36 „ DE. STAMP AjSTD ME. WOOLDEIDGE ON THE [vol. lxxix,. 



the cementing-material in no way similar. Further, if bogen- 

 struktur is o£ any value whatever as a pyroclastic criterion, the 

 nature of the rock is determined beyond doubt. In any case, the 

 limited lateral extent of the bed and the admixture of broken 

 crystals must be considered significant, and there can be little 

 doubt that the breccias mark the site of one or more vents. 



The spilites present no very unusual features. The normal 

 lavas, occurring in pillows, are pale grey-green vesicular rocks, 

 invariably weathering brown. Under the microscope the rock is. 

 seen to be composed, in the main, of a number of felspar-laths ; 

 there is little or no sign of fluxional disposition, which points to a, 

 high viscosity in the magma. This absence of fluxion -structure- 

 has been observed elsewhere in rocks of the spilite-keratophyre- 

 suite. 1 



We follow Dr. F. H. Hatch 3 in restricting the term ' pyro- 

 clastic ' to explosion-products. Dr. Alfred Harker seems to include- 

 flow-breccias under this term. 3 



[There seems to be some difference of emphasis in recent petro- 

 logical writings, in relating the volume of a tuff to the properties 

 of the magma. Prof. A. H. Cox (see p. 33) thinks it probable 

 that the massive tuffs of spilitic series are due to high viscosity of' 

 the magma. He quotes Dr. H. H. Thomas (see above), as stating^ 

 that the keratophyres of Skomer Island show an absence of 

 fluxion-structure which suggests high viscosity. It is to be noted,, 

 however, that Dr. Thomas emphasizes the rarhVy of pyroclastic 

 rocks in the Skomer Series, which thus appears to be an exception 

 to a rule that certainly holds in many Welsh localities. Further,, 

 it is interesting to note that Weinschenk states explicitly 4 that 

 the amount of associated tuff depends on the gaseous content of 

 the magma — ' the more gas, the more dust-like material hurled 

 forth by its escape.' High gaseous content, however, should lower- 

 the viscosity of the magma.] 



On the whole, the rocks are of even grain, although some might 

 be described as microporphyritic. There are, moreover, occasional, 

 true phenocrysts, as well as glomeroporphyritic groups of felspar- 

 crystals. 



The roeks differ from certain typical spilites (for example, those 

 of Pentire Point and many from Merioneth) in being relatively 

 coarse. The felspars are not, on the whole, microlitic, and they 

 are somewhat wanting in the bifurcating and swallow-tail aspect 

 so characteristic of the felspars of these rocks in general. In some 

 cases, indeed, the term ' andesitic ' would not be a misnomer in 



1 See H. H. Thomas, Q. J. G. S. vol. lxvii (1911) p. 195. 



2 ' Text-Book of Petrology : the Igneous Rocks ' 1914, p. 291. 

 * ' Petrology for Students' 1908, p. 277. 



4 ' Fundamental Principles of Petrology ' English transl. by Johannsen,. 

 p. 43. 



