60$ J. V. LEWIS OEIGIN OF PILLOW LAVAS 



gests a genetic connection. Magmatic vapors or solutions rich in dis- 

 solved silicates of soda and other bases would be exlialed from the rocks 

 as they cooled and some would escape into the sea. It is thought that 

 these siliceous waters may have furnished favorable conditions for the 

 increase of radiolarians, diatoms, and other silica-secreting organisms. 

 The direct deposition of the silica among the pillows does not seem to 

 have been considered by these authors, although many of the British and 

 other examples have jaspers and nonradiolarian cherts between the masses, 

 and in some cases the rock itself has been extensively replaced by silica, 

 as shown by Cole.°- 



Ir eland. — In the Lough Mask region of western Connaught the pillows 

 are compressed into rudely polygonal forms, says Geikie,-^^ and the vesicles 

 are greatly drawn out in the direction of the tension (at Bohaun, 9 miles 

 south of Westport). By more shearing the pillows disappear and their 

 crusts are broken up as fragments in a matrix of green schist. The lavas 

 are andesites and more basic rocks of Bala (Ordovician) age. Some are 

 strongly vesicular- and sack-like structure is conspicuous in places. 



An immense development of pillow lavas is described by Gardiner and 

 Eeynolds^^ as the most marked feature of a great series of igneous rocks 

 of the Kilbride Peninsula, County Mayo. The rocks are "spilites'^ asso- 

 ciated and interbedded with cherts, grits, breccias, flow-breccias, and 

 shales of Arenig (Ordovician) age. They are dark green to purplish 

 rocks, somewhat vesicular, some of them markedly so, and the pillows are 

 generally more vesicular around the borders than in the center, or con- 

 sist of concentric layers of vesicular and solid lava. Irregular strings and 

 patches of chert fill spaces between the masses and form a network around 

 many of them. Rarely a large mass of chert is found in the center of a 

 spheroid. These cherts are believed to have been formed subsequently to 

 the consolidation of the rock and chiefly through infiltration. An illus- 

 tration of the Kilbride pillow structure accompanying this paper is re- 

 markably like that of West Paterson and Great Xotch, New Jersey (com- 

 pare plates 15-17). 



On the east coast of Ireland (Dungarvan Harbor and northeastward) 

 crushed cherts, igneous rocks, sandstones, etcetera, occur between the 

 Paleozoic on the south and the schists on the north, in the same relations 

 as along the border of the Scotch Highlands, but in a much broader area 

 and with better opportunity to observe the relations. Besides intrusives, 



•"^2 G. A. .T. Cole : Proc. Roy. Diiblin Soc. new ser.. vol. vii. 1891-1892, pp. 112-120. 

 ^ A. Gelkle : Ancient Volcanoes of Great Britain. London. 1897. p, 252. 

 ^ C. I. Gardiner and S. H. Reynolds : The Ordovician and Silurian rocks of the Kil- 

 bride Peninsula (Mayo). Quar. Jour. Geol. Soc. London, vol. 68, 1912, pp. 75-102, 



