part 4] study of the st. attstell granite. 547 



avoid spots of intense secondary changes, and specimens were 

 chosen only from good exposures where the general character of 

 the rock could be ascertained, and typical granite-specimens 

 selected. Further, all slides were subjected to a preliminary 

 examination, and, if any showed excessive greisenization or 

 secondary tourmalinization, or were in any respect abnormal 

 samples, they were rejected. 



The samples, as finally chosen, represent altogether fifty localities 

 on the main intrusion, besides a few, treated separately, on out- 

 lying satellitic masses. The distribution of these localities is 

 shown by reference-numbers (which are also given with the locality- 

 names in Table IV, facing p. 552) on the map (fig. 1, p. 548). 

 The more northerly areas, outlying among Red and Conce Moors, 

 did not provide good exposures, and all surface-blocks appeared 

 to be modified, so this district is not included in the work. 



It was found that the rock-types fell very closely into four 

 distinct areas, and, in order to facilitate later desci'iption, it will 

 be useful to anticipate the results and define these areas here: — 



(a) The Re tew Area, the small portion of the exposure west of the 



River Fal. 



(b) The St. Stephen Area, included between the River Fal and a line 



drawn northwards through St. Stephen's Beacon. 



(c) The Hensbarrow Area, stretching' from St. Stephen's Beacon 



nearly to the Luxullyan Valley. 



(d) The Luxullyan Area, including' all the mass cropping' out east of 



the Luxullyan River, and a thin strip of country to the west of it. 



The boundary-lines of these districts are shown in fig. 1. and 

 also in most of the later maps. 



Most of the previous general petrographic description is due to 

 Dr. J. S. Flett, and the present investigation continues his work, 

 extending it chiefly in a quantitative direction. I beg leave to 

 extend my thanks to Dr. Herbert H. Thomas, by whose courtesy 

 I have been enabled to examine and measure the slides from this 

 district in the collection of the Geological Survey. 



II. Mineral Composition op the Granite. 



Dr. A. Johannsen's 1 areal method of micrometric measurement 

 was chosen for the work in hand, partly because it adapts itself 

 especially to the analysis of the coarse-grained rocks, and partly 

 because it offers advantages arising from the concentrated study 

 of a single field at a time. Measurement was carried out under a 

 1-inch objective for all slides, for this power enables the amount 

 of accessories to be estimated, without at the same time unduly 

 limiting the field. The scale of measurement should be deter- 

 mined, for the actual areas of the minerals present are sometimes 

 of use in estimating the distribution of inclusions, etc. The mean 

 specific gravities of minerals, as given by Dr. H. A. Miers, 3 were 



1 Journal of Geology, vol. xxvii (1919) pp. 276-85, 

 - ' Mineralogy ' 1902, pp. 556-61, 



