part 4] THE PETROLOGY OE THE ARNAGE DISTRICT. 481 



Arnage. 



A. Norite of the Arnage Mass, lepeated from p. 457. 



Ai. Contaminated rock (see p. 463) : Quarry at the roadside, half a mile north-east 



of Mains of Drumwhindlp, Arnage. Analyst, E. G. Radley. ' Summary of 



Progress for 1921 ' Mem. Geol. Surv. 1922, p. 107. 

 A 2 . Contaminated rock (see p. 463) : Carding Hill, 700 yards south-south-east of 



Arnage House, Arnage. Analyst, E. G. Radley. loc. cit. 

 A3. Contaminated rock of Ardlethen Type (see p. 472) : Quarry at the roadside, 



340 yards south-south-east of Ardlethen. Analyst, E. G. Radley. 



Insch. 



B. Hypersthene-gabbro of the Insch Mass. Analyst, E. G. Radley, loc. cit. 



Bi. Contaminated rock, Insch Mass, 400 yards south by east of Easter Saphock, Old 

 Meldrum. Analyst, E. G. Radley, loc. cit. 



Huntly. 



C. Norite of Huntly Mass. Analyst, E. G. Radley, loc cit. 



Ci. Contaminated rock, right bank of the River Deveron, 359 yards east of Castle 



Bridge, Huntly. Analyst, E. G. Radley. 

 CV Contaminated rock, Cuternach, Huntly. Analyst, E. G. Radley; quoted from 



W. R. Watt, Q. J. G. S. vol. lxx (1914) p. 289. 



Le Pallet. 



D. Gabbro, Le Pallet (Loire Inferieure), France. Analysts, Lacroix & Pisaui. 



Average of two analyses. A. Lacroix, Bull. Carte Geol. France, vol. x (1898- 

 99) p. 363, Analysis j. 

 Dj. Contaminated rocks, Le Pallet. Analysts, Lacroix & Pisani. Average of six 

 analyses. A. Lacroix, ibid. p. 363, Analysis i. 



Minnesota. 



E. Olivine-gabbro, Birch Lake (Minnesota). Analyst, A. N. Winchell, Amer. 



Geol. vol. xxvi (1900) p. 181. 

 Ej. Contaminated gabbro (eordierite-norite), Snowbank Lake (Minnesota). Analyst, 

 A. N. Winchell, ibid. p. 303. 



It is now pertinent to enquire into the source of the oxides 

 gained, and the destination of the oxides lost, hy the initial 

 magmas. The answer to both these queries can only be the 

 xenoliths. In Table VI, Analyses X-XIII (p. 482) we have a 

 fine series of analyses which add considerable strength to this 

 view. These four analyses, from the Huntly Mass, are of the 

 initial sediment, a xenolith, the contaminated rock adjacent to 

 this xenolith, and the initial magma. It will be seen that, corn- 

 pared with the initial magma, the contaminated rock is poorer in 

 lime and magnesia, while, compared with the initial sediment, the 

 xenolith is richer in these two oxides. In other words, the sums 

 of lime, magnesia, and potash of the two initial rocks approxi- 

 mately equal the sums of the same oxides of the two resultant 

 rocks. 1 It must be admitted that the other oxides show no very 

 intelligible variation, but this series of analyses demonstrates that 

 certain of the oxides lost by the magma are found in the xenolith s. 



As } r et, I can advance no such striking chemical proof from the 

 Arnage Mass. The analysis of the Arnage cordierite-hornfels 

 (Table VI, Analysis XIV) gives surprisingly low alumina and 



1 For a fuller discussion of this series of analyses, see H. H. Read, '" Geology 

 of the Country round Banff, Huntly, & Turriff ' (Explanation of Sheets 86 

 & 96) Mem. Geol. Surv. Scot. 1923, Chap. vii. 



