DISTRIBUTION BRITISH ISLES 601 



roids are compared to Johnston-Lavis's experiment of injecting a dense 

 viscous liquid into another, which produced spherical forms, with narrow 

 necks, that may be severed, leaving the globes detached ; and the structure 

 is thought to be "probably due to the injection of magma into a thick 

 stratum of submarine silt, which occupies . . . the interspaces between 

 the different globes." 



BRITISH ISLES 



Wales. — In 1881 Bonney^^ reported a "compact green rock with indu- 

 bitable spheroidal structure ... in places very distinctly spheroidal," 

 in what had formerly been called a serpentine at Porthdinlleyn, Carnar- 

 vonshire. 



Eaisin^* has described the spheroidal structure in the variolitic rocks 

 of the Lleyn, Carnarvonshire, in which the masses are long and cushioned 

 in appearance and some of them as much as 6 feet in diameter. Broken 

 ones show a radial jointing toward the outer surfaces and a rhomboidal 

 jointing within. Many of them contain "a zone of vesicles radially elon- 

 gated as if gases had been prevented from escaping. Thus evidently a 

 crust was formed at an early period, before the structures were com- 

 pletely solid." The outer portions consist of a chloritic and serpentinous 

 aggregate which is thought to represent a glassy crust. The spheroids 

 are regarded as contraction products and the glassy material as due to a 

 secondary movement of lava which "partially thrust itself in between 

 them," or possibly to differences in the glassy and cryptocrystalline char- 

 acter of the outer and inner portions of the spheroids, respectively, and 

 to the different results of crushing on these portions. The structure at 

 this locality has been more recently described as a pillow lava by Cox and 

 Jones,^^ who call attention to the association of the rock with a peculiar 

 "limestone" with beds and strings of jasper, which in places wrap round 

 the pillows. The structure is extremely complicated and the rocks are 

 probably of pre-Cambrian age. 



The same authors describe from Sam Mellteyrn, near Pwllheli, in 

 Carnarvonshire, a bed of pillow lava 10 to 12 feet thick overlain by about 

 the same thickness of a rock of allied character which is not pillowy, and 

 this is followed in turn by flinty mud stones and micaceous shales, prob- 



23 T. G. Bonney : On the sernentine and associated rocks of Anglesey, with a note on 

 the so-called serpentine of Porthdinlleyn (Caernarvonshire). Quar. Jour. Geol. Soc. Lon- 

 don, vol. 37, 1881, pp. 40-51. 



^ Q. K. Raisin : Varlolite of the Lleyn and associated volcanic rocks. Quar. Jour. 

 Geol. Soc. London, vol. 49. 189.'?, pp. 145-165. 



25 A. H. Cox and O. T. Jones : On various occurrences of pillow lavas in North and 

 South Wales. Rept. Brit. Asso. Adv. Sci., Birmingham meeting, 1913, p. 495 ; also Geol. 

 Mag., vol. X, 1913, pp. 516, 517. 



