part 2] OF SOUTHERN EYEE PENINSULA. 123 



While in the rocks under discussion the garnet is often developed 

 as a border to original augite, and intervening between this mineral 

 and felspar there is a conspicuous absence of quartz, this reaction, 

 however probable the associations may appear, cannot represent the 

 change which has resulted in the production of garnet. The absence 

 of quartz in all those rocks in which garnet is a constituent is 

 indicative that this type of reaction has pla^^ed no important 

 part. 



Of the reaction of Type 2 we can speak with less certainty. 

 Olivine is known as a constituent of a number of these rocks, but 

 here the associations requisite to conform to this type have not 

 been observed. In some cases this change may, however, have 

 proceeded to completion, and there are certain features which 

 suggest that this reaction has sometimes played a part. Espe- 

 cially is this so where garnet-grains are developed isolated in the 

 plagioclase. It may well be that these garnet-crystals represent 

 original olivine which has reacted with felspar to yield this mineral, 

 the felspar becoming richer in the albite molecule as a consequence. 



There can be little doubt, nevertheless, that reaction 3 has 

 actually been in progress. The typical garnet-pyroxene borders 

 to ' relict ' pyroxene are most readily explained by this reaction. 

 The secondary granulitic monoclinic pj^roxene, which is of a 

 pale-green colour, would appear to represent the diopside of this 

 equation, or at least the pyroxene poorer in aluminous constituents. 

 A secondary pyroxene of hypersthenic composition is an important 

 constituent of these aggregates in many cases, and the evidence 

 suggests strongly that this hypersthene is also a product of 

 the recrystallization of the original augite. It is unmistakably 

 associated with the granulitic boi'ders to an original monoclinic 

 pyroxene, besides being developed in the complexes in which the 

 'relict' pyroxene has ultimately disappeared (see fig. 11, p. 124). 



These coronas around ' relict ' augite, consisting of the three 

 above-mentioned minerals, and in which no constituent takes part 

 other than the coalesced aggregates of magnetite (with some 

 biotite), are most readily explained by a derivation from the nuclear 

 pj^'oxene. It would appear, therefore, that the hypersthene mole- 

 cule is a constituent of the original augite, and the equation 

 expressing this disintegration of the pyroxene is as follows : — 



r a;CaMgSi.,0,. 1 



< Mg-Al,SlO, V^{x- l)CaMgSi,Og + CaMg.Al ,Si,Oi , + yMgSiO, 

 [^MgSiO.3 J (Diopside.) (Garnet.) (Hypersthene.) 



(Augite.) 



The normal pyroxenes of igneous rocks of basic composition can 

 be represented b^^ various mixtures of the ferromagnesian silicate 

 molecules, CaMgFeSi^O^, and Mg-FeAl^SiO^; and further, as Sir 

 Jethro TealU and W. WahP have shown, the inolecule MgSiO^ 



1 Q. J. G. S. vol. xl (1884) p. 640. 



~ Min. Petr. Mitth. vol. xxvi (1907) pp. 1-131. 



Q. J. G. S. No. 306. L 



