Tol. 67.] SKOMER VOLCANIC SERIES. 195 



conspicuous vesicles which are usually drawn out slightly iu the 

 direction of flow. These lavas were the most viscous of all but 

 the rhyolitic rocks, and evidence of explosive outbursts may be 

 gathered from the small patches of silicified pyroclastic material 

 (for instance, E 7050) occupying hollows in the surfaces of the 

 lavas (p. 185). 



Porphyritic crystals of acid plagioclase are sometimes present, 

 but seldom reach any considerable size. The ground-mass, as in 

 the case of the trachytes, consists of lath-shaped microlites of 

 albite-oligoclase felspars. The laths are usually well bounded, but 

 in some cases they are somewhat indefinite, both as to their lateral 

 boundaries and as to their terminations. They are usually twinned, 

 sometimes only once, but more often if their breadth will allow of 

 it. Mow-structures are not well shown, the structure being best 

 described as ' felted.' The vesicles, which are numerous and large, 

 are usually filled with chlorite, which may be either the ordinary 

 feebly birefringent variety or the highly birefringent minerals 

 allied to delessite. In both cases the material filling the cavity 

 has a radiate structure. Occasionally the vesicles may be filled 

 with secondary quartz (E 7111) ; but this is unusual, and probably 

 represents surface-silicification. Granules of secondary sphene are 

 abundant, and often good crystals project from the rock into the 

 vesicular cavities. 



Iron-ores in general are less common than in the soda-trachytes, 

 but the felspars are often deeply iron-stained. 



The felspar phenocrysts are albite-oligoclase, but are slightly more 

 basic than those of the trachytes. Some appear to be microper- 

 thitic, and exhibit a curious vermicular structure, due presumably 

 to the alteration or absorption of one constituent. They have, in 

 most cases, suffered a fair amount of decomposition, giving rise to 

 patches of calcite and less frequently epidote. Original ferro- 

 magnesian minerals are very rare, and almost always absent from 

 these rocks. 



The interstices between the microlites are filled with chlorite 

 and a little cryptocrystalline material which might represent 

 residual glass. Generally speaking, chlorite is not very plentiful in 

 the ground-mass, but it may become a prominent constituent, and 

 the same remark applies to the dusty iron-ores (E 7047). 



In composition these rocks are evidently allied to the soda- 

 trachytes, but differ chiefly in containing a greater amount of 

 chloritic material and granular sphene, and structurally in the 

 absence of well-defined fluxion-phenomena. They are so similar in 

 mineralogical character and structure to Dr. Elsden's lime-bostonites 

 that it was thought unnecessary to analyse them specially. Their 

 composition will be approximately represented by the analysis of 

 the Abercastle rock, tabulated on p. 192 (Analysis YI). It will be 

 seen by comparing Analyses IV & VI that they are slightly more 

 basic than the soda- trachytes and richer in lime and magnesia, but 

 that the total alkalies approximate in value. 



