part 4] STUDY OF THE ST. AUSTELL GJU.TSTTE. 563 



Evidently, a high negative correlation does exist, and it follows 

 that increase in orthoclase is accompanied by a corresponding 

 decrease in quartz, and conversely. 



Now, the two felspars might be expected to behave similarly 

 with regard to quartz, and one would expect good correlation 

 also between quartz and plagioclase. The following are the 

 calculated results : — 



Correlation- coefficient, quartz and plagioclase --0"26 



Probable error of coefficient 0*09 



There is, indeed, a small negative correlation ; but, for some 

 reason, the fluctuations of quartz are more dependent on those of 

 orthoclase than on those of plagioclase. 



Biotite and tourmaline. — The case of these two minerals is 

 of especial interest, for it has been discussed wherever Cornish 

 granites have been investigated. Mr MacAlister l held, at least 

 for the granite of St. Agnes Head, that tourmaline actually 

 replaces biotite. Dr. Flett 3 is more cautious in reviewing the 

 matter in connexion with the St. Austell mass, and says : — ■ 



' It is evident that the schorl is a primary mineral to a large extent, and 

 though it may rejn'esent biotite, it does not replace it in every case.' 



First, dealing with the district as a whole and assuming that 

 the pale gilbertite of the St. Stephen's area represents the lithionite 

 elsewhere, the calculation gives : — 



Correlation- coefficient +O04 ; probable error 0'09. 



There is, therefore, no correlation at all under these conditions. 

 Secondly, excluding the gilbertite values of the St. Stephen's area, 

 and taking into account onby the coloured micas, we have the 

 following result : — 



Correlation-coefficient —0*2; probable error 0*09. 



Consequently, a very small, but definite, negative correlation does 

 exist between biotite and tourmaline ; but the latter can only be 

 regarded as representing the biotite to a very slight extent. 



Topaz and andalusite. — The suggestion that the topaz of 

 St. Austell is equivalent to the andalusite so common in other 

 Cornish granites is due to Dr. Flett (op. cit. p. 58). The data of 

 the present investigation go far to confirm it. Consider, first, the 

 quantitative conditions as summarized in the following table : — 



District. Cordierite. Andalusite. Topaz. Totals. 



Luxullyan 1-4 tr. 0-1 1-5 



Hensbarrow and Retew ... 0'1 n.f. 1"6 1'7 



St. Stephens 0-2 n.f. VI 1-9 



St. Denis 0-5 M n.f. 1-6 



1 ' The Geology of the Country near Newquay ' Mem. Geol. Surv. 190G, 

 p. 38. 



- ' The Geology of the Country around Bodmin & St, Austell ' Ibid, 1909. 

 p. 59, 



