322 



The Ordovician Lavas of Mynydd Prescelly. 



In the columu below each, analysis I have shown the mineral 

 composition of the rock calculated from it. That in column IV has 

 been quoted from Professor Cox's paper. The remainder I have 

 calculated from the analyses, and in the case of II and III will be 

 found to differ somewhat from the figures given in the original 

 descriptions, particularly in regard to the contents of anorthite. 



1 believe that in these sufficient account was not taken of the 

 accessories containing lime, such as sphene, apatite, and calcite, 

 which collectively reduce considerably the anaount of this con- 

 stituent available for felspar. In the case of the rock from Cnwc-y- 

 felin, this adjustment is not possible, except in a very general way, 

 in the absence of figures for titanium, carbon dioxide, and phosphorus 

 pentoxide, but the appearance of the rock under the microscope 

 suggests that apatite is present, but certainly not in excess of one- 

 half per cent, and sphene in greater amount, probably over one 

 per cent. If the figures for titanium and phosphorus pentoxide 

 are as low as those shown in column II, there is only sufficient lime 

 left over for 1 per cent of anorthite, making a felspar of the com- 

 position Ab. 98 An. 2, which is not an unreasonable composition in 

 view of the optical characters of the plagioclase seen in the rock. 

 In any case, there is only sufficient alumina for the formation of 



2 per cent anorthite, and this leaves out of account any chlorite 

 present in the rock. 



As regards the value given for quartz in column I, no accurate 

 figure can, of course, be determined in the absence of a full analysis, 

 but an approximate indication can be arrived at in the following 

 manner. A consideration of the other analyses shows that the 

 proportion of silica taken up by constituents other than alumina, 

 lime, and alkalies, lies between 60 and 70 per cent of the total of 

 such constituents. If we take the higher figure of 70 per cent, the 

 4-6 per cent of constituents " not determined " will account for 

 3-2 per cent of silica. The felspars (using the " corrected " figure 

 for anorthite) account for a further 47-3 per cent of silica, making 

 50-5 per cent in all, and leaving 21'6 per cent present as free quartz. 



The foregoing descriptions and analyses show clearly that the 

 affinities of these rocks lie with the soda-rhyolites and trachytes 



