Dear Mr. Bekele:

I-type granitic rocks lithologically reflect subduction processes where gabbroic and ultramafic melts intruded into the arc and back-arc region followed by normal I-type granitoid melts with mantle signature.

There are also post-collisional granodioritic and tonalitic I-type.

Two-mica or only muscovite-bearing S-type granites, developed from granitic magmas of crustal origin. They developed syncollisionally in the course of crustal melting and nappe stacking. It may be a suite of syeno- or monzo-granites with basic predecessors of granodioritic composition.

With kind regards

H.G.Dill

Dear Barsisa,

I think it is not complicated because S-Type granites have been formed during crustal processes and have surficial origins whereas the mafic rocks have deeper origins. So they can not be formed in the same levels.

Regards,

Rahim

Dear Rabim: I don't think that a S-type granite formed by anatexis of crustal rocks is a "surficial" phenomenon, as you stated, as partial fusión of metasedimentary material happens at least at 20 km below the surface or even more. One thing is sure, S-type plutons have no mafic associates, and that's easy to understand, as usually the sedimentary material which undergoes anatexis in collisional orogens represents a former passive margin or miogeoclinal prism, where there is a very great lack of mafic materials. This also explains their peraluminous character (main pelitic protolith), and abundance of micas, garnet, cordierite and aluminosilicates (some cordierite granites could be also products of contamination by crustal rocks, of course).

Regards, Sebastián

Another perspective is that S types are typically peraluminous whereas I types typically metaluminous. Experimental petrology suggests that S type/peraluminous granites are generated from dehydration melting of dominantly micas (pelitoic protoliths as indicated by Sebastian above) whereas I types/metaluminous are generated by dehydration melting of dominantly amphiboles, typical of medium to high grade metabasites.

The source for S-type granite is sedimentary protolith. The mafic mineral phases more abundant deeper part of the earth. Most of the S-type granites formed in the upper part of the crust. The low temperature minerals of Mica and K, Na rich feldspar minerals are more abundant in the upper part of the crust and it is generating through dehydration melting and occurs in continental margin.

I agree in general with the comments above. S type granites are generated by partial melting of metasediments in the mid-crust. But is it possible that some of the heat required for melting might come from a gabbroic melt from the mantle below? In that case there may be be mafic rocks associated with the S-type.

Dear Brian: quite probably a cushion of mafic magma or hot rock located at the roots of continental active margin arcs could supply the necessary heat to cause partial fusion of metasediments, and originate S-type granites, but the mafic material will stay quite a few km below the granitic plutons, since these ascend in the crust to lower levels, therefore they couldn't be associated to each other. Perhaps in fractured crust some basaltic dikes could rise to the surface and intrude the granite plutons, but even this is quite difficult in compressional regimes, such as a active margin. But in collision zones there will be no magma raising from the mantle, so partial fusion will be a matter of rising temperatures during deep burial of supracortical material. Regards, Sebastián

We have published a paper on Neoarchean S-type granites present at terrane boundary zones or collisional zones in the western Superior province, suggesting that the S-type granites may have been formed by partial melting of sedimentary materials in thickened crust due to terrane juxtaposition or collision. We discussed the heat source leading to melting in the paper, including radiogenic heat, crustal scale shear induced heat, and subduction failure magmatism. For details, please see ' S-type granites in the western Superior Province: a marker of Archean collision zones, Canadian Journal of Earth Sciences, 2018,

DOI: 10.1139/cjes-2018-0056'

Dear Xue-Ming: here in Venezuela we don’t have access to International magazines, so I beg you to send me the paper using this way. There are quite large Mesoproterozoic orogenic batholiths of S-type granites in the Cuchivero Province of the Guiana Shield, in Southern Venezuela, maybe they are also markers of ancient collisions! Thanks in advance.

Regards, Sebastián

Thank you very much all for your valuable comments and references

Harald G. Dill Xue-Ming Yang Sebastián Grande Brian P. J. Stevens Vijay Anand Sundarrajan Geoff Grantham

Martin and Piwinskii (1972) pointed out that igneous rocks generated at convergent plate boundaries tend to be characterized by unimodal petrochemistry, whereas those generated in rift zones are characterized by bimodal petrochemistry. This bimodality, especially with regard to silica content of volcanics, appears to provide a powerful tool in term of recognition of ancient rifting events (Sawkins, F.J, 1984). The trace element chemistry of ancient igneous rocks has also been used ,again by analogy with data from modern terrains , as a fingerprint for the tectonic environments o magma generation (Floyd and Winchester, 1975)