56 ■ I'fCNTTI F.SKOLA. M.S. Kl. 



niinciMl.s arc mentioned. ICcloj^itc occurs at tlu- i)en'j)heries oï the roundec) 

 rock-mass surrounded hy gneiss. The analy.sis of the eclogite proves that 

 its garnet must Ix- much richer in ahnandite than that in the oil vine-rock. 



in this conncctir)!! wc ha\e still to mention the eclogite nodules in 

 the diamond-bearing kimberlite of .South-Africa. These are of 

 the gi'eatest importance for our subject and we shall have repeated occasion 

 to refer to them. The literature dealing with these nodules is very extensive, 

 but we do not uccd to abstract it, as a good monographical treati.se is 

 available'. The following data are quoted from the \olume of Percy 

 Wagner. 



That variety of kimberlite poor in mica, building up the greater part 

 of the pipes in which the diamonds and the nodules are found, has the fol- 

 lowing composition (p. 63): SiO.2 30.95, 'V\0., 2.34, AUO.^ 4.02, Fe203 6.16, 

 FeO 2.66, CaO 8.92, MgO 32.30, MnO tr, NiO + CoO tr, Na.jO 0.35, 

 K,,0 1. 61, H.,0 7.36, CO._, 2.54, P.,0-, 0.15. .Sum 99.36, Sp.g. 2.835. 



The rock is composed of olivine (50 — 70 'N) I, mica, pyrope, enstatite, 

 ilmenite, diopside, perowskite, ores, apatite. As may be seen, this kimber- 

 lite differs from the pyrope-diopside-enstatite-olivine-rocks of Søndmøre only 

 in the presence of mica and titanium minerals, and in the decomposed state 

 of its olivine. The chemical nature is the same. 



Wagner names 21 different mineral associations represented in the 

 nodules. Of the minerals named by him the mica, ilmenite, apatite and 

 cyanite are not met with in the segregations in the dunite near Lien. 

 The other minerals, just as in the Lien inclusions, are garnet, diopside, 

 enstatite and olivine, and most common in , both is the combination 

 diopside-garnet (eclogite) to which may be added enstatite or enstatite and 

 olivine. The presence of diamond in some nodules in the kimberlite is 

 of much interest. 



The nodules arc ellipsoidal, often flattened or almost tabular. Usually 

 they are not more than 5 cm in diameter. 



From this comparison a close analogy between the nodules in the 

 diamond-bearing kimberlite and the band-formed masses in the Norwegian 

 dunite is apparent. Difference occurs chiefly in the shape, but this 

 is naturally accounted for by the different development: in one case 

 volcanic explosions, in the other crystallization in connection with mountain- 

 folding, whereby all the rock-masses have become stretched out and be- 

 come thin bands. 



T. G. Bonney, A. W. Rogers and A. L. Du Toit, F. P. Mennell, and 

 certain other petrographers ha\-e interpreted the nodules in the kimberlite 

 as fragments derived from some formation of eclogite existing at a great 

 depth, while others, as R. Beck, Geo S. Corstorphine, R. Scheibe, F. W. 

 Vogt and, though somewhat doubtfully, Pei'cy Wagner, ha\e seen in them 



' Percy Wagner, Die Diamantführenden Gesteine Südafrikas, Berlin 1909. 



