64 DR. A. HOLMES ON THE PRE- CAMBRIAN [vol. lxxiv, 



may not enter into the composition of the biotite-gneisses. The 

 latter would then he composite rocks like the hornblende-gneisses, 

 and their richness in biotite compared with the gneissose granites 

 into which they pass, would thus find a simple explanation. 



The distribution of radium in the gneisses helps to support this 

 view. The general argument may be stated as follows l : — 



On an average granites contain three times as much radium as 

 argillaceous sediments and mica-schists ; while gneisses, apparently 

 of igneous origin, average only twice as much radium as schists — 

 a fact which harmonizes with the possibility of their composite 

 origin. In all these rocks a considerable proportion of the radium- 

 content occurs in the biotite, but gneisses are generally richer in 

 biotite than granite, thus implying that the biotites of granites 

 are much richer in radium than those of gneisses and richer still 

 than those of schists. Thus, if the biotites of gneisses are 

 relatively rich in radium, a primary igneous origin is suggested ; 

 while on the other hand, if they are relatively poor in radium, a 

 sedimentary or composite origin may be suspected. In neither 

 case, however, does the evidence amount to actual proof. 



For the Mozambique rocks, part of the radioactive evidence has 

 been already given in Tables V & VI. The figures for Nos. 45 a 

 & 45 b are particularly significant. The following averages bear- 

 ing on the subject are given for completeness. In each case the 

 radium-content is stated in grams per billion grams of rock or 

 mineral, and, with the exceptions of averages 1-3 and 9, the 

 materials examined were from Mozambique. 



Table VII. 



Average Radium-Content of Granites, Gneisses and Schists, 

 and of Biotites sefarated from them. 



Grm. — V2 per c/rm. 

 of Bod-.' 



(1) Granites average 3"0 "") 



(2) Gneisses „ 2'0 \- 



(3) Mica-schists „ 1*0 J 



(4) Three specimens of sediment-gneisses and schists... „ 1'34^| 



(5) Fourteen specimens of hiotite-gneiss ,, 1'87 | 



(6 a) Four specimens of biotite-gneiss „ 2*11! 



(6) Three specimens of gneissose granite ,, 2'64 , 



(7) Six specimens of granulitic granite (of later age 



than the preceding) „ 3"24J 



(8) Four specimens of biotite from gneissose granite ... „ 8'0(T] 



(9) Biotites from mica-schists „ 2*00 y 



(10) Six specimens of biotite from biotite-gneiss „ 2'63j 



[1-3. See J. Joly, Phil. Mag. ser. 6, vol. xxiv (1912) p. 694; A. Holmes, 'Science 

 Progress ' 1914, No. 33, p. 35. 4-10. Anabases published for the first 

 time.] 



In every case the results for gneisses — whether the gneisses 

 themselves or their biotites are considered — fall between those for 

 granites and mica-schists. The simplest explanation is that the 



1 Contribution to the discussion on Prof. G. A. J. Cole's paper on ' A Com- 

 posite Gneiss near Barna ' Q. J. G. S. vol. lxxi (1915-16) .pp. 186-87. 



