

fluidal groundmass, marked by the presence of oriented microlites.

Hadjer el Hamis metaluminous rhyolites contain phenocrysts of quartz (up to 1.5 mm), alkali feldspar (~4.6 mm), amphibole (~1.5 mm), clinopyroxene (~1 mm) and Fe-Ti oxides (~1 mm), scattered in the groundmass, consisting of alkali feldspar, amphibole and Fe-Ti oxides microlites, enclosed by a thin glassy phase.

This groundmass is very finely crystallized. Quartz phenocrysts are euhedral to subhedral, with corrosion gulfs. Euhedral and subhedral feldspar phenocrysts are irregular, colourless to yellow brown, with spongy structure. In the interstices left by these phenocrysts occur microlites of feldspar and Fe-Ti oxides. The dark greenish amphibole phenocrysts are sometimes pseudomorphosed or enclosed with fine-grained Fe-Ti oxides. Apatite microcrysts are included in amphibole. Clinopyroxene phenocrysts are green or brown, with anhedral to subhedral shape, overgrown and partly replaced by Fe-Ti oxides.

Major-elements compositions of metaluminous rhyolite from Hadjer el Hamis are more or less similar to those of the peralkaline rhyolites (see Table 5), except that, CaO and Na₂O contents decrease in metaluminous rhyolite. SiO₂, TiO₂, Al₂O₃, Fe₂O₃, MnO contents are roughly constant from peralkaline to metaluminous rhyolites.

Except the slightly enrichment of Rb content from peralkaline (~56.8 wt%) to

Fe₂O₃*: Total Fe as Fe³⁺; L.O.I.: Loss on ignition; D.I.: Differentiation index; P.I.: Peralkaline index; D.L.: Detection limit.

metaluminous (65.5 wt%), the composition of the studied rocks are more or less similar. Whereas Ba, Sr and Y contents are low.

The total REE concentrations ranging from 359 - 387 ppm to 345 ppm, respectively from peralkaline to metaluminous rhyolites. Overall, peralkaline and metaluminous rhyolites from the Lake Chad basin exhibit the similar Zr/Nb ratios (see Table 5), just below 7 (Zr/Nb: 6.5 - 6.9).

Primitive mantle-normalized [9] REE (Rare Earth Elements) patterns for all samples are subparallel (Figure 4). The rhyolites from the Lake Chad basin exhibit high enrichment in LREE and display fractionated REE pattern with (La/Yb)_N ratios ranging from 9.1 to 10.4.

The REE patterns (Figure 4) show negative Eu anomalies, with Eu/Eu* ( Eu / Eu * = Eu N / ( Sm N × Gd N ) ) ranging from 0.53 to 0.59, which may indicate removal of feldspar by crystal fractionation. Primitive mantle-normalized [9] multi-element patterns (Figure 5) show negative Ba, P, Sr and Ti anomalies.

The peralkaline and metaluminous rhyolitic magmas from Lake Chad basin derive likely from the same source, according to their coexistence in the same dome (Hadjer el Hamis) and their similar Zr/Nb ratios. According to the Y vs Nb diagram (Figure 6), the silicic rocks of the Lake Chad are from within plate domain and likely the product of the intraplate magmatism [2] . The values of K/Rb, Rb/Sr, Rb/Ba and Ga/Al ratios (see Table 5; Figure 6) for the Lake Chad rhyolites are similar to those exhibited for the A-type granite [10] . A-type granites ( [11] [12] [13] ) are generated in association with uplift and major strike-slip faulting.

Primary diversity in source rocks, varying amounts of extraction of an earlier melt phase, or tectonic juxtaposition can be suggested as causes of magma heterogeneity [6] .

However, other processes such as different degrees of partial melting or removal of successive batches of magma from a suitable source might also produce ranges of subalkaline to peralkaline magmas. Indeed, the trace-element variations of the Lake Chad rhyolites suite, such as subparallel primitive mantle-normalized REE and multi-element patterns, indicate that they are likely the cogenetic products.

The occurrence of nearly pure albite and sanidine are regarded as reflecting alkali exchange during sub-solidus hydrothermal reactions or as corresponding to late-stage recrystallization at hydrothermal conditions [4] . Quartz phenocrysts with corrosion gulfs were probably generated during ascent to the surface upon eruption, during decompression of the rhyolitic magmas.

The destabilization of hedenbergite phenocrysts core, breaking down into limonite and ilmenite, was probably caused by a sudden increasing of fO₂ in silicic magmas, triggered from a hydrothermal process, associated with F- and alkali-bearing fluids influx; which promoted the enrichment of Na in the hedenbergite rims and the crystallization of arfvedsonite. This hydrothermal process accompanied by an influx of F- and alkali-rich fluids is evidenced by the presence of F-arfvedsonite and the Na-rich phases such as pure albite. Thus, the presence of limonite, in addition to showing variable water content, frequently have other elements reported due to the intimate mixture of the hydrous iron oxides with colloidal silica, as well as small of hydrous aluminum oxides.

The occurrence of phases such as arfvedsonite (K, Ba, Ti), ilmenite (Ti), apatite (P) and alkali feldspar (K, Sr, Eu) in the rhyolitic lavas from Lake Chad basin are consistent with the well-developed negative Ba, P, Sr, Ti and Eu anomalies. The oxygen fugacity of melts, the minerals in fractional process and the stability of minerals in the magma system ( [14] [15] [16] [17] ) are sometime affected by the halogen-rich fluids such as F and Cl. As described for the destabilization of hedenbergite from the Lake Chad rhyolites, hydrothermal fluids can react with pre-existing minerals ( [18] [19] [20] ) to generate others (i.e. arfvedsonite, limonite).