410 F. RTJTLEY OU THE MICKOSCOPIC STRUCTURE OE 



due to the usual process of devitrification ; while in some instances 

 it is probable that the rocks have developed a spherulitic structure 

 throughout, during solidification, in which cases they must be re- 

 garded as having been spherulite rocks from the first. Although 

 none of these rocks retain a vitreous appearance, the minute struc- 

 tures which are developed in them are perfectly preserved, and are 

 as clearly demonstrable as they would be in the most recent 

 lavas. 



The close resemblance in minute structure between these Skomer- 

 Island lavas and those of the Yellowstone district in the United 

 States is very striking, although their respective periods of eruption 

 are so far removed from one another, namely — by all the time which 

 elapsed between the deposition of the upper beds of the Lower Silu- 

 rian series and certainly the lowest, if not the highest, beds of the 

 Tertiary epoch. There is also a close resemblance between the 

 Skomer-Island lavas and those of the Snowdon district. The period 

 of eruption in both areas is nearly the same. 



In vol. i. of the £ Memoirs of the Geological Survey of England 

 and Wales ' an account of the rocks of Marloes Bay and Musclewick 

 Bay is given by Sir Henry De la Beche ; but, although Skomer Island 

 is situated close by, he makes no mention of it. At Wooltack Park, 

 on the north of Marloes Bay, there are fossiliferous shales with some 

 sandstones, beneath which it is stated that trap occurs resting on 

 conglomerate and associated with shales. The whole of the igneous 

 rocks of Skomer Island and the adjacent promontory have been 

 mapped as greenstone. 



It is, however, evident that Sir Henry De la Beche generalized to 

 some extent in the mapping of these rocks, as the following extract 

 from the ' Transactions of the Geological Society ' * will prove : — 



" All the north of Skomer Island consists of massive trap, having 

 the character principally of fine-grained compact greenstone, and 

 sometimes approaching to cornean. The small peninsula, however, 

 to the east of the landing-place must be excepted, where some strati- 

 fied rocks of ambiguous appearance occur. The southern part of the 

 island consists of stratified greenstone dipping at about 48° to the 

 south-east. Between this and the greenstone belonging to the north 

 of the island a quartzose cornean, mostly striped, occurs. In some 

 parts of the island hornblende is the prevailing ingredient of the 

 rock.''" 



In the same paper (at p.* 2) Sir Henry states that he considers 

 these traps to be " forcibly intruded amongst the other rocks at a 

 period subsequent to their consolidation ; " and he adds that, in 

 applying the term "stratified" to trap, he only means to imply 

 " that there is a parallelism of texture in the trap, which it has in 

 common with a contiguous rock belonging to some other formation, 

 and that this texture is also parallel to the common surface of sepa- 

 ration between the trap and that other rock." 



* Second series, vol. ii. p. 8. 



