392 Leonard Hawkes—Building up the North Atlantic 
in Fig. 8. It would be impossible to conceive such a regular curved © 
boundary to result if the liparite were intrusive. The marked 
sagging of one of the basalts at 7’ is very suggestive of a lava 
flowing round a hill. We have clear evidence of an acid outbreak 
in the vicinity before the extrusion of the main mass in the lower 
tuff-breccia, and our conclusion is strengthened by the similarity of 
this flow to the post-glacial acid flows of the Torfajokull district 
described by Thoroddsen (9 and 3, p. 155). These lavas are in the 
main composed of a bluish-grey rock with flow-banding and occasional 
columnar structure, and grade at the periphery first into obsidian and 
finally into pumice. The Hrafntinnuhraun stream is a light-grey rock 
40-50 feet thick, with an outer shell of obsidian 5-10 feet thick and 
a final 2-8 feet thick casing of pumice (9, p. 612). The Seydisfjord 
liparite was a surface flow. It formed an eminence on the plain, 
being gradually surrounded by successive basalt lavas and finally 
submerged by them. Estimating the liparite as 300 feet in maximum 
thickness and taking 26 feet as the average thickness of the basalts, 
twelve of the latter were required to cover it. 
55 
Ca. 80ft. 
€-------- 
Fic. 3.—Section, western end of the liparite exposure, Shalanes, Seydisfjord. 
F' = fault throwing two or three feet; ZL = liparite; O = obsidian border 
to L; B= basalt; & = red parting. 
The most important feature is the undenuded character of the 
liparite. There is no sign of any debris round the hill, and the outer 
obsidian shell remains intact. The readiness with which lparite is 
broken up, especially in comparison with basalt, is well known and 
abundantly illustrated in East Iceland, where basic and acid rocks are 
so intimately associated and attain such wide development. Acid rocks 
are fissile and brittle, constantly breaking up and forming scree slopes 
where no plants gain a footing owing to the ever falling stream of 
blocks, but the tough basalt is much more resistant and its scree 
slopes are typically green mantled. Whilst obsidian is more resistant 
than liparite and provides a protective sheath for acid lavas (8, p. 158) 
the unaltered form of this Seydisfjord flow shows that no great 
interval of time elapsed during its submergence by the basalts, and 
also confirms the conclusions reached above respecting the tuff, or 
rather original fragmental, nature of the red partings, for how 
