84 DR. A. HOLMES OX THE PRE-CAMERIAX [vol. lxxiv, 



Base Camp. The same mineral is a constant, though rare, acces- 

 sory of the Bwibwi residues. 



In almost every case ilmenite was found to be more abundant 

 than magnetite, and this relation was reflected by the residues 

 from rocks. 



Pink or red garnet is present in all sands, and the examination 

 of the rocks shows that it is derived from the gneisses. 



Zircon is everywhere abundant or common, while rutile is 

 generally a rare mineral. In two localities, however, rutile is com- 

 mon, and here neighbouring rocks rich in rutile also occur — the 

 rutile- granite of Nrassi, and the rutile-muscovite pegmatite of the 

 Monapo Limestone district. 



S phene is very variable in its distribution. It is most plentiful 

 in the Monapo and Nrassi sand, where its relation to sphene- 

 bearing hornblende- or augite-rocks is apparent. Apatite is 

 noticeably more abundant in the residues from rocks than in those 

 from the river-sands. 



X. Correlation oe the Gneisses and Granulitic 

 Granites oe Mozambique. 



In a paper on ' The Measurement of Geological Time,' read 

 before the Geologists' Association in 1915, l I suggested that the 

 method of determining geological time by means of the lead- 

 ratios of radioactive minerals mischt 



*■&' 



' be used comparatively for the correlation of igneous intrusions in various 

 parts of the world, and in particular for the correlation of the Pre- Cambrian 

 rocks.' 



I now propose to apply this method to the correlation of the 

 Mozambique gneis'sose granites and of the later Nrassi, Monapo, 

 and Ligonia granites, with corresponding rocks in other well-known 

 Pre- Cambrian areas. It should be noticed that no assumption is 

 here made as to the absolute age of the rocks ; for such an assump- 

 tion is in no way necessary to the argument. If, as has been 

 suggested by Prof. Joly, 2 a proportion of the uranium atoms that 

 formerly existed disintegrated more rapidly than do the uranium 

 atoms still existing at the present time, then it follows that the 

 .ages calculated from lead-ratios, on the assumption that the rate of 

 decay of uranium has not varied during geological time, must be 

 too high. The detailed structure of pleochroic haloes provides 

 some evidence in favour of this view ; but, on the other hand, the 

 thermal condition of the Earth speaks against it. 3 The only 

 assumption made for purposes of correlation is that, if uranium 

 produced lead at a certain rate in Mozambique at any given time, 



1 A. Holmes, Proc. Geol. Assoc, vol. xxvi (1915) p. 308. In this paper the 

 principles underlying the method are briefly described, and full references are 

 given to the original sources of information. 



2 J. Joly, ' The Genesis of Pleochroic Haloes ' Phil. Trans. Eoy. Soc. 

 ser. A, vol. ccxvii (1917) p. 77. 



3 A. Holmes, Kep. Brit. Assoc. (Manchester, 1915) 1916, p. 432. 



