190 DE. C. E. TILLET ON THE PETROLOGY OP [vol. lxxix,. 



The refractive index of the garnet (as measured in methylene 

 iodide saturated with sulphur) exceeds T78, and the specific gravity 

 of carefully selected crystals is 3 - 94. In none of these rocks clo- 

 the garnets exhibit any signs of incipient alteration. Hornblende 

 is a rare constituent of the composite rocks, and (if present) is 

 characteristically developed in those areas that are free from white 

 mica. Biotite is the most notable absentee, and there is no reason.' 

 to believe that it has ever figured as a constituent of these rocks. 



It was of interest to determine whether the garnet of these- 

 rocks was a normal almandine, or if other bases than ferrous oxide- 

 were present in notable amount. Qualitative analysis showed a 

 high content of ferrous oxide. Manganese was determined 

 colorimetrically, and found to constitute 5 - 5 per cent, as man- 

 ganous oxide, corresponding to a spessartine content of 12"8 percent. 

 The ferrous iron content exceeds 20 per cent., and the garnet must 

 therefore be regarded as a spessartine -almandine. 



There is reason to believe that the early appearance of garnet in 

 these rocks is connected with the manganese content of this 

 mineral. That point will be considered in § YII of this paper ; 

 but Ave may note at this stage that an early appearance of garnet 

 is conditioned by a similar peculiarity in rocks of the Ardennes, 1 

 and in metamorphosed phyllites of the Stavanger district of" 

 South-Western Norway. 3 



VI. The Metamoephic Boundaey. 



The literature devoted more especially to the relations existing 

 between the metamorphosed rocks of the Start area and the 

 Devonian rocks lying on the north has been summarized by Sir 

 Jethro Teall in W. A. E. Ussher's Geological Survey memoir. 

 Ussher himself gave an admirable description of the evidence 

 concerning the boundary, but preferred to leave undecided the- 

 mutual relations of the two groups of rocks. 



The problem resolves itself as follows : — 



(i) Absence of a definite boundary, but a progressive metaniorphism 



increasing in intensity southwards, 

 (ii) A definite boundary (a) of unconformity, and (b) of dislocation. 



Of these solutions the first was supported by, among others, 

 H. B. IIoll, A. R. Hunt, and A. Somervail. These writers 

 regarded the evidence as one of progressive metamorphism, and 

 concluded that the Start group of rocks is satisfactorily explained 

 as consisting of Devonian rocks in a higher grade of meta- 

 morphism. 



On the other hand, Prof. T. G. Bonney and Miss C. A. Raisin 

 have given reason to believe that the Start rocks are separated 



1 A. Renard, 'Les Roches Grenatiferes & Amphiboliques de la Region de 

 Bastogne ' Bull. Mus. Roy. Hist. Nat. Belg. vol. i (1882) p. 10. 



2 V. M. Goldschmidt, ' Die Injektionsmetamorphose im Stavanger- Gebiete ' 

 Vidensk. Selsk. Skrifter, No. 10 (1920) pp. 68-69. 



