Vörös A. szerk.: Fragmenta Mineralogica Et Palaentologica 11. 1983. (Budapest, 1983)

wt %) and high in CaO (1.82 to 1.96 wt %) and in MgO (4.0 to 4.5 wt %). Crystal Tb-8/2 is similar to Tb-8/3 and 4, but with lower quantities of FeO (30.9 to 31.8 wt %). Similary, zoning patterns may also differ e.g. in Tb-8/3 Mn is increasing towards the border zones, while in Tb-8/5 it is decreasing (Fig. 5). DISCUSSION Garnets originating from andésites In Hungary, andesitic rocks are by far the most widespread among all other volcanics. Thus the Mountains of Visegrád, Börzsöny, Cserhát, Mátra and Zemplén are built essentially of Miocenic andésites and pyroclastics, yet garnetiferous rocks can be rarely met with and they are restricted to those discussed in this paper. The rarity of garnetbearing andésites, dacites and rhyolites compared to the huge volumes of garnet-free rocks of otherwise identical mineralogical and chemical characteristics in marginal volcanic belts have also been pointed out by other workers (e.g. OLIVER 1956 and FITTON 1972 in the Ordovician Borrowdale volcanic rocks, English Lake District; MIYASHIRO 1955 and KANO & YAHIMA 1976 in Japan; as well as BROUSSE et all. 1972 in the Miocene andésites, dacites and rhyolites of the Carpathian Chain in Slovakia). Many of the pétrographie al features mentioned in this paper seem to be present in a surprisingly similar manner elsewhere. Thus, the maximum size of 2 cm of the Királyrét garnets is identical with that of Victoria, Australia (GREEN & RINGWOOD 1968) but most of the Hungarian garnets just like the others are between a few millimeters and 1 cm. The facj that garnets in more acid volcanics are well-developed with regular outlines in contrast with the rounded, resorbed and ragged crystals of the more basic rocks (OLIVER 1956, HILLS 1959, FITTON 1972, BROUSSE & al. 1972) could also be confirmed just as the presence of flow-lines around some garnets. The sensibly different inclusion content of the "basic-type" garnets (Plate III, Figs. 1,2) greater than that of the "acid-types" (Pilisszentlélek, Bajdázó Hill) has not however been signaled from elsewhere. In chemical respect too, the Hungarian garnet megacrysts do not differ from those analysed by other workers in that they are almandine-rich (Aim > 55%), contain 12 to 23% pyrope, 6 to 24% grossular, < 5 mol % spessartine and Ti0

2 generally between 0.3-0.5 wt %. Earlier investigations have shown however that garnet megacrysts originating from different andesitic and dacitic rocks of the world are not entirely homogeneous chemically and at least two main types can be distinguished. The chemistry of garnet megacrysts of this work resembles indeed more closely that of Slovakia (BROUSSE & al. 1972) than that of Borrowdale, England in that the latter has a consistently and markedly lower content of grossular (cf. OLIVER 1956, FITTON 1972). In general there is no consensus if all or any of the garnet megacrysts originating from acid and intermedier volcanics of continental plate margins are cognate or accidental (see GILL 1981 p. 185). Theoretically, the following possibilities may be considered: 1. Garnets are xenocrysts in andésites, dacites and rhyolites and they were detached from some deepseated rock in the upper mantle such as eclogite. 2. Again the megacrysts of garnet are considered as xenocrysts but they originate from shallow crustal metamorphic rock of the basement. 3. Garnets have been created in the volcanics itself by some ill-understood late magmatic (metasomatic, hydrotermal) process such as proposed by GREEN (1915). 4. Finally, garnets are cognate and they were nucleated at greater depths. The first proposition that the garnets were eclogitic in origin can be rejected because of the much higher pyrope content in eclogitic garnets. The third suggestion according to wich garnets would be metasomatic or hydrothermal, is at variance with the textural observations especially with the flow lines and plagioclase corona of some crystals that clearly indicate the presence of garnet in the magma before the eruption. The difficulty to decide in favour of either one of the second or the fourth (metamorphic xenocryst versus cognate megacryst) solutions reside in the fact that available petrological evidence especially in the light of experimental petrology has yielded somewhat ambiguous results for the interpretation. The apparently rare and highly irregular