84 REV. EDWIN HILL A:NI) PROF. T. G. BONNET ON THE 



a greater number than usual of the included quartz-grains. Unless 

 a very large amount of the rock be pounded up, discrepancies of this 

 kind are inevitable. About 70 per cent, of SiO., would probably be 

 near the general average. Thus the rock represents an ancient 

 dacite rather than a rhy elite, but, as is often the case in the Forest, 

 is somewhat intermediate in character. The Sharpley rock, when 

 fresh, allowing for the presence of the quartzes, might have had a 

 general resemblance to the older andesite of Krakatoa *. 



4. Field Relations of the Feldar and Sliarpley FocJcsf. — It is re- 

 markable that the areas occupied by these two porphj'roids are so 

 constantly environed by agglomerates. With only one exception J, 

 fragments will be found in the outcrops, nearest to au}- point of 

 their boundaries, and the agglomerates in the spinneys east of the 

 Peldar moorland are among the coarsest in the Forest. 



The agreement between two boundaries of this region and the 

 usual directions of strike suggest that they may be in some way 

 connected with surfaces of deposition. The Peldar rock also occu- 

 pies a position which would agree with its being part of a stratified 

 series, regularly overlying the Sharpley. But this view is nega- 

 tived by the abrupt termination of the former against the agglome- 

 rates and ashes of Eatchet Hill, which are exactly in the usual 

 direction of strike ; it is limited as completely in the opposite direc- 

 tion. If we define the two areas of Sharpley and Peldar rocks by 

 straight lines drawn joining their outermost outcrops, we find the 

 Sharpley rock covering a rude parallelogram §, about three times 

 as long as it is broad, while the Peldar occupies a much ruder 

 parallelogram of only one-third the size, touching along a small 

 portion of its longer side the Sharpley area (see Map, facing p. 80). 

 It is scarcely possible to account for the restriction of these rocks to 

 such strictly limited and peculiarly-shaped areas by faulting. Put 

 these difficulties disappear if the rocks are regarded as two lava-fiows. 



A rock w^hich we found in a spinney (rudely trilobate in form) 

 not quite one-third of a mile W.N.W. of Swanymote Hock, 

 between the northern end of Cademan Wood and the road, may be 

 mentioned here. It occurs near the outside of the southern end of 

 the spinnej^ and can be traced fairly continuously over an area 

 extending about 30 yards in a northerly direction, and perhaps 4 or 

 5 yards wide. Macroscopically, it is rather intermediate in cha- 

 racter between the normal Sharpley and the " purple porphyritic " 

 rock which is common in the agglomerates ; that is, it is like the 

 former, but the individual crystals are not quite so large. Under 

 the microscope the felspars are similar, suggesting in their outline 

 fracture or corrosion. They contain sometimes frequent enclosures, 



* Eeport of Krakatoa Committee of the Royal Society, plate iii. Gg. 1. 



t See Map, facing p. 80. 



\ Drybrook Wood (S. end), where Sharpley rock is seen within a few feet of 

 ashy rocks. 



§ It is also noteworthy that, corresponding to an indentation in the Sharpley 

 boundary at Ratchet Hill, there is a similar indentation in the opposite bound- 

 ary by Gun Hill, as if the parallelogram had been broken across and the parts 

 displaced. 



