562 MK. W. A. RICHAElJSOtf : A MICROMETRIC [vol. lxxiX, 



the boundaries of what I have defined as the St. Stephen's area 

 (fig. 7, p. 558). In this district two types of rock are recognized 

 in the china-stone quarries, one with, and one without fiuorite ; 

 but the relative proportions of the two could not be determined. 

 Most of the localities are represented by slides of both types ; 

 but, in slides of the ' purple ' variety only, fiuorite bulked often as 

 largely as 7 per cent. 



The fiuorite is secondary, though it is not at all clear what 

 mineral it replaces. Small grains are found in micas and felspars, 

 sometimes in the heart of the crystals, sometimes at their edges, 

 and often arranged along cleavages or other lines, Some 228 

 occurrences of the mineral were counted, and they were situated 

 as follows : — 99 cases in mica; 69 in orthoclase; 54 in plagioclase; 

 and 6 cases in quartz. Each single ' host ' crystal was counted as 

 one occurrence, although it might contain a dozen grains of fiuorite. 

 It is not easy to see any chemical reasons for this order of selection, 

 and the preference shown to mica may be merely due to its greater 

 permeability. 



Finally, the investigation of the first-order grouping of the sum 

 of the pneumatolytic group of minerals shows two belts of high 

 values (fig. 8, p. 558). One occupies adjacent parts of the lietew 

 and St. Stephen's areas, and is mainly due to topaz and fiuorite. 

 The second is in the Hensbarrow area, stretching from Watch 

 Hill to just above Hensbarrow Beacon, and is mainly topaz and 

 tourmaline. Both these zones have a south-westerly and north- 

 easterly trend, and possess this feature in common with tbe 

 principal lodes and the china-clay pits. 1 It must be emphasized, 

 however, that the tourmaline and topaz of these granite-belts are 

 minerals of the primary crystallization, and thus Ave have an 

 indication that the main lines of pneumatolytic activity were to 

 some extent determined, even before the complete consolidation of 

 the granite. 



IV. CORRELATIONS AMONG THE MINERALS. 



At different times correlations have been suggested between 

 such pairs of minerals as biotite and tourmaline, and one of tbe 

 objects of this enquiry was to provide a body of data whereby 

 these relations mio-ht be investigated and tested. 



Quartz and felspars. — When values were plotted on the 

 maps, a striking antipathy between quartz and orthoclase was 

 noticed. Naturally, the principal minerals will fluctuate inversely 

 one to the other, but there appeared to be something more than 

 this. The high zones of orthoclase and quartz interdigitate so 

 persistently, that a high correlation appeared evident. On this 

 account the correlation-coefficient for these two minerals was 

 calculated, with the following results : — 



Correlation-coefficient — 0'69 



Probable error of coefficient O'Oo 



1 ' The Geology of the Country around Bodmin & St. Austell ' Mem, Geol. 

 Surv. 1909 ; map. p. 108. 



