I see a lot of problems with nomenclature, even things other than ultramafites. In the case of ultramafites, the problem is mostly facilitated by the fact that the rocks are very often serpentinized. Not much can be done with the rare remnants of olivine and other minerals.

Now my colleagues and I are working on the teschenite association around alkaline volcanics. I think, or I believe that the term enjoy is in LeMaitre et al. 2002 defined incorrectly. At the type localities, the rocks do not correspond to analcimic gabbro even by chance. The rocks do not contain plagioclase, so they would be gabbro without plagioclase :-).

The approach to analcime is flawed in my opinion. It is still taken as foid even though it is obviously an alteration product. Analcime is secondary and quite often occurs in rocks in the same position as natrolite. "Natrolitic syenite" is nonsense isn't it :-).

As our rocks are altered, we have obvious problems distinguishing picrites from olivine foidolites. The limit is not even theoretical.

I probably won't even mention lamprophyres yet, it's too complicated.

That's it for now, best regards

Dalibor M.

Yes Dalibor.

The new classification would emphasize that rocks have to be classified for what they are effectively, avoiding any inference on secondary products.

Analcime is not a foid, but a zeolite and its stability at magmatic conditions has never demonstrated.

The original terms reported in the Glossary often refers to rocks that are different from present-day definitions. The classical example is Gabbro.

Cheers,

michele

Of course I know that analcime is zeolite. It is not a foid and its formation in volcanogenic conditions is extremely unlikely. But Le Maitre 2002 still includes it there (page 4, F = feldspatoids or foids). It will need to be treated somehow.

If the new classification really wants to be strictly based on the current state, i.e. the phase composition, then the majority of analcimic rocks will necessarily become metamorphosed rocks belonging to the zeolite facies.

And what about those lamprophyres, where zeolites are sometime normally recognized?

Dalibor M.

DEAR MICHELE

THERE ARE ULTRAMAFIC ROCKS WHICH ARE NOT COARSE GRAINED. AN EXAMPLE ARE THE ULTRAMAFIC LAMPROPHYRES (AILLIKITES, ALNOITES ETC.) WHICH ARE PORPHYRITIC. THEY ARE TERMED "ULTRAMAFIC" BECAUSE OF THE VERY LOW SiO2 THEY CONTAIN (

Thanks Ioannis.

Ultramafic is a concept that should not be mixed with ultrabasic.

michele

YOU ARE RIGHT, MY MISTAKE. HOWEVER, ULTRAMAFIC LAMPROPHYRES ARE BOTH ULTRABASIC (

Michele - are we discussing the common coarse-grained ultramafic rocks (ie, dunite, harzburgite, lherzolite, various pyroxenites), that have historically been considered igneous? If so, I think referring to these as 'metamorphic rocks' is problematic. First, all of these form in equilibrium with melt, either during partial melting (peridotites) or crystallization (pyroxenite, commonly). That many/most have been deformed later does not make them metamorphic. Even if we assume they originally crystallized from a magma ocean, they began as coarse-grained igneous rocks with basically the same mineralogy that they have now. To be considered metamorphic, we should see one mineral assemblage transformed to another in response to changes in PT conditions. Typically metamorphic rocks reflect increases in PT, but of coarse metamorphosed ultramafic rocks go the other way, from high PT to low PT (serpentine, etc). My opinion is that the classic UM rocks of mantle origin (listed above) are most definitely igneous and not metamorphic in any sense. [if a granite is deformed, is it not still a granite?]

Thanks for the comment John.

The classification of a rock should not be associated to its origin. Never. We should have the possibility to give a name to a rock independently on where it has been collected, what is its age, who collected it, the associated tectonic setting and, definitively, what is its petrogenetic origin. A rock should be classified based on petrography and, if available and if necessary and if useful, on its geochemistry. That's all. This is the first principle of the IUGS classification of rocks.

What you write is correct, but it is not directly linked to the choice to classify a mantle rock ultimately to an igneous origin. I agree with you that most of the peridotites and pyroxenites have a mantle origin (i.e., they crystallized from a magma ocean), but you should agree with me that their petrography is almost always typical of metamorphic rocks. All the classical articles dealing with the petrography of igneous rocks use terms exclusive of metamorphic rocks (granoblastic, pyrometamorphyc, mosaic, heteroblastic, and so on).

The problem is indeed here: we know that mantle rocks have a mantle origin, but their ultimate aspect in thin section is typical of metamorphic rocks. IUGS reccommends to use petrography as the first step to classify a rock as igneous or metamorphic. Following this rationale, most of the mantle rocks should be classified as metamorphic. Of course nobody would accept to delete peridotites and websterites from the igneous rock family, but I believe that it is important to write in the third edition of the IUGS book some comment stating that there is a great difficulty to classify this kind of rocks as igneous, considering their metamorphic textures. Harzburgites, dunites, lherzolites, wehrlites, olivin websterites, websterites and pyroxenites will remain in the igneous rock classification scheme, but some "caveat" has to be reported.

A last comment on your last phrase. I am sorry, but I disagreee with you. A metamorphic granite is no longer a granite, but it becomes as a metagranite (i.e., a metamorphic rocks with an igneous origin). It could be classified as gneiss or a migmatite, for example, but certainly a metamorphic granite is no longer a granite, otherwise we would have no metamorphic rocks at all.

Cheers,

michele

Michele, this brings up a more philosophical question: what do we mean by metamorphism? Is it necessary to grow a new mineral assemblage, or is it sufficient to impose a penetrative fabric on a pre-existing igneous mineral assemblage? Especially since formation of a penetrative fabric can occur in the presence of a melt? The answer is above my pay grade, there is clearly a grey area between igneous rocks and metamorphic rocks, as understood classically in the past. In any case, good luck with your efforts; it's a thankless task that someone needs to do. John

It is not a phylosophical question John. Everything is reported in the IUGS book edited by Desmond and Fettes in 2007 (https://assets.cambridge.org/97805218/68105/frontmatter/9780521868105_frontmatter.pdf). What rules is petrography. If the texture is of specific types, the rocks should be classified either as volcanic, plutonic or metamorphic (or sedimentary, of course).

I believe that we can continue leaving the classification as it is, but simply adding some comment referring to the difficulty to properly classify ultramafic rocks.

Cheers,

michele