/ TA L I A N PE TR OL OGICAL SKE TCHES 355 



Closely related to the ciminites, if not to be classed with 

 them, are the interesting rocks of Radicofani. This is a small 

 hill a few kilometers east of Monte Amiata. The rock has 

 been studied microscopically by Weiss,' Bucca,^ and Mercalli.3 

 Two varieties are to be distinguished — a doleritic and an ande- 

 sitic. The former is dark gray to black, very compact and of 

 specific gravity 2.79 (Mercalli). Augite, olivine, and feldspar 

 phenocrysts are visible. Microscopically they resemble the 

 dolerites, numerous large crystals of pale augite, olivine, and 

 plagioclase (with a few of orthoclase, according to Mercalli), 

 lying in a groundmass composed of microlites of plagioclase, 

 magnetite, and apatite in a dark glassy base. The olivines in 

 this rock are generally altered to fibrous green serpentine. The 

 andesitic variety is light gray, of a trachytic aspect and specific 

 gravity 2.683 (Mercalli). It shows phenocrysts of olivine. In 

 thin sections are seen phenocrysts of plagioclase, orthoclase, 

 augite, and olivine (the last altered to a reddish substance on 

 the borders) lying in a groundmass of abundant feldspar micro- 

 lites and magnetite grains, with a rather scanty yellowish glass 

 base. It is seen that while the first variety is mineralogically a 

 basalt, the second is a ciminite. 



Three analyses of these rocks are given in Table I. Vom 

 Rath's (No. 9) is of a specimen of the andesitic variety and is 

 similar to those of Ricciardi. Ricciardi's (Nos. 10 and 11) dif- 

 fer from each other notably only in the silica, which is quite low 

 in the doleritic variety though still higher than in normal basalts, 

 and in the magnesia, which is extremely high in the more basic 

 variety. Thus while the silica percentage is about that of the 

 ciminites, alumina is decidedly lower (about that of the toscan- 

 ites) , iron higher, magnesia and lime very high, and alkalies low, 

 but still higher than in normal basalts, and with potash above 

 soda. From these analyses it seems probable that some of 

 Bucca's and Mercalli's plagioclase is to be referred to alkali 



'Cf. VOM Rath, Zeit. d. d. geol. Gesell., XVII, 405, 1865. 



='BUCCA, Boll. Com. Geol. Ital., 274, 1887. 



3 Mercalli, AUi. See. Ital. Sci. Nat., XXX, 1887. 



